The Economics of Entrepreneurship

This page intentionally left blank

The Economics of Entrepreneurship

Entrepreneurship is an integral part of economic change and growth. Yet until recently it has been largely neglected by economists. In The Economics of Entrepreneurship, Simon C. Parker draws on theoretical insights and recent empirical findings to show how economics can contribute to our understanding of entrepreneurship. The book is based on an earlier work, The Economics of Self-employment and Entrepreneurship (Cambridge University Press, 2004), that has quickly become an essential reference for academics researching the economics of entrepreneurship. Written in a more accessible style, this book contains much that made this earlier work so successful and, in addition, includes improved pedagogical features and new material on the theory of the firm, spin-offs, nascent entrepreneurship, growth-enhancing knowledge spillovers and social entrepreneurship. It can be used both as a reference text for academics from a variety of disciplines and as a textbook for graduate students. s i m o n c . p a r ke r is Director of the Entrepreneurship CrossEnterprise Leadership Centre at the Richard Ivey School of Business, University of Western Ontario. He is also a fellow of the IZA Research Institute for the Study of Labour in Germany and a fellow of the Amsterdam Centre for Entrepreneurship.

The Economics of Entrepreneurship

Simon C. Parker

CAMBRIDGE UNIVERSITY PRESS

Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo, Delhi, Dubai, Tokyo Cambridge University Press The Edinburgh Building, Cambridge CB2 8RU, UK Published in the United States of America by Cambridge University Press, New York www.cambridge.org Information on this title: www.cambridge.org/9780521899604 © Simon C. Parker 2009 This publication is in copyright. Subject to statutory exception and to the provision of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published in print format 2009

ISBN-13

978-0-521-89960-4

Hardback

ISBN-13

978-0-521-72835-5

Paperback

Cambridge University Press has no responsibility for the persistence or accuracy of urls for external or third-party internet websites referred to in this publication, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

To Charlie, Lydia, Olivia and Julia

Contents

List of figures

1

List of tables

xvi

Preface

xvii

Introduction 1.1 What economics adds to the study of entrepreneurship 1.2 Coverage and structure of the book 1.3 Defining and measuring entrepreneurship 1.3.1 New venture creation and nascent entrepreneurs 1.3.2 Small firms 1.3.3 Self-employment/business ownership 1.3.4 Appraisal 1.4 International evidence about entrepreneurship rates in developed countries 1.5 The transition economies of Eastern Europe 1.6 Developing countries 1.7 Appendix: habitual entrepreneurs

Part I 2

page xv

Selection

Theories of entrepreneurship 2.1 ‘Early’ theories of entrepreneurship 2.2 The occupational choice model of entrepreneurship I: homogeneous agents 2.2.1 Definitions of risk aversion and risk 2.2.2 Simple static models 2.2.3 Dynamic models 2.3 The occupational choice model II: heterogeneous ability – the Lucas (1978) model 2.3.1 The Lucas model

1 2 5 6 7 10 10 15 15 19 22 24 29 31 32 36 37 38 39 41 41 vii

viii

Contents

2.4 2.5

2.6

2.7

2.8 2.9 2.10

3

2.3.2 Criticisms of the Lucas model 2.3.3 Variants and extensions of the Lucas model The occupational choice model III: heterogeneous risk attitudes – the Kihlstrom and Laffont (1979) model The very existence of an entrepreneurial option 2.5.1 Theory of the firm considerations 2.5.2 Non-profit-maximising ventures Incumbents’ characteristics: entrepreneurial spawning 2.6.1 Organisational limitations of incumbent firms 2.6.2 Agency cost theories 2.6.3 Learning theories Macroeconomic theories of entrepreneurship and growth 2.7.1 Wealth-based theories 2.7.2 Technology-based theories 2.7.3 Knowledge-based theories Multiple equilibrium models Conclusion Appendices 2.10.1 Technical definitions of risk aversion and risk 2.10.2 A simple ‘hire or outsource’ model 2.10.3 Landier’s serial entrepreneurship multiple equilibrium model 2.10.4 Parker’s human capital multiple equilibrium model

Empirical methods in entrepreneurship research 3.1 Cross-section regression models: sample selection bias and IV 3.1.1 Sample selection bias 3.1.2 Endogeneity and IV 3.2 Cross-section binary models of occupational choice 3.3 Extensions of the cross-section binary model 3.3.1 The inclusion of relative incomes 3.3.2 Multiple occupational choices 3.3.3 Multiple equation systems 3.3.4 Non-binary occupational choices 3.3.5 Heteroscedastic probit 3.4 Time-series models 3.5 Panel-data models 3.6 Entrepreneurial duration models 3.7 Appendices 3.7.1 The reduced form of the two-equation simultaneous equation model 3.7.2 Fraser and Greene’s (2006) heteroscedastic probit model

43 44 49 51 52 57 61 63 65 67 67 68 70 72 74 76 77 77 79 81 82 86 87 87 88 90 92 92 94 94 96 96 96 98 100 101 101 102

Contents

ix

4

Evidence about the determinants of entrepreneurship 4.1 Pecuniary and non-pecuniary incentives 4.1.1 Pecuniary incentives: relative earnings 4.1.2 Desire for independence and job satisfaction 4.2 Human capital 4.2.1 Age 4.2.2 Experience 4.2.3 Formal education 4.3 Social capital 4.4 Risk attitudes, over-optimism and other psychological traits 4.4.1 Risk attitudes and risk 4.4.2 Over-optimism and over-confidence 4.4.3 Other psychological trait variables 4.5 Demographic and industry characteristics 4.5.1 Marital status 4.5.2 Health issues 4.5.3 Family background 4.5.4 Industry characteristics 4.6 Macroeconomic factors 4.6.1 Technology as a determinant of entrepreneurship 4.6.2 Knowledge spillovers and growth 4.6.3 Entrepreneurship and the business cycle 4.6.4 Unemployment 4.6.5 Regional factors 4.7 Nascent entrepreneurship 4.7.1 Characteristics of nascent entrepreneurs 4.7.2 Venture development paths of nascent entrepreneurs 4.8 Dependent starts and firm characteristics 4.9 Conclusion

106 107 109 110 113 113 115 117 119 121 121 124 128 132 132 133 134 138 139 139 140 142 143 147 151 151 152 155 157

5

Ethnic entrepreneurship and immigration 5.1 Discrimination 5.1.1 Discrimination in the labour market 5.1.2 Discrimination in the capital market 5.1.3 Discrimination in the product market 5.2 Positive factors 5.2.1 Positive expected relative returns in entrepreneurship 5.2.2 Ethnic enclaves 5.2.3 Culture 5.2.4 Role models and inculcation of positive attitudes 5.3 Further evidence on determinants of ethnic differences in entrepreneurship

163 165 165 166 170 171 171 172 173 174 174

x

Contents

5.4

5.5

6

Female entrepreneurship 6.1 Some basic facts about female entrepreneurship 6.2 Family factors 6.2.1 Marriage and household production 6.2.2 The impact of children 6.3 Performance of women entrepreneurs 6.3.1 The gender earnings gap 6.3.2 Explanations of the earnings gap 6.3.3 Other performance gaps: growth and survival rates 6.4 Women and entrepreneurial finance 6.5 Conclusion

Part II 7

Immigration and entrepreneurship 5.4.1 Immigrants’ entrepreneurial propensities 5.4.2 The determinants of immigrant entrepreneurship 5.4.3 The effects of immigration on entrepreneurship Appendix 5.5.1 The Borjas (1986) decomposition

Financing

Debt finance for entrepreneurial ventures 7.1 Background and useful terminology 7.1.1 Background 7.1.2 Terminology 7.2 Theories of credit rationing and redlining 7.2.1 Type I credit rationing 7.2.2 Type II credit rationing, redlining and under-investment 7.3 Rebuttals of the credit rationing hypothesis and counter-rebuttals 7.3.1 Rebuttals 7.3.2 Counter-rebuttals 7.4 Over-investment 7.4.1 The de Meza and Webb (1987) over-investment model 7.4.2 Over-optimism 7.5 Conclusion 7.6 Appendices 7.6.1 Bernhardt’s (2000) model of Type I rationing 7.6.2 Stiglitz and Weiss’ (1981) model of Type II credit rationing 7.6.3 Bester’s (1985a) screening model

176 176 177 180 181 181 184 184 187 187 188 189 190 191 194 195 197 201 203 205 205 209 212 212 212 216 216 221 223 223 225 226 228 228 229 230

Contents

xi

8

Venture capital and other sources of finance 8.1 Venture capital and entrepreneurs 8.1.1 Organisational structure 8.1.2 Size of the entrepreneurial venture capital market 8.2 Advantages of venture capital finance for entrepreneurs 8.2.1 Value-adding activities by VCs 8.2.2 Equity finance as an optimal financial contract 8.3 Drawbacks of venture capital and equity finance for entrepreneurs 8.3.1 Factors inhibiting the use of equity finance 8.3.2 Equity rationing, funding gaps and under-investment 8.4 Informal equity finance: business angels 8.5 Other informal sources of finance 8.5.1 Family finance 8.5.2 Micro-finance schemes 8.5.3 Other non-profit-making lending organisations and schemes 8.5.4 Co-operative schemes 8.5.5 Trade credit 8.6 Conclusion

234 235 235 238 239 239 242 244 244 246 248 250 250 251 257 257 259 261

9

Wealth and entrepreneurship 9.1 The role of entrepreneurs in aggregate wealth accumulation and inequality 9.2 The ‘private equity premium puzzle’ 9.3 Entrepreneurial wealth diversification 9.4 Wealth and entrepreneurship: theories 9.4.1 The Evans and Jovanovic (1989) model 9.4.2 The Banerjee and Newman (1993) model 9.4.3 The Aghion and Bolton (1997) model 9.4.4 Newman’s (2007) moral hazard model 9.5 Wealth and entrepreneurship: evidence 9.6 Alternative interpretations of a wealth/entrepreneurship relationship 9.7 Wealth and performance in entrepreneurship 9.7.1 Effects of wealth on venture survival 9.7.2 Effects of wealth on venture investment decisions 9.7.3 Borrowing constraints and profitability 9.8 Evidence relating to Type II credit rationing 9.8.1 What does not constitute evidence of credit rationing 9.8.2 Berger and Udell’s (1992) approach 9.8.3 Other evidence 9.9 Conclusion 9.10 Appendices 9.10.1 Preferences for skewness and entrepreneurship 9.10.2 The Paulson et al. (2006) structural model

263 264 266 268 269 269 271 272 272 273 275 279 279 280 280 281 282 282 283 284 285 285 286

xii

Contents

Part III Performance

291

10

Entrepreneurship, job creation and innovation 10.1 Job creators 10.1.1 Some basic facts about entrepreneurs’ labour demand 10.1.2 Theoretical analysis 10.1.3 Evidence 10.2 Job creation by small firms 10.3 Innovation by small firms 10.3.1 Theoretical arguments 10.3.2 Evidence about innovation 10.4 Conclusion 10.5 Appendix 10.5.1 Carroll et al.’s (2000a) model of labour demand and taxation

293 293 293 294 295 297 300 301 303 305 306

11

Entrepreneurship and growth 11.1 Theories of venture growth 11.1.1 Gibrat’s Law 11.1.2 Jovanovic’s (1982) model of industry selection 11.1.3 Innovation, growth and shakeouts 11.1.4 Other theories of growth 11.2 Evidence about the growth of entrepreneurial ventures 11.2.1 Definitional and measurement issues 11.2.2 Evidence about determinants of venture growth 11.3 Entrepreneurship and economic growth 11.4 Appendix

310 311 311 312 314 315 316 317 319 324 330

12

Entrepreneurial effort 12.1 Unproductive and destructive entrepreneurship 12.1.1 Conceptual issues 12.1.2 Evidence 12.2 Work effort 12.2.1 Hours of work 12.2.2 Explaining entrepreneurs’ work hours 12.3 Ageing, retirement and entrepreneurship 12.4 Entrepreneurial learning 12.5 Appendices 12.5.1 Frank’s model of entrepreneurs’ labour supply and ageing 12.5.2 Parker and Rougier’s model of the entrepreneurial retirement decision

333 334 334 338 340 341 343 351 356 357

306

357 359

Contents

xiii

13

Entrepreneurs’ incomes and returns to human capital 13.1 Measurement issues: tax evasion and income under-reporting 13.2 Other measurement problems 13.3 Evidence relating to entrepreneurs’ relative incomes 13.4 The inequality and volatility of entrepreneurs’ incomes 13.5 Determinants of entrepreneurs’ incomes: theory and methods 13.6 Determinants of entrepreneurs’ incomes: findings 13.6.1 Entrepreneurs’ rate of return to education 13.6.2 Other explanatory variables and extensions

363 364 368 369 372 375 378 378 380

14

Survival 14.1 Closure is not necessarily failure 14.2 Survival rates and their distribution 14.3 Theoretical determinants of survival 14.3.1 The liabilities of newness and smallness 14.3.2 Other factors 14.4 Empirical determinants of survival 14.5 Conclusion

385 386 388 389 389 391 391 396

Part IV Public policy

401

15

Principles of entrepreneurship policy 15.1 The case for and against pro-entrepreneurship public policies 15.1.1 The case for pro-entrepreneurship policies 15.1.2 The case against pro-entrepreneurship policies 15.2 Principles of entrepreneurship policy design 15.3 Entrepreneurship policy evaluation

403

Finance and innovation policies 16.1 Loan guarantee schemes 16.1.1 Organisation of LGSs 16.1.2 Theoretical perspectives on LGSs 16.1.3 Evaluation of LGSs 16.2 Other credit market interventions 16.3 Policies to promote equity finance 16.3.1 Regulatory policies 16.3.2 Taxation policies 16.4 Innovation policy and entrepreneurship 16.5 Appendix: the Keuschnigg-Nielsen (2004a, 2006) double moral hazard model

412 412 412 413 415 417 418 419 421 423

16

405 405 407 408 409

425

xiv

17

Contents

Taxation, regulation and other policies 17.1 Taxation 17.1.1 Income and payroll taxes: theoretical issues 17.1.2 Income and payroll taxes: empirical evidence 17.1.3 Corporation tax 17.1.4 Wealth and inheritance taxes 17.2 Labour and product market schemes 17.2.1 Employment assistance schemes 17.2.2 Procurement and affirmative action schemes 17.2.3 Business advice and assistance 17.3 Regulation 17.3.1 The regulation of entry 17.3.2 The regulation of exit (bankruptcy) 17.3.3 Property rights and corporate governance 17.3.4 The regulation of employment 17.3.5 The regulation of credit markets 17.3.6 Costs of compliance with regulations 17.3.7 Conclusions 17.4 The welfare state, trade unions, enterprise culture and instability 17.4.1 The size of the welfare state 17.4.2 Social security transfers and trade unions 17.4.3 Fostering an enterprise culture 17.4.4 Political and economic instability 17.5 Conclusion

References Index

428 428 429 433 436 437 438 438 440 440 442 442 446 448 450 451 453 454 455 455 457 457 458 459 466 541

Figures

2.1 2.2 2.3

7.1 7.2 7.3 7.4 9.1 11.1 12.1 12.2

Utility functions page 37 Occupational choice with heterogeneous entrepreneurial ability 42 Occupational choice with two occupations, entrepreneurship (EN) and paid employment (PE). (a) PE attracts the ablest entrepreneurs: x > x˜ enter PE (b) EN attracts the ablest entrepreneurs: x > x˜ enter EN (c) Multiple marginal entrepreneurs, x˜ and x˜ 2 48 The supply of and demand for loans 210 Stiglitz and Weiss’ credit rationing model. (a) Banks’ expected returns (b) The market for funds 215 Redlining 216 The use of two-term contracts to separate hidden types 218 The convex outer envelope of returns in entrepreneurship, V(B), and paid employment, R(B) 267 The lag structure of the impact of new business formation on regional employment change 326 Payoffs to entrepreneurship and rent-seeking in Murphy et al.’s (1993) model 336 Work mixing in entrepreneurship and paid employment 343

xv

Tables

1.1 1.2 1.3 1.4 2.1 4.1 4.2 4.3 4.4 4.5 9.1 10.1 10.2 11.1 11.2 12.1 12.2 14.1 14.2 16.1

xvi

International rates of nascent entrepreneurship page 9 Aggregate self-employment rates in some selected OECD countries, 1960–2000 (per cent) 16 Aggregate self-employment rates in some selected transition economies, 1980–1998/99 (per cent) 20 Aggregate self-employment rates in some selected developing countries, the 1960s to the 1990s (per cent) 23 Demand- and supply-side factors influencing the ‘make or buy’ decision 55 Summary of determinants of entrepreneurship 108 Reasons given for becoming self-employed in the UK (per cent) 109 Observations about social capital and implications for entrepreneurship 120 Self-employment rates in the British regions, 1970 and 2000 147 What happens to nascent entrepreneurs 154 Estimates of a simultaneous equation model of entrepreneurship and wealth 275 Characteristics of job creators 296 Innovation and entrepreneurship 301 Determinants of entrepreneurial venture growth 320 Useful empirical growth models 330 Average work hours in Britain by gender and occupation 342 Estimates of the determinants of entrepreneurs’ work hours 346 Summary of economic determinants of venture survival 392 Firm birth and death interactions 396 Features of LGSs in the UK, the USA, France, Germany and Canada 413

Preface

This book follows directly in the footsteps of its predecessor, The Economics of Self-employment and Entrepreneurship (referred to hereafter as Parker, 2004). That book was primarily a monograph for the specialist which could be used secondarily as a course text. By updating the post-2003 literature (to mid-2008) and covering more of the pre-2003 literature, the present book is more comprehensive than its predecessor. It has also been completely rewritten to facilitate its use as a course text for undergraduate and postgraduate courses in entrepreneurship, taught from an economics perspective. In addition, the emphasis of the present book is on entrepreneurship rather than self-employment. For all of these reasons, this book both is, and is not, a ‘second edition’. To help enhance the appeal of this book as a course text I have made two presentational changes to Parker (2004). First, I have wherever possible removed mathematical formalism from the core text. When particular formal models are deemed to be of especial interest, they are discussed in footnotes or in chapter appendices. These can be skipped entirely by readers of a non-technical persuasion. Second, I have made greater efforts to explain economics ideas, terminology and jargon more clearly than before, to help readers not principally trained in economics. Both changes are designed to enhance the accessibility of the text, which I use to teach my Economics of Entrepreneurship students on the doctoral entrepreneurship programme at the University of Louisville, KY. No doubt some unexplained jargon remains, for which I apologise. Compared with its predecessor, the book extends its coverage to discuss several important areas of literature, including social (not-for-profit) entrepreneurship; spin-offs and intrapreneurship; and nascent entrepreneurship. There is also a deeper treatment of several topics covered in Parker (2004), including venture capital, knowledge spillovers, female entrepreneurship, regulation and economic growth. Three areas which I lack the space to cover in detail in the present volume are academic entrepreneurship (including university spinoffs), family firms and entrepreneurship education. Finally, I would like to thank the Social Sciences editor at Cambridge University Press, Chris Harrison, and his colleagues for ongoing support of this ‘economics of entrepreneurship’ project and help in bringing it to fruition. xvii

1

Introduction

The entrepreneur is at the same time one of the most intriguing and one of the most elusive characters…in economic analysis. He has long been recognised as the apex of the hierarchy that determines the behaviour of the firm and thereby bears a heavy responsibility for the vitality of the free enterprise society. (Baumol, 1968, p. 64)

Entrepreneurship is increasingly in the news. Governments all over the world extol its benefits and implement policies designed to promote it. There are several reasons for this interest in, and enthusiasm for, entrepreneurship. Owner-managers of small enterprises run the majority of businesses in most countries. These enterprises are credited with providing specialised goods and services that are ignored by the largest firms. Entrepreneurs generate productivity gains from dynamic entry and exit, which spurs economic development. This comes about either by selection or by competition. Selection involves replacing incumbents who are inefficient or do not satisfy consumer demand by entrants who are more efficient or better meet demand by offering new or better-quality products. Entrants intensify competition and thereby discipline incumbents to provide cheaper or more innovative goods. The most dynamic entrepreneurs pioneer new markets for innovative products, creating jobs and enhancing economic growth. As a striking example, four of the largest US companies by market capitalisation in 1999, accounting between them for about one-eighth of US GDP (Microsoft, Dell, Cisco Systems and MCI), did not exist twenty years earlier (Jovanovic, 2001). Hence it is reasonable to expect that some of today’s new start-ups will grow to become tomorrow’s industrial giants. Even those which do not do so can create positive externalities, for example by developing supply chains that help attract inward investment, or by creating wealth and facilitating social mobility. It is sometimes also claimed that the decentralisation of economic production into a large number of small firms is good for society and democracy, promoting the ethos of a self-reliant and hardy ‘entrepreneurial spirit’. As the wellspring of industrial dynamism, wealth creation and innovation, entrepreneurship is an integral part of economic change and growth. Yet entrepreneurship has only recently come to be regarded as a field. A complete view of it recognises its multi-disciplinary academic underpinnings, drawing from economics, finance, business studies, sociology, psychology and other subjects. This heterogeneous provenance 1

2

Introduction

reflects the multidimensional nature of entrepreneurship, which partly contributes to the elusiveness of the entrepreneur alluded to by William Baumol. 1.1 What economics adds to the study of entrepreneurship

Today, the economics of entrepreneurship is a thriving research field. Although the ‘business studies’ approach to entrepreneurship research remains dominant in terms of field journals, conference activity and academic posts – in other words, in most practical respects – the economics of entrepreneurship literature continues to develop rapidly, generating numerous insights about how entrepreneurship interacts with the economy. However, many non-economists continue to ignore the economics of entrepreneurship literature, while a minority actively denigrates economics, sometimes claiming that the discipline itself is intrinsically unsuited to the study of entrepreneurship. One of the objectives of this book is to rebut the anti-economics arguments, by demonstrating constructively what the subject can and does say about entrepreneurship. It is the author’s belief that anti-economics arguments mainly reflect ignorance about the current state of economics. Before going on to define what the economics of entrepreneurship is, and what it brings to the analysis of entrepreneurship as an academic field, it is worth briefly trying to understand these claims, which can be summarised as follows: 1. Economics (it is alleged) assumes that agents know prices and goods and, automatonlike, optimise resource usage via mathematical rules. But entrepreneurs cannot optimise because they cannot know the prices of goods or services which do not yet exist; they must therefore use heuristics and exercise idiosyncratic judgement. 2. Economics entails the analysis of equilibrium. But the essence of entrepreneurship is that entrepreneurs recognise disequilibrium opportunities and exploit them, destroying the status quo in a ceaseless progression of disequilibrium states. 3. Economics assumes perfect information and competition, so in equilibrium profits are eliminated. But without a profit motive there can be no entrepreneurship; and in the real world imperfect information and imperfect competition prevail so even small entrepreneurial ventures can possess some market power. 4. Economists have chosen not to write the entrepreneur into their models. For this reason the entrepreneur is absent from economics textbooks. But the entrepreneur is central to economic growth so neoclassical growth theory is at best incomplete and at worst misleading. I will take these criticisms point by point. The first one is based on a simple misunderstanding about optimisation in economics. For example, Bayesian methods are ideally suited to modelling situations of entrepreneurial uncertainty (Alvarez and Parker, 2009); and economists have a long tradition of assuming that agents act on the basis of subjective probabilities about the future, even if subjective probabilities differ from objective probabilities. That is, it is recognised that individual agents can and do make mistakes. Although the ‘rational expectations hypothesis’ does not allow agents to

Introduction

3

make systematic errors, this is far from being the only school of thought in modern economics. Economic models are increasingly beginning to incorporate persistent overoptimism, bounded rationality and other cognitive biases into individual behaviours and choices (Minniti and Lévesque, 2008). So nowadays the criticism of hyper-rationality in economics is wide of the mark. The second criticism seems to be based on another misunderstanding, this time about the notion of equilibrium in economics. ‘Equilibrium’ describes a resting point which is eventually obtained after some change occurs. Even if the economy never arrives at a predicted equilibrium, because it is disrupted by another event, it is still helpful to predict the eventual likely outcomes of a given change. As it happens, many economic models now analyse the behaviour of individuals in environments which undergo continual unpredictable change, and deal with equilibrium as a dynamic concept (captured, for example, by the notion of an ‘equilibrium growth path’). A further example relates to innovation, where some economists model the dynamic processes that generate new knowledge and opportunities, rather than taking them to be exogenous as in much of the business studies entrepreneurship literature (King and Levine, 1993; Audretsch, 2003). It is surprising to see some critics continuing to make the third point, which is now hopelessly out of date. As numerous examples in this book attest, imperfect information and imperfect competition play a central role in modern economic analysis, including applications to entrepreneurship. It is essential not to erroneously conflate ‘normal’ and ‘supernormal’ profits. The former is the return needed to keep factors of production employed in their present use. It is not competed away to zero. Economists merely claim that when markets are competitive or contestable, ‘supernormal profits’ (i.e. profits in excess of normal profits) will eventually be competed away. It is a mistake to claim that this precludes exploitation of temporary or even ongoing entrepreneurial opportunities. Indeed, economists would say that one manifestation of entrepreneurship is precisely entry by new firms to compete for profits with incumbents. Other manifestations and definitions of entrepreneurship are also possible, including those based on innovation, managing uncertainty and owning a business; these come well within economists’ambit too (Bianchi and Henrekson, 2005). The first part of the fourth criticism states that economists do not write entrepreneurs into their models, firms or the broader economy. That might have been true when Baumol wrote that ‘the theoretical firm is entrepreneur-less – the Prince of Denmark has been expunged from the discussion of Hamlet’ (1968, p. 66); but with the development of new theories, perspectives and subject areas such as agency theory, personnel economics and game theoretic work on innovation, this is no longer the case. As this book will hopefully show, numerous economics journal articles now treat the entrepreneur as a distinctive economic actor, albeit (to use the terminology of Baumol, 1993b) usually as a ‘firm-organising’ rather than an ‘innovating’ entrepreneur. Baumol (1993b) points out that it is the innovating entrepreneur, and not the firm organiser entrepreneur, whose role is inherently difficult to describe and analyse systematically, and who is really absent from conventional economic models of the firm. As he wrote at an earlier time, ‘one

4

Introduction

hears of no…brilliant innovations, of no charisma or any of the other stuff of which entrepreneurship is made’ (Baumol, 1968, p. 67). But this entrepreneur is doomed to be absent from all scientific theories, economic or otherwise. Criticising economics for this state of affairs is hardly fair. The second part of the fourth criticism has greater substance, however. The terms ‘entrepreneur’ and ‘entrepreneurship’ are still missing from leading economics textbooks in microeconomics, macroeconomics and industrial organisation (Rosen, 1997; Kent and Rushing, 1999). In my opinion these are unfortunate and unnecessary omissions and this criticism is a fair one. In short, and allowing that economists can do more to incorporate the entrepreneur into mainstream textbooks, it is time for the anti-economists to stop caricaturing economics as a subject locked in a 1970s neoclassical time-warp, where economies are characterised by perfect information, perfect foresight, perfect markets and perfect price flexibility. They should instead start to consider what economics can add to our understanding of entrepreneurship. In essence, the economics of entrepreneurship analyses how economic incentives affect entrepreneurial behaviour, and how entrepreneurial behaviour in turn affects the broader economy.1 This is clearly a broad definition and covers a wide variety of issues, as the various chapters of this book amply testify. Consider by way of example a manager’s decision problem of whether to retain employees who develop new innovations within the firm, in order to foster ‘intrapreneurship’, or whether to let them quit and start up in independent entrepreneurship. In this problem, economic incentives are clearly a key issue. Of course, incentives also shape behaviour more generally. Individuals do not have to become entrepreneurs, but choose to do so when the incentives (not necessarily financial) are sufficiently favourable. Indeed, the whole idea of public policy towards entrepreneurship is premised on the notion that government interventions (through taxation, regulation, grants, etc.) affect entrepreneurs’ incentives and thereby their behaviour. One could in fact go further and argue that one cannot fully understand issues like female entrepreneurship, ethnic minority and immigrant entrepreneurship, or entrepreneurial effort without some knowledge of labour economics. Labour economics is also at the heart of participation choices and work participation decisions, as is the microeconomics of incentives. The latter in turn underpin much of the contemporary cutting-edge research on entrepreneurial finance, both debt finance and venture capital. And for their part, these issues cannot be understood without some knowledge of financial economics. Likewise, public economics informs the analysis of public policy towards entrepreneurship. Finally, one can also point out some limitations in some non-economics approaches to entrepreneurship which the economics approach appears well placed to avoid. One is a lack of predictive theory, and ad hoc (or post hoc) hypothesis generation. For instance, it is not much of a theory which merely states that people lacking entrepreneurial intentions are less likely than others to become entrepreneurs; or that individuals who lack access to resources needed to start a business are less likely to actually start a business.

Introduction

5

This type of obvious reasoning, which is deemed uninteresting and therefore unpublishable by mainstream economics journals, can nevertheless be found frequently in other approaches to entrepreneurship. Nor does the economics approach to entrepreneurship content itself with merely listing descriptive and anecdotal evidence which lacks conceptual or causal interpretation and which is not obviously generalisable. By applying its armoury of sophisticated theoretical and econometric methods, the economics of entrepreneurship seeks to extend the understanding of all entrepreneurship scholars, whether they are economists or not. My hope is that this book will help to convince the sceptical reader of this potential. 1.2

Coverage and structure of the book

This book builds on my previous volume (Parker, 2004) by continuing to organise and assess the current state of the branching, acquisitive and rapidly growing literature on the economics of entrepreneurship. The book is intended to serve as a comprehensive overview and guide to researchers and students of entrepreneurship in a variety of disciplines, not just in economics. I have tried to make the text more accessible by providing verbal explanations of analysis in the text, and relegating technical details wherever possible to chapter appendices. This way, the non-mathematically inclined reader can skip the maths altogether without missing any of the major points and insights. For brevity and focus, some topics will be mentioned only in passing and will not be explored in depth. These include academic entrepreneurship (see Rothaermel et al., 2007, for a review); family firms (see Anderson and Reeb, 2003); and entrepreneurship education (see Lee and Wong, 2006). Some alternative approaches will also receive only fleeting attention, including organisational, strategic and managerial decision-making by entrepreneurs; ‘organisational ecology’ and ‘evolutionary economics’ approaches to entrepreneurship; and practical advice (‘how to’ information) to entrepreneurs. Nor will I provide descriptive case studies of individual entrepreneurs, small firms or the industries in which they operate. These topics are ably covered in numerous business studies texts, and will not be repeated here. The book is organised in four parts. The first part deals with selection into entrepreneurship, analysing which people become entrepreneurs and why. Chapter 2 discusses prominent theories in the economics of entrepreneurship, while chapter 3 describes some useful econometric techniques commonly employed in applied research. Chapter 4 provides evidence derived from testing theoretical constructs articulated in chapter 2, using the methods outlined in chapter 3. This evidence base explains what drives some people to engage in entrepreneurship. Chapters 5 and 6 then focus on entrepreneurial selection for some particular groups of interest: ethnic minorities, immigrants and women. The second part of the book analyses the financing of entrepreneurial ventures. There are two major chapters here. Chapter 7 deals with debt (bank) finance, while chapter 8 treats venture capital and other ‘informal’ sources of finance. The third part

6

Introduction

of the book then considers several aspects of entrepreneurial inputs and performance, from the standpoint of individual entrepreneurs and the broader economy. Chapter 9 explores one important aspect of performance, namely wealth accumulation. Chapters 10 and 11 examine several more: job creation, innovation, venture growth and the relationship between entrepreneurship and aggregate economic growth. Chapter 12 analyses the essential venture input of entrepreneurial effort. Chapter 13 discusses entrepreneurial incomes and the returns to human capital, while chapter 14 considers a different performance outcome, namely venture survival. The final part of the book deals with public policy. There are three chapters here. Chapter 15 sets out some principles of entrepreneurship policy. Chapter 16 analyses finance and innovation policies towards entrepreneurship, while chapter 17 concludes the book with a discussion of taxation, labour and product market policies towards entrepreneurship; regulation as it impacts on entrepreneurs; and other important macro issues including the role of the welfare state, trade unions, the role of ‘enterprise culture’ and macroeconomic instability.

1.3

Defining and measuring entrepreneurship

The first and most pressing task is to define entrepreneurs and entrepreneurship. It should be said immediately that there is no general agreement about the meaning of these terms. Some researchers identify entrepreneurs with residual claimants such as small business owners or the self-employed; others restrict their definition of entrepreneurs to business owners who employ other workers. Others again take a Schumpeterian standpoint and argue that entrepreneurship entails the introduction of new paradigmshifting innovations rather than a particular occupation. A popular definition of an entrepreneur in business studies is someone who ‘perceives an opportunity, and creates an organisation to pursue it’ (Bygrave and Hofer, 1991, p. 14). This definition implies that new venture creation is the essence of entrepreneurship. Part of the divide between the economics and business studies approaches to entrepreneurship is attributable to the different definitions of entrepreneurship they utilise. Economists are often content to utilise business owners (in industrial organisation and macroeconomics), the self-employed (in labour and microeconomics) and small firms (in industrial organisation) as working definitions. These definitions all rely implicitly on residual-claimant and risk-taking aspects of entrepreneurship, and facilitate the analysis of incentives, investments, resource allocation decisions and occupational choices. In contrast, many business studies researchers feel there is nothing entrepreneurial about merely being an owner-manager of a small business. They usually prefer to study behaviours entailed by starting a new business, and speculate about cognitive and perceptual constructs entailed with it. Economists tend to eschew this approach as overly subjective, insisting instead on inferring motives only from actual observed behaviour. This is the so-called ‘revealed preference’ principle.

Introduction

7

In empirical work, researchers of all persuasions either have to gather their own data or are obliged to use whatever measure of entrepreneurship comes to hand. The present section presents three of the most commonly used empirical measures, and discusses their advantages and drawbacks. These are new venture creation, small firms and self-employment/business ownership. The final subsection concludes with a brief appraisal.

1.3.1

New venture creation and nascent entrepreneurs

Equating entrepreneurship with opportunity recognition and new venture creation is now standard practice in the business studies approach to entrepreneurship (Shane and Venkataraman, 2000). It is operationalised empirically in the ongoing Global Entrepreneurship Monitor (GEM) data collection exercise (Reynolds et al., 2005). GEM defines as an ‘entrepreneur’ an adult who is engaged in setting up or operating a new venture which is less than forty-two months old. The index of ‘Total Entrepreneurial Activity’ (TEA) is the proportion of the population who are entrepreneurs according to this definition. For example, the 2005 GEM reports that the TEAs of most industrialised countries lie in the 5–10 per cent range. An advantage of GEM data is that definitions and measurement constructs are largely comparable across countries. Thirty-one countries participated in GEM in 2003, involving interviews of some 100,000 adults. GEM also compiles some individual-level data. However, the new venture creation conception of entrepreneurship suffers from several drawbacks. First, many new ventures are mundane, hobby businesses which generate little private or social value. These are included in TEA, despite being far from ‘entrepreneurial’ in a Schumpeterian sense. Second, by excluding businesses over forty-two months old, GEM implicitly categorises even dynamic and enterprising business owners as ‘non-entrepreneurial’. This hardly chimes with popular views about entrepreneurs. Third, focusing only on new ventures excludes growth and exit as part of the entrepreneurship phenomenon, even though many people regard growth and strategic closure (e.g. ‘harvesting’) as essential aspects of entrepreneurship.2 GEM also suffers from limited numbers of covariates and a short time-series. This has resulted in numerous cross-country studies based on as few as twenty or thirty observations. It is unclear what can be learned from such small and heterogeneous samples. Another problem is substantial year-to-year volatility in TEA as a result of excluding older firms. While annual movements of countries up and down the TEA ‘league table’ no doubt make good headlines, it is questionable whether this measure fully reflects the range of entrepreneurial activities. By failing to net out the numerous business exits which occur, the new venture creation and GEM approaches probably overstate sustained, wealth-creating entrepreneurship (Gartner and Shane, 1995). A useful distinction operationalised within GEM is the difference between ‘necessity’ and ‘opportunity’ entrepreneurs. Necessity entrepreneurs are those who face no better alternative to work than entrepreneurship, while opportunity entrepreneurs are those who pursue an entrepreneurial opportunity even though attractive alternative ways of

8

Introduction

earning a living are open to them.3 The 2001 GEM report argued that economic growth is associated with opportunity entrepreneurship while social welfare programmes affect necessity entrepreneurship (Reynolds et al., 2002). A different part of the GEM data collection effort specifically measures ‘nascent entrepreneurs’. Someone is classified as a ‘nascent entrepreneur’ (NE) if they answer ‘yes’ to each of the following questions: (i) ‘Are you, alone or with others, now trying to start a new business?’, (ii) ‘Do you expect to be owner or part owner of the new firm?’, (iii) ‘Have you been active in trying to start the new firm in the past twelve months?’, and (iv) ‘Has your start-up not yet generated a positive monthly cash flow that covers expenses and the owner-manager’s salary for more than three months?’. In addition, respondents must still be in the start-up or gestation phase of an independent firm (Gartner et al., 2004). There are two advantages of studying NEs when exploring the entry process. These are the avoidance of ‘survival’ and ‘hindsight’ biases. Survival bias arises because only about one-half of all aspiring business founders ultimately succeed in creating new organisations which eventually appear in public records (Aldrich, 1999). Firms which ultimately start up are not generally representative of all those which originally tried, and contain relatively few of the smallest and youngest start-up efforts. So inferring aspects of NEs from data sets of established firms is akin to ‘studying gamblers by exclusively investigating winners’ (Davidsson, 2006, p. 3). ‘Hindsight bias’ occurs when established entrepreneurs misreport events which occurred prior to start-up, perhaps because of memory loss or selective re-interpretation of the past. Comparing expectations with outcomes, Cassar (2007) showed that NEs are prone to substantial recall bias. This problem is avoided by interviewing NEs at the time they start up. Two major types of data set focus explicitly on NEs. Both types screen large random samples of households or individuals and use the definition of NE given above. GEM is one; the other is the Panel Study of Entrepreneurial Dynamics (PSED) (Reynolds et al., 2004; Gartner et al., 2004). The PSED originated in the USA, but versions are now available in many other countries too, as well as a new version (PSED II) in the USA. The original US PSED I identified NEs from 64,622 random telephone interviews conducted between July 1998 and January 2000. PSED II identifies NEs from 31,845 random telephone interviews conducted between October 2005 and January 2006. Both GEM and PSED have advantages and drawbacks with respect to measuring nascent entrepreneurship. GEM data on dependent and independent variables are comparable across countries while the various versions of PSED are not. On the other hand, unlike PSED, GEM lacks rich information about individual-level variables. This together with its limited sophistication of measurement makes GEM less useful than PSED for micro-level analysis (Davidsson, 2006). Arguably, both data sets are vulnerable to the charge that despite their emphasis on individual-level factors, their conceptualisations of NE and measurement instruments refer to the venture rather than the person. As many as one-fifth of NEs are starting a new venture for the second, third or nth time (Alsos and Kolvereid, 1998). Another problem is that both data sets probably underestimate entrepreneurial activities by failing to register ‘spontaneous’ starts.

Introduction

9

Table 1.1. International rates of nascent entrepreneurship Venezuela Chile New Zealand USA Australia Brazil Ireland Canada Spain China

0.192 0.109 0.093 0.081 0.066 0.065 0.051 0.051 0.044 0.043

Finland Germany UK Singapore South Africa Italy Netherlands Hong Kong Japan France

0.041 0.035 0.034 0.030 0.027 0.020 0.017 0.017 0.014 0.009

Source: GEM 2003. Nascent entrepreneurship rates by country, extracted from Wagner (2006b, Table 2.1).

Thus, building on earlier longitudinal data analysis by Katz (1990), Henley (2007) finds that the majority of actual transitions observed in Britain are not preceded by declarations of NE status to survey interviewers a year earlier. This might mean that the majority of start-ups are ‘hastily conceived’, having less than a year of preparation. Lack of preparation might in turn explain the high closure rates of many new ventures (see chapters 4 and 14). Let us now turn to evidence about the prevalence of nascent entrepreneurship. According to PSED I data cited by Reynolds et al. (2004), 6.2 per cent of American adults are NEs, corresponding to over 10 million people and 5.6 million new firms. Wagner (2006b, Table 2.1) provides NE rates for all thirty-one countries participating in the 2003 GEM; an abstract of these data appears in Table 1.1. Note the higher estimate of US nascent entrepreneurship in this table compared with the PSED I. Arobust finding both for the USAand many other countries is that men are about twice as likely to be an NE as women.4 But there seem to be few gender differences in venture organisation structures and performance outcomes once NEs are actually engaged in the process (Davidsson, 2006). Another important feature of nascent entrepreneurship is team starts, which involve just over one-half of American NEs. Seventy-four per cent of NE teams comprise two members, followed by 17, 7 and 5 per cent for three, four and five or more members, respectively (Aldrich et al., 2004). Most team members are spouses, with non-spouse teams mainly comprised of people who are similar to each other (‘homophilious’) in terms of ethnicity, gender and occupational background (Ruef et al., 2003). ‘Homophily’ is most pronounced along ethnic and occupational lines, and is even stronger in large teams. Among non-spousal teams, homophily also has a strong gender aspect. Ruef et al. (2003) conjectured that homophily is valued because it embodies familiarity and makes trust easier to establish. This issue is explored further in chapter 4. Despite the relatively recent emergence of this topic, there is already a vast business studies literature devoted to nascent entrepreneurship. Davidsson (2006) reviews some

10

Introduction

of this literature; another useful review, with more of an economics emphasis, is Wagner (2006b).5 Evidence about the characteristics of NEs and their venture development paths are discussed further in chapter 4. 1.3.2 Small firms A more ‘traditional’ measure of entrepreneurship, which pre-dates the 1980s, is the number (or share) of small and medium sized firms (SMEs) in the economy. This definition has the advantage of being easily measurable, since most national statistics agencies tabulate data on economic outcomes by firm size ranges. Nowadays, few researchers or practitioners believe that SMEs are congruent with entrepreneurship. Firm size definitions are arbitrary and industry-specific, and do not obviously represent notions of entrepreneurship. Not all entrepreneurs run small firms, and not every small firm is run by an entrepreneur (Brock and Evans, 1986; HoltzEakin, 2000). The number of SMEs also includes part-time and ‘hobby’ businesses that are not truly entrepreneurial in the sense of being innovative, growth- or profitdriven (Carland et al., 1984). And it can be objected that small business is about firms, whereas entrepreneurship is about the individual, in particular individuals exploiting new opportunities.6 1.3.3

Self-employment/business ownership

The rationale for using self-employment or business ownership as a measure of entrepreneurship is that entrepreneurship is a risk-taking activity. Since all entrepreneurs do not have an employer and own their own business, these measures possess the merit of inclusivity. A problem with them though, as we shall see, is that they can include individuals who are unlikely to be entrepreneurs by other criteria. Self-employment also fails to capture many nascent entrepreneurs. According to GEM data, about 80 per cent of nascent entrepreneurs either have a current job or are managing another business while they work on developing their new business. Hence they are not measured in household surveys (the primary source of self-employment data) as self-employed. And while (as we have seen) just over one-half of all new business creation efforts by nascent entrepreneurs are performed by teams, the conception of self-employment is always at the level of a single individual. Self-employment and business ownership classifications overlap but are not identical. For example, some employees own businesses or shares of businesses ‘on the side’, while other ‘casual’ self-employees do not own a business in any concrete sense. However, for expositional ease, I will talk mainly about self-employment hereafter: the reader should recognise that similar arguments generally apply to business owners too. A practical advantage of using self-employment as a measure of entrepreneurship is that it is widely implemented – both at the individual level within household surveys and at the national level, via the OECD Labour Force Statistics database, allowing

Introduction

11

international comparisons to be performed. The OECD data date back to the 1960s. Useable international comparisons on a large panel of countries go back as far as 1972 (Parker and Robson, 2004) and the data series continues to be published. Although there are some problems of comparability between countries, van Stel (2005) presents algorithms to resolve these problems. In short, given the widespread availability of data on the self-employed in government surveys worldwide, self-employment is one of the easiest measures of entrepreneurship to operationalise in empirical research (Katz, 1990). The self-employed are usually classified formally as individuals who earn no regular wage or salary but who derive their income by exercising their profession or business on their own account and at their own risk. Many of them operate sole proprietorships. This is the simplest form of business organisation. A sole proprietorship is an unincorporated business owned by one person, which makes no distinction between the assets of the business and the personal liabilities of the owner. Likewise, partners of an unincorporated business are usually classified as self-employed. It is sometimes helpful to partition the self-employed into employers and own-account workers (the latter work alone); and into owners of incorporated or unincorporated businesses. Most self-employed people in most countries own unincorporated businesses, rendering their incomes liable to income tax. For example, CPS data suggest that 62 per cent of self-employed Americans run unincorporated businesses (Bregger, 1996). Incorporated business-owners tend to be older and employ others, which might explain why incorporation rates among new (and typically small) entrants to self-employment are about one-half of this rate (Evans and Jovanovic, 1989). Incorporated businesses face corporation tax (see also chapter 17). To appreciate the economic importance of self-employment and business ownership, consider the following facts:

1. Around 10 per cent of the workforces in most OECD economies are self-employed. The figure climbs to about 20 per cent when individuals who work for the selfemployed are also included (Haber et al., 1987). Two-thirds of people in the US labour force have some linkage to self-employment either by having experienced self-employment themselves; by coming from a background in which the household head was self-employed; or by having a close friend who is self-employed (Steinmetz and Wright, 1989). By the end of their working lives, about two-fifths of theAmerican workforce has experienced at least one spell of self-employment (Reynolds and White, 1997). 2. Between 80 and 90 per cent of all businesses in the economy are operated by selfemployees (Acs et al., 1994; Selden, 1999). 3. Many employees in industrialised countries claim that they would like to be self-employed. For example, according to Blanchflower (2000), 63 per cent of Americans, 48 per cent of Britons and 49 per cent of Germans declared a preference for self-employment over paid employment.7

12

Introduction

The remainder of this subsection focuses on the problems that can be encountered when researchers use self-employment data. Many of these problems boil down to difficulties of classifying self-employed people accurately in sample surveys. A further problem, of the measure being too broad to capture genuine entrepreneurship, is discussed at the end. The first classification problem is that in many countries (including the UK and USA), owners of incorporated businesses are defined for tax purposes as employees rather than self-employed. Yet they resemble in all other respects (e.g. in terms of residual claimant status) the ‘self-employed’. In cross-country comparisons of self-employment rates, it is therefore important for consistency to count these individuals as self-employed (van Stel, 2005). Second, for some individuals, legal and tax-based definitions of self-employment are sometimes at variance with each other (Harvey, 1995; Dennis, 1996). In law, the issue comes down to whether there is a contract of service or a contract for services. The first indicates paid employment, the second self-employment. For example, Harvey (1995) cites the UK legal case of Young and Woods v. West, whereby the criteria for a worker being under a contract of service includes the worker not determining their own hours, not supplying their own materials and equipment, not allocating or designating their own work, not being able to nominate a substitute to work in their place, and not setting their rate of pay (see also Leighton, 1983). UK legislation IR35, announced in 1999, has given these restrictions on self-employment legal force, to help Her Majesty’s Revenue and Customs better tackle tax evasion by ‘spurious self-employed’ workers (see below). This measure has helped to harmonise the UK’s legal and tax definitions of self-employment. The tax system can have other perverse effects as well, leading to wrongful classification of self-employment. For example, married partners might declare themselves self-employed purely to enable the chief income earner to minimise tax payments by ‘splitting’ tax liabilities with the spouse who pays tax at a lower marginal rate. Canadian evidence suggests that this problem entails over-estimation of the number of genuinely self-employed people (Schuetze, 2006; and see chapter 17). Third, many household surveys stipulate that self-employment status is to be self-assessed by the survey respondents. This can lead to further differences in the classification of workers, compared with legal and tax-based definitions (Casey and Creigh, 1988; Boden and Nucci, 1997). Partly for this reason, some surveys (e.g. the UK Labour Force Survey (LFS) and the US Characteristics of Business Owners (CBO)) use tax-based definitions of self-employment. Fourth, as intimated above, many people who create new businesses are categorised as employees in national surveys because wage and salary work is currently their primary source of income. Thus self-employment typically only measures entrepreneurship once a venture is up and running. It does not capture the process of creation itself, which some (but by no means all) scholars identify most closely with entrepreneurship. Fifth, there is a ‘grey area’ between paid employment and self-employment. Some workers classified as self-employed, with apparent autonomy over their work hours,

Introduction

13

are effectively employees. These are peripheral workers subordinated to the demands of one client firm (Pollert, 1988; Harvey, 1995). They are sometimes referred to as ‘dependent self-employed workers’ (Böheim and Muehlberger, 2006; Cabeza Pereiro, 2008). According to an ILO Report cited by Böheim and Muehlberger (2006), these workers ‘provide work or perform services to other persons within the legal framework of a civil or commercial contract, but…are dependent on or integrated into the firm for which they perform the work or provide the service in question’. Defining a dependent self-employed worker as a sole proprietor with only one customer, Böheim and Muehlberger (2006) report that the dependent self-employed amount in number to about one-tenth of the number of non-dependent self-employed in the UK. Böheim and Muehlberger argue that the dependent self-employed lose their rights under labour law, are entitled to fewer social security benefits and are often (but not always) beyond trade union representation and collective bargaining.8 According to these authors, the dependent self-employed tend to be white, older men, with modest education and job tenures but persistence as dependent self-employees. They are concentrated in construction, financial services and skilled trades. It is sometimes argued that employers actively seek to organise their workforce in self-employment contracts, to cut costs and to avoid their social obligations (Parker, 2007b). Böheim and Muehlberger (2006) estimate that fewer worker entitlements in self-employment, together with the removal of employer payroll contributions for selfemployed workers, can reduce an employer’s labour costs by 20–30 per cent. Firms do appear to exercise some discretion about the mode of employment contract they offer. For example, Moralee (1998) reported that in response to new laws penalising companies that misclassify employees as self-employed to avoid tax payments, the number of ‘employees’ in the construction industry apparently increased sharply while the number of ‘self-employed’ workers decreased sharply. This debate is related to one on contracting out of labour by large firms, a phenomenon that was thought to have been particularly pronounced in the 1980s, a decade when self-employment rates in several countries increased dramatically. The theory behind this issue is discussed further in chapter 2. Other examples of workers in the ‘grey area’ between employment and selfemployment include commission salespersons, freelancers, homeworkers and teleworkers, workers contracted through temporary employment agencies and franchise holders. According to Moralee (1998), 13 per cent of the UK self-employed in 1997 were homeworkers, with little change in self-employed homeworking taking place over the 1990s. Moralee (1998) also observed that 61 per cent of teleworkers were self-employed. It is difficult to know whether to classify franchisees as ‘entrepreneurial’ or merely ‘managerial’. Much of the distinction comes down to the behaviours and actions of the franchisees themselves. Felstead (1992) argued that many ‘self-employed’ franchisees are effectively directed by their franchisor, who holds most of the ownership rights and has a senior claim on profits. With relatively little discretion about the format of their business, one could certainly claim that some franchisees resemble branch managers more than independent entrepreneurs.9 Likewise, one could argue that a

14

Introduction

self-employed retailer who is pressurised into stocking the goods of mainly one goods manufacturer also has limited discretion about the nature of his or her business. Like other self-employed workers, franchisees face risk and uncertainty. In their case, not only is their income risky, but there is also a possibility that the franchisor will either go out of business or refuse to renew the franchise agreement at the expiration of its term.10 Two other ‘grey’ categories include unpaid family workers who work in a business run by a self-employed person; and members of worker co-operatives, who are not obviously either employees or self-employed workers in the conventional sense of the terms. Both groups tend to be more numerous in developing than in developed countries, and in rural than in urban areas. According to Bregger (1996, p. 5), ‘Unpaid family workers are persons who work on a family farm or in a family business for at least 15 hours a week and who receive no earnings or share of the profits of the enterprise.’ Blanchflower (2000) detected substantial variation within developed countries in the proportion of self-employed workers who are unpaid family workers, being one-third in Japan, compared with one-seventh in Italy and just 1.7 per cent in the USA. As Blanchflower points out, it may not make sense to simply discard unpaid family workers from the self-employment count, since they often share indirectly (e.g. via consuming household goods) in the proceeds generated by the business. Worker co-operatives are also relatively uncommon in developed countries, although there are exceptions, such as France, Italy and Spain.11 Finally, one should stress that the self-employed are a very diverse group of people. They include small business proprietors, whose businesses may or may not be incorporated; independent professionals such as doctors and lawyers; skilled manual and craftworkers; farmers; some categories of homeworkers; and ‘labour-only’ subcontractors such as construction workers (Meager and Bates, 2004). Guiso et al. (2004) claim that a measure of ‘pure’ entrepreneurship should exclude from self-employed samples all professionals, artisans, plumbers, electricians and other tradesmen. This can be quite easily implemented in practice using micro-data containing detailed occupational codes.12 What are we to make of this discussion? It seems that self-employment is a convenient but imperfect measure of entrepreneurship. It has the merit of versatility, capturing a stock measure of entrepreneurship in a cross-section sense, as well as flow measures if entries to and exits from this state are identified longitudinally. Arguably, future work should seek to refine the self-employment definition to produce a more focused group of subjects who are in some sense ‘entrepreneurial’, yet which is still wide enough to cover a variety of different aspects of ‘entrepreneurship’ – a concept whose theoretical dimensions will be analysed in chapter 2.

1.3.4 Appraisal We have seen that there is no common definition of entrepreneurship, with three main empirical measures dominating the literature. Although this might seem to be

Introduction

15

unsatisfactory or even problematic at first blush, Baumol (1993b) argues that disagreement about ‘who is an entrepreneur’, or ‘what is entrepreneurship’, scarcely matters in practice. Many of the definitions are ‘complementary rather than competitive, each seeking to focus attention on some different feature of the same phenomenon. It is surely appropriate for two authors to differ in the aspects of the entrepreneur that they find it useful to emphasize, and the choice of one author need constitute no impediment to the pursuit of a slightly different side of the subject by the other’ (Baumol, 1993b, p. 198). Indeed, claims that the field cannot progress without resolving this issue in a single agreed definition seems to be a counsel of despair. It is unlikely if not impossible that any single measure of entrepreneurship could or even should ever be regarded as portraying all the nuances of entrepreneurship. One could certainly argue that the existence of more than one practical entrepreneurship measure is a positive advantage rather than a drawback. The researcher has greater choice to employ an empirical measure that relates more closely to their theoretical construct, whatever that may be. Different measures contain different information, which make them complements rather than substitutes. Some researchers have recognised this, suggesting that researchers should use a mixture of entrepreneurship measures in empirical work (Gartner and Shane, 1995). It is more likely in practice, however, that researchers will either take an agnostic approach and use whatever measure is available, flying as it were under a flag of convenience; or that they will remain wedded to their own strict disciplinary paradigms and continue doggedly to use only one type of measure. For clarity, the following convention is adopted in this book. At the conceptual level, the terms ‘entrepreneur’and ‘entrepreneurship’will be used throughout. At the practical level, where issues of measurement, estimation and policy are involved, I will try to refer to the particular measure that was either used in applied research or which appears most suitable. That could be any of new venture creator, nascent entrepreneur, small firm, self-employed or business owner. Finally, it is sometimes convenient to disaggregate entrepreneurs into ‘habitual’ and ‘novice’ subgroups. ‘Habitual’ entrepreneurs run multiple businesses, either sequentially (‘serial’ entrepreneurs) or simultaneously (‘portfolio’ entrepreneurs) (Westhead and Wright, 1998, 1999). In contrast, ‘novice’ entrepreneurs are entrepreneurs who establish a single business. The appendix to this chapter contains more information about these distinctions, which are slowly but rightfully beginning to attract the attention of entrepreneurship researchers. 1.4

International evidence about entrepreneurship rates in developed countries

This section documents some international evidence about levels of, and trends in, entrepreneurship in developed countries. An advantage of using a long time span of data is that it can capture the effects of fundamental changes in important environmental influences (e.g. technological changes), which shorter time spans cannot do.As explained in the previous section, the data for which the longest time spans are available relate to

16

Introduction

Table 1.2. Aggregate self-employment rates in some selected OECD countries, 1960–2000a (per cent) 1960

1970

1980

1990

2000

13.83 18.81 22.68 34.25 15.86c 30.51 25.93 21.87 21.79 38.97 7.28

8.94 13.20 19.18 31.29 14.09 22.17 23.59 16.65 17.90 35.59 7.36

8.70 9.74 17.18 21.67 16.16 16.79 23.26 12.23 10.03 30.47 8.05

8.50 9.52 14.05 25.64 15.05 13.26 24.53 9.64 9.24 26.27 13.32

7.33 10.66 11.34 28.53 13.49 10.56 24.48 10.46d 7.03 20.49 11.34

B. Non-agricultural workers USA 10.45 10.17 Canadab Japan 17.38 Mexico 23.01 Australia 11.01c 16.90 Franceb Italy 20.60 15.08 Netherlandsb Norway 10.14 23.60 Spainb UK 5.89

6.94 8.33 14.44 25.20 10.00 12.71 18.97 12.02 8.61 21.55 6.27

7.26 7.05 13.75 14.33 12.73 10.71 19.20 9.06 6.53 20.63 7.11

7.51 7.40 11.50 19.89 12.34 9.32 22.24 7.84 6.12 20.69 12.41

6.55 9.46 9.35 25.48 11.72 8.06 23.21 9.25d 4.83 17.69 10.83

A. All workers USA Canadab Japan Mexico Australia Franceb Italy Netherlandsb Norway Spainb UK

Notes: a Self-employment rates defined as employers plus persons working on their own account, as a proportion of the total workforce. b Includes unpaid family workers. c 1964 not 1960. d 1999 not 2000. Source: OECD Labour Force Statistics, issues 1980–2000, 1970–81 and 1960–71.

aggregate self-employment rates. Hence the results in this and the following sections present evidence of levels and trends in aggregate self-employment rates. It should be borne in mind however that to the extent that different countries use different definitions of self-employment, cross-country comparisons of levels should be treated with caution. Table 1.2 demonstrates considerable diversity in the levels and time-series patterns of self-employment rates across OECD countries. Results are presented separately for both all workers and non-agricultural workers, reflecting the substantial difference agriculture makes to measured self-employment rates in many countries. Indeed,

Introduction

17

most researchers tend to exclude agricultural workers from their definitions of selfemployment, on the grounds that farm businesses have very different characteristics from non-farm businesses. The agricultural sector tends to decline as an economy develops, which may distort self-employment trends in other sectors of the economy (Blanchflower, 2000). I will first single out for discussion a few countries which have received substantial attention from the research community, and then chart some broad historical trends. Consider the USA first. Although its self-employment rates look relatively low, the primary source for the US entries in Table 1.2 is the CPS Monthly Household Labour Force Survey, which unlike comparable surveys in other countries excludes incorporated business owners as self-employed (see the previous section). Were these business owners to be included in the figures, the US rates would rise by about 2.5 percentage points, making them more similar to other industrialised countries (Manser and Picot, 1999). Looking at trends, if agricultural workers are included, the US self-employment rate has been in continual decline since at least 1960. According to Steinmetz and Wright (1989), this decline can actually be traced back to the 1870s, when the selfemployment rate stood at just over 40 per cent of the labour force; and it underwent an especially steep decline between 1950 and 1970 (see also Table 1.2). This led Phillips (1962) in his historical study to characterise US self-employment as a ‘shrinking world within a growing economy’. Phillips (1962) predicted that self-employment would eventually serve as a refuge only for older, handicapped or unproductive workers as a safeguard against unemployment. Few experts would agree with this characterisation today. Excluding agricultural workers, Table 1.2 shows that a revival in US self-employment occurred during the 1970s and 1980s, a finding that has been highlighted by several authors.13 But unlike previous authors, Table 1.2 reveals that this revival in nonagricultural US self-employment apparently came to an end by 2000, by which time the non-agricultural self-employment rate had fallen back to below its 1970 level. Thus the last two decades of the last century can best be interpreted as witnessing selfemployment rates that stabilised after a long period of decline punctuated by a brief window in the mid-70s to mid-80s of modest but consistent growth (Arum and Müller, 2004). The US experience is mirrored by France, which has also seen its overall selfemployment rate decline steadily since the start of the twentieth century (Steinmetz and Wright, 1989). It should be noted that low-skilled workers in France such as childcarers and gardeners are still mainly classed as wage earners (with multiple employers) rather than self-employed (Amossé and Goux, 2004). This could partly explain the relatively low self-employment rates in France, although not its inexorable decline. However, France’s downward trend is not replicated in every OECD country. For example, Table 1.2 shows that both Canadian self-employment rates exhibited U-shaped patterns, increasing particularly strongly in the 1990s, where it was concentrated among own-account workers and accounted for the bulk of overall job growth.14

18

Introduction

In contrast, both measures of the UK self-employment rate increased dramatically in the 1980s, a finding that attracted substantial research interest when it was first discovered (Hakim, 1988; Campbell and Daly, 1992). But the expansion of UK selfemployment ceased in the 1990s, at the end of which decade male self-employment rates declined, while the female self-employment rate was static (Ajayi-obe and Parker, 2005). Empirical work by Taylor (2004) attributed the increase in self-employment during the 1980s to an increase in the inflow rate; and the cessation of the trend increase in the 1990s to a rise in the outflow rate. According to Storey (1994a), the UK historical trend in self-employment was one of steady decline between 1910 and 1960, followed by increase from 1960 to 1990, with the rate in 1990 being similar to that in 1910. The data in Table 1.2 reveal a striking variety of patterns over 1960–2000. Four countries (Japan, France, Norway and Spain) had steadily declining self-employment rates. Six witnessed a revival in self-employment at some point within the period (USA, Canada, Mexico, Italy, the UK and the Netherlands); and one (Australia) had a relatively stable self-employment rate. This prevents grand conclusions from being drawn about trends or ‘revivals’in entrepreneurship across the developed world, at least in the second half of the last century. In contrast, between the mid-nineteenth and mid-twentieth century the picture seems to have been less ambiguous, with declining self-employment shares in Germany, Switzerland, Canada, the UK, the USA and France.15 One reason why it is hard to draw sweeping cross-country conclusions is that different countries experienced different events which masked any common trends (Blanchflower, 2004). It does seem clearer, however, that the last two decades of the previous century saw increased heterogeneity of self-employment in most developed countries, as witnessed by the decline of the traditional artisan class and the rise of the managerial-professional class as well as unskilled self-employment. At the same time, more people had greater exposure to self-employment over the course of their lives (Arum and Müller, 2004). Consistent with this, UK evidence points to greater instability in the labour market with more movements into and out of self-employment, which involved more frequent spells of self-employment for a greater number of people (Taylor, 2004). Finally, I will say a word about the relative importance of small firms in the economy. The overwhelming majority of US businesses employ fewer than five workers.16 There are high rates of business formation and dissolution among small firms, especially in industries like retailing, where low capital requirements make entry easy and keep profits modest. The aggregate number of small businesses grew in the USA in the post-war period, but their relative economic importance (measured in terms of their employment share or share of GDP) declined over that period. Recent evidence suggests that the share of private non-farm GDP accounted for by small businesses in the USA has now stabilised, at around 50 per cent over the last two decades (SBA, 2002a). That the decline was not greater is mainly attributable to the growth of the service sector, in which small firms are disproportionately concentrated. We will say more about the role of changing industrial structure in chapter 4.

Introduction

19

1.5 The transition economies of Eastern Europe

Of special interest are the so-called transition economies of Eastern Europe, which underwent a switch from communist central planning to a more market-based system at the end of the 1980s. In the words of Earle and Sakova (2000, p. 583): ‘it is difficult to imagine a regime more hostile towards self-employment and entrepreneurship than the centrally planned economies of Eastern Europe’. These regimes fixed prices and wages, placed restrictions on hiring workers and acquiring capital and levied confiscatory taxes on entrepreneurs. This is reflected in the low non-agricultural selfemployment rates, shown in Table 1.3, which held in Poland in 1980 and in Russia in 1992.17 Part of the interest in studying the transition economies is that they serve as a testbed for the strength of dormant entrepreneurial vigour that could be released after market liberalisation. According to Blanchflower et al. (2001), Poles topped the list of respondents to a survey of 25,000 people in twenty-three countries asking whether they would prefer to be self-employed to being a wage worker: 80 per cent responded in the affirmative. Blanchflower et al. (2001) concluded that there is no shortage of potential entrepreneurs in the transition economies. Numerous opportunities to make profits and run businesses became available after liberalisation that had been proscribed before. Many of these opportunities involved supplying services and consumer goods which were not provided by state-run producers. Rates of new entry and profits tended to be very high in transition economies just after liberalisation, before declining over time as niches became filled and competition intensified (McMillan and Woodruff, 2002). Estrin et al. (2006) describe in detail the kinds of entrepreneurial opportunities that emerged from the collapse of central planning. They include the liquidation of vertically integrated state enterprises, which created niche markets; privatisations, which released fixed and working capital that could be used to finance entrepreneurship; the rise of grey and black markets, which responded to the lack of consumer goods; and the emergence of politically networked middlemen with access to crucial production inputs. This led to a dramatic expansion in the share of small firms in many transition economies throughout the 1990s, towards Western levels. The incentive to become an entrepreneur in the transition economies was no doubt further enhanced by declining opportunities in wage employment and growing unemployment as the sprawling former state-run companies began to contract in the 1990s. Perhaps surprisingly in this regard, GEM data identifies relatively few ‘necessity’-type entrepreneurs in eastern Europe, compared with the West (Estrin et al., 2006). Although legal barriers to private enterprise and self-employment came down after 1989, bureaucracy and the limited rule of law continues to stifle productive entrepreneurship in many transition economies.18 For example, Baumol (1990, 1993a) contends that an important factor which retards productive entrepreneurship in countries that are prone to bureaucracy and lawlessness is the scope to engage in privately profitable (but socially unproductive) rent-seeking activities or (socially destructive)

20

Introduction

Table 1.3. Aggregate self-employment rates in some selected transition economies, 1980–1998/99 (per cent)

Poland (all workers) (non-agricultural) Russian Federation Czech Rep. (all workers) (non-agricultural) Hungary (all workers) (non-agricultural) Slovak Rep. (all workers) (non-agricultural)

1980

1990

25.44 3.37

27.17 9.16

1992

1994

1998/99

10.18 10.25 16.93 16.94 6.21 6.49

22.44 11.70 5.29 14.59 14.50 14.56 12.81 7.80 8.00

0.76

Notes: Self-employment rates defined as employers plus persons working on their own account, as a proportion of the total workforce. Source: UN Yearbook of Labour Statistics, various issues. ‘1998/99’ is 1998 for Poland and 1999 for the Russian Federation.

organised crime. If payoffs to unproductive and destructive activities are sufficiently high, entrepreneurs will rationally divert effort from productive innovation to exploit them. This general issue is discussed further in chapter 12. In the remainder of this section, I will discuss some specific obstacles to entrepreneurship in the transition economies, as well as some empirical evidence on the issue. Researchers have highlighted the following obstacles to entrepreneurship in transition economies (Estrin et al., 2006): • Individuals lacked wealth when communism ended because they had hitherto been prevented from legally accumulating assets. This reduced personal financing of new ventures. The problem was exacerbated by a ‘missing middle class’ and underdeveloped financial markets reluctant to lend to new small ventures. Formal trading was further restricted by reliance on a variety of informal rules and behaviour which were remnants of the communist past. • Massive economic uncertainty because of major restructuring of entire industries simultaneously, and ongoing imbalances between supply and demand. Recessionary and inflationary conditions in the early years of transition restricted purchases of shares of newly privatised companies for many citizens and led to a chaotic business environment which fostered corruption and mafia operations. • Weak and fragile institutions resulted in uneven, lengthy and arbitrary enforcement of contracts, which created fertile ground for organised rent-seeking. Entrepreneurs in these economies remain unusually vulnerable to corruption because they cannot rely on secure property rights enshrined in the rule of law and usually lack bargaining power against powerful gangster-capitalists and the public bureaucracy. • Negative social attitudes towards private business and entrepreneurship.

Introduction

21

Some commentators claim that although rates of entrepreneurship are comparatively low in all transition economies, advanced Central and Eastern European countries face superior entrepreneurial prospects compared with former Soviet republics and Russia in particular (Estrin et al., 2006; Aidis et al., 2008). Although one must be careful about making cross-country comparisons because of data limitations and the sparseness of the empirical research base, there are several reasons why Russia in particular faces even steeper challenges to promote entrepreneurship than its neighbours (Table 1.3). First, as we will see in chapter 4, one of the strongest influences on an individual’s likelihood of becoming an entrepreneur is having a parent who was an entrepreneur. Russia experienced communism for the longest of all the former Soviet countries, and suffered from not just one but three missing generations of entrepreneurs. Second, Russia has a limited productive heritage, given its tradition of trading natural resources for goods produced by satellite countries in the former USSR. Its reliance on natural resource extraction continues to this day. Third, Russia has an adverse age profile, with low average rates of male longevity (only fifty-eight years), a declining population and relatively few young people. In asking whether government can help address these problems, one should recognise that the role of public policy in the promotion of entrepreneurship is quite subtle. McMillan and Woodruff (2002) argue that government may not be initially important to the success of liberalisation and entrepreneurship in transition economies, but becomes more important as these economies begin to develop. At first, entrepreneurs can find ways around missing institutions. The establishment of reputations can substitute for legal enforcement of contracts between new firms and suppliers. Entrepreneurs play a repeated game with their suppliers and creditors, so incentives to cheat each other are diluted. This is especially important in the stages immediately following liberalisation, when search costs for new partners are the greatest if trade breaks down. Also, trade credit can substitute for bank credit, and reinvestment of profits can substitute for outside equity. But this can only go so far. Entrepreneurs need access to reliable contracts if they are to expand their companies into markets where they have fewer personal contacts; and they need financial regulation to access bank loans and outside shareholding if they are to grow and achieve economies of scale. This is where governments can play a key role by upholding contracts and property rights, effectively setting the ‘rules of the game’. Government may also be able to tackle impediments to investment such as corruption and mafia, costs of licensing and registering new starts, as well as macroeconomic instability. McMillan and Woodruff (2002) argue that governments in transition economies differ considerably in these respects. Turning to the evidence base about personal characteristics of entrepreneurs in transition economies, Djankov, Miguel et al. (2005) surveyed 400 entrepreneurs and a similar number of non-entrepreneurs in Russia in 2003, and looked for significant differences in characteristics and backgrounds. Entrepreneurs in Russia tend to be more willing to gamble with their incomes than non-entrepreneurs. They also have stronger links with other entrepreneurs. For example, the proportion of parents, aunts and uncles running a business is 42 per cent among entrepreneurs but is only 20 per cent among

22

Introduction

non-entrepreneurs. Also, more than one-quarter of Russian entrepreneurs claim that having friends who were entrepreneurs influenced their decision to become one too. Broadly similar findings hold for China in most but not in all respects (Djankov et al., 2006). According to Aidis et al. (2008), Russian entrepreneurs are disproportionately older, male and well educated compared with their Western counterparts. Where Russian and Chinese entrepreneurs appear to differ most is in terms of the influence of business networks. In principle, business networks may have positive or negative impacts in transition economies. They can be positive because they can foster trust and substitute for missing institutions. But they can have negative effects if they create barriers to entry and promote business inertia. Estrin et al. (2006) report positive effects in China, but negative ones in Russia. In particular, Russian business networks strongly promote the interests of established firms at the expense of new entrants (Aidis et al., 2008). It is noteworthy that relatively little is known about ethnic and female entrepreneurship in transition economies. This situation, which contrasts with what is known about them in developed economies (see chapters 5 and 6), constitutes an important gap in our knowledge base which needs to be closed. 1.6

Developing countries

In the early post-war period, researchers attached great importance to fostering entrepreneurship in developing countries. In the words of W. Arthur Lewis (1955, p. 82): ‘economic growth is bound to slow unless there is an adequate supply of entrepreneurs looking out for new ideas, and willing to take the risk of introducing them’. According to Leff (1979), interest in the issue had dwindled by the 1970s. Leff asserted that this was because of a perception that the entrepreneurial problem had been ‘solved’, with high rates of real output growth serving as evidence of entrepreneurial vigour. Subsequently, slower growth, high rates of population growth, widespread failures of state-owned enterprises, constraints on public sector employment and the spread of free-market beliefs reactivated interest in promoting entrepreneurship in developing economies. Table 1.4 shows that, on average, developing countries have markedly higher selfemployment rates than developed countries. Nevertheless, Table 1.4 also reveals considerable variation in these rates. It has been noted elsewhere that many Asian countries have very high self-employment rates, sometimes exceeding half the workforce. This includes the Philippines and Indonesia (Le, 1999) and Nepal (Acs, Audretsch and Evans, 1994). According to Acs, Audretsch and Evans (1994), the self-employment rate in Nepal in the 1980s exceeded 85 per cent, compared with only 3.1 per cent in Botswana. However, in contrast to claims by some previous researchers that the trend in developing countries is away from self-employment (Blau, 1987; Schultz, 1990), the evidence in Table 1.4 reveals that no such trend can be generally established. It is noteworthy that the evidence in Table 1.4 relates to relatively small developing countries. Conspicuous by their absence are the ‘BRIC’ (Brazil, Russia, India and

Introduction

23

Table 1.4. Aggregate self-employment rates in some selected developing countries, the 1960s to the 1990s (per cent)

Africa Mauritius Egypt Americas Bolivia Costa Rica Dominican Rep. Ecuador Asia Bangladesh Korean Republic Pakistan Sri Lanka Thailand

1960s

1970s

1980s

1990s

13.03 29.19

10.30 26.14

n.a. 28.20

16.72 27.19

n.a. 20.78 44.79 42.97

48.86 17.10 29.44 37.81

40.27 21.80 36.46 37.27

34.81 24.70 37.11 37.03

7.33 44.04 21.94 26.94 29.83

45.56 33.92 46.90 22.90 29.65

38.83 33.07 55.95 24.74 29.75

29.59 28.02 48.18 26.68 28.45

Notes: Self-employment rates defined as employers plus persons working on their own account, as a proportion of the total workforce. Includes agricultural workers. Source: UN Yearbook of Labour Statistics, various issues. ‘1960s’ is either 1960, 1961, 1962 or 1963 for all countries; ‘1970s’ is some year between 1970 and 1976; ‘1980s’ is 1980 or 1981 except Ecuador (1982), Costa Rica (1984) and Bolivia (1989); and ‘1990s’ is some year between 1990 and 1996.

China) economies. The last two countries in particular are rapidly becoming major players in the world economy, making it likely that entrepreneurship researchers will pay them greater attention in future (Bruton et al., 2008). Why are self-employment rates so high in developing countries? I will return to the broad issue of economic development and entrepreneurship later in the book; here just three specific factors will be mentioned. First, the data in Table 1.4 (and those used in most other studies of these countries) include agriculture. The agricultural workforce is dominated by the self-employed; and agriculture usually plays a prominent role in the economies of developing countries. Second, much labour market engagement in urban locations in developing countries (at least one-fifth according to Yamada, 1996) takes the form of informal self-employment.19 Third, high self-employment rates may reflect limited development of formal economic and financial markets. For example, Leibenstein (1968) argued that entrepreneurship in developing countries often simply involves overcoming constraints caused by poor economic and financial infrastructure, and is quite basic in nature. This viewpoint is related to the long-standing dual labour market model of development, comprising a formal urban sector in which employees earn premium wages, and an informal rural sector in which entrepreneurs receive belowaverage incomes (Lewis, 1955; Harris and Todaro, 1970). That model predicts that as poor economies develop, labour will move from the informal to the formal sector, with

24

Introduction

a decline in self-employment rates. However, evidence from the field refutes both the prediction of higher average incomes in paid employment in developing countries,20 as well as the prediction that workers shift from self- to paid employment as they gain experience. This evidence could have an important bearing on entrepreneurship policies in developing countries. If entrepreneurship really does improve living standards and reflects ‘development’ rather than ‘distress’ (or ‘opportunity’ rather than ‘necessity’) then proentrepreneurship policies seem appropriate, at least in principle. An interesting insight into this question comes from rural China, which has experienced a huge rise in rural self-employment, with up to 30 million new rural self-employed emerging between 1988 and 1995 alone. Almost 40 per cent of all new off-farm jobs in rural China now belong to the newly self-employed. Self-employment as a proportion of the rural labour force was 4 per cent in 1981 but had grown to 16 per cent by 1995 (Mohapatra et al., 2007). Mohapatra et al.’s (2007) survey evidence shows that relatively capital-intensive self-employment is driving this trend; low-productivity self-employment (handicrafts and custom labour services) has declined in importance. Furthermore, education and wealth are positively correlated with entry into and persistence within high-productivity self-employment. This evidence is indicative of a ‘development’ rather than a ‘distress’ story. Arguably, therefore, economic theories of growth and development offer greater scope for explaining high rates of entrepreneurship in developing countries than stories of deprivation alone. I will return to this issue in subsequent chapters of this book. 1.7 Appendix: habitual entrepreneurs

According to Wright et al. (1998), nearly two-thirds of British entrepreneurs are novice entrepreneurs whose current business is their first business. One-quarter are serial entrepreneurs who owned a previous business but no longer own it. The remaining 10 per cent are portfolio entrepreneurs who own two or more businesses contemporaneously. Together, serial and portfolio entrepreneurs are referred to as ‘habitual’ entrepreneurs. From their interviews of 9,553 Norwegians, Alsos and Kolvereid (1998) reported that 205 individuals (2.2 per cent of the randomly sampled population) were nascent entrepreneurs (NEs). Of these NEs, 64 per cent were novice entrepreneurs, 20 per cent were serial entrepreneurs and 16 per cent were portfolio entrepreneurs. These rates are similar to British estimates reported by Westhead and Wright (1999). Univariate comparisons reveal similar personal characteristics by NE type, with a slightly higher proportion of young people and women among novice NEs. More recent Swedish evidence suggests that human and social capital are both associated with portfolio entrepreneurship, and that habitual entrepreneurs tend to exploit new opportunities in the form of new organisations – unlike novice entrepreneurs who are more likely to incorporate them within their current business (Wiklund and Shepherd, 2008). US evidence shows that 20 per cent of new starts in the 1987 CBO were businesses operated by portfolio entrepreneurs (Astebro and Bernhardt, 2003), while the founders

Introduction

25

of 20–30 per cent of all small business owners had started other businesses in the past (Baird and Morrison, 2005, p. 2329; Hyytinen and Ilmakunnas, 2007). According to Hyytinen and Ilmakunnas (2007), Finns with both greater aspirations and capabilities to become serial entrepreneurs are significantly more likely to do so. In summary, despite some agreement across studies, estimates of the proportions of habitual versus novice entrepreneurs still vary (Ucbasaran et al., 2006, pp. 464–65). This may suggest a need for more representative survey data. Some, but by no means all, serial entrepreneurship entails entrepreneurs bouncing back from business failure. The German Regional Entrepreneurship Monitor reveals that about one-fifth of active NEs are ‘failed’ entrepreneurs (Wagner, 2003b), while Landier (2004) cites a pilot study by the SBA showing that 50 per cent of bankrupt ‘Chapter 7’ business owner filers between 1989 and 1993 went on to start another new business in 1993. These rates are high enough to suggest that stigma from failure is not very widespread in either Germany or the USA. Wagner (2003b) observed that the probability of ‘taking a second chance’ is significantly lower among older and more risk-averse entrepreneurs located in regions characterised by high past business failure rates. This probability is however significantly higher for individuals who know other entrepreneurs and who are located in regions with a high share of NEs. Several conceptual and empirical problems arise in studies of habitual entrepreneurs. Rosa (1998) observes, for example, that portfolio business activity is not necessarily entrepreneurial: diversification is a standard efficiency-enhancing strategy of large, established firms as well as of new and small ventures. Rosa (1998) also criticises the ‘cross-section’ emphasis on the latest businesses of habitual entrepreneurs, rather than the performance of entrepreneurs evaluated across all of their businesses. At the time of writing, the body of theory about serial entrepreneurship is relatively under-developed. One is left instead to conjecture about its determinants. One interesting possibility of a non-economic nature is that some people become serial entrepreneurs in order to escape corporate bureaucracy. Entrepreneurs grow their firms to the point where the internal bureaucracy becomes unbearable, at which point they sell them and start new small-scale companies. Thus it is possible that the inevitably recurring nature of both bureaucracy and the emergence of new opportunities in the firms that entrepreneurs found, rather than any innate entrepreneurial ability or characteristics, might explain the serial entrepreneurship phenomenon (Dobrev and Barnett, 2005). More prosaically, tax advantages might provide an alternative explanation of serial entrepreneurship. In some jurisdictions, entrepreneurs receive tax credits on their positive income ‘draw’ if their company profit is negative up to a given time limit – after which they lose the favourable income tax treatment even if the company continues to make losses net of the draw. This gives entrepreneurs incentives to draw excessive incomes out of their businesses, putting their companies into the red, before closing and re-forming the same ventures as ‘new’ businesses after the tax-credit time-limit expires. As owners of an ostensibly new business, such entrepreneurs can continue to claim the tax credit. Clearly, though, these are fairly sketchy and heuristic accounts

26

Introduction

of the phenomenon. More formal economic theories of serial entrepreneurship will be discussed in the next chapter. Also surprising is the fact that relatively little research has so far been conducted on portfolio entrepreneurs. Storey et al. (1987) observed that 80 per cent of the directors of ‘fast-growth’ companies (those hiring more than fifty employees within five years of incorporation) owned other businesses. In quoting these results, Storey later concluded that ‘portfolio owners are therefore of key [economic] importance’ (Storey 1994a, p. 112). This highlights the need for further theoretical and empirical research on portfolio entrepreneurship.

Notes 1. The introduction to a recent special issue on the Economics of Entrepreneurship in the Journal of Business Venturing stated: ‘Economics helps us understand how individuals make decisions, why and how they create and grow organisations, and what the intended and unintended consequences of these actions are at both the micro and macroeconomic levels. Economics further helps us analyze how entrepreneurship influences growth and development and how, in turn, the macro structure of a region or country influences the type and quantity of entrepreneurship. Economic analysis provides insights for scholars and road maps for practitioners and policymakers’ (Minniti and Lévesque, 2008, p. 603). 2. In fact, GEM does collect some data on established firms, though these are rarely used or discussed in public forums. 3. Taking this distinction further, Buenstorf (2007) distinguishes between ‘opportunity’ and ‘necessity’ spinoffs. The former are new ventures formed by employees who commercialise ideas that would otherwise be shelved by the parent firm (Klepper and Thompson, 2006). The latter are new ventures stimulated by adverse developments at the parent firm which render future employment at this firm less attractive. According to Buenstorf (2007), both types of spin-off performed similarly well in the German laser industry. 4. See Reynolds (1997), Delmar and Davidsson (2000), Davidsson and Honig (2003), Reynolds et al. (2004), Wagner and Sternberg (2004), Arenius and Minniti (2005), Rotefoss and Kolvereid (2005) and Wagner (2006b). 5. Johnson, Parker and Wijbenga (2006) introduce a special interdisciplinary issue of Small Business Economics on the topic. 6. The same drawbacks also apply to Gartner and Shane’s (1995) little-used measure of the number of organisations per capita, whose rationale is Hawley’s (1907) argument that entrepreneurship depends on ownership rights. 7. Of course, these types of survey can be criticised for asking hypothetical questions, without forcing individuals to bear the constraints of self-employment as they would if they acted upon their declared preferences. Indeed, much lower rates of serious entrepreneurial intention emerge from longitudinal analyses of wage and salary workers (Katz, 1990), implying that much ‘latent entrepreneurship’ is probably only cheap talk. This has not stopped some researchers studying its determinants (Grilo and Irigoyen, 2006; Masuda, 2006). Because of the danger of cheap talk, I will usually adopt in this book the economist’s standard practice of ignoring studies that report interviewees’ declared preferences, focusing instead on revealed preferences (i.e. actual behaviour). The principal exception relates to a discussion of job satisfaction (see chapter 4). 8. See Bone (2006) for a highly entertaining though hopelessly tendentious account of ‘direct selling’ self-employed salesmen and women, in which it is claimed that managerial control over these self-employees can be even greater than in conventional wage contracts.

Introduction

27

9. ‘They [franchisees] operate without close and direct supervision, but are required to operate within procedures laid down and subject to unilateral change; they appropriate the profits generated by the business, but only after they have made turnover-related payments to their franchisor; and, despite buying or leasing the means of production, are subject to restrictions on how they are used both during and after the currency of the agreement’ (Felstead, 1991, pp. 38–9). 10. Williams (1999) argues that franchisees take less risk than independent self-employed business owners, because of profit-sharing arrangements with franchisors and lower demand uncertainty resulting from selling a known product. In support of his contention, Williams went on to observe a lower variance of self-employment incomes among franchisees than non-franchisees in his 1987 CBO sample of full-time self-employed workers. However, this might overstate the case, not least because exit rates can be higher among franchisees than independent business owners (Bates, 1994a). 11. See Spear (1992). Pérotin (2006) supplies a recent analysis. 12. According to Glaeser (2007), only 1 per cent of American self-employees are engaged in skilled manufacturing. Also, San Jose in California has among the lowest self-employment rates in the USA, despite being home to Silicon Valley, often regarded as the most entrepreneurial cluster in the world. Glaeser (2007) distinguishes high-value from low-value entrepreneurs by disaggregating this group by industry sector and income level. 13. Blau (1987), Steinmetz and Wright (1989), Aronson (1991), Bregger (1996) and Williams (2000). 14. See Lin et al. (2000), Manser and Picot (1999), Kuhn and Schuetze (2001), and Moore and Mueller (2002). 15. See Kuznets (1966, Table 4.2). Numerous studies report self-employment rates which are significantly negatively correlated with GDP per capita over shorter time spans. See Ilmakunnas and Kanniainen (2001), Torrini (2005), Bergmann and Sternberg (2007), Ho and Wong (2007) and Wennekers et al. (2005). Yamada (1996) reports a correlation coefficient of −0.85 between self-employment rates and GDP per capita for over thirty developing countries. 16. See e.g. White (1984) and Brock and Evans (1986, chap. 2) for details. 17. Agriculture was never fully collectivised in Poland, which accounts for its high rate of selfemployment for all workers inclusive of agriculture. 18. Baumol (1990, 1993a); Dutz et al. (2000) and Aidis et al. (2008). 19. Some researchers contend that entrepreneurship in developing countries has been unfairly linked to low-value informal sector work, owing to early work in the development area which linked this sector to a makeshift outcome when mobility between agriculture and industry was impeded (van der Sluis et al., 2005). 20. See Yamada (1996) for evidence from a Peruvian sample and Zhang et al. (2006) for evidence from rural China. In the latter study, the rural self-employed earn more than twice as much than rural employees, albeit with nine times the standard deviation of income. Early studies also generally report higher returns to the self-employed in developing economies (see Parker, 2004, chap. 1, n. 19, for references), although there are exceptions. In Kenya, only 10 per cent of the self-employed appear to earn more than the average wage (Daniels, 1999; Daniels and Mead, 1998).

Part I Selection

2

Theories of entrepreneurship

Numerous thinkers have speculated about both the origin and function of the entrepreneur and the nature of entrepreneurship. A large body of research in economics now addresses these topics. Section 2.1 briefly surveys ‘early’(chiefly pre-1975) views about entrepreneurship. These are mainly concerned with defining and identifying salient aspects of entrepreneurship in a general way. The remaining sections treat ‘modern’ (post-1975) contributions to the economics literature on entrepreneurship. These are typically framed as models of occupational choice between entrepreneurship and paid employment. They tend to be less concerned with definitional issues and usually take entrepreneurship to be any activity where individuals work for themselves and trade off risk and returns, rather than opting for safe returns in a different occupation. For simplicity, the latter will generally be called simply ‘paid employment’ hereafter. Section 2.2 introduces a basic occupational choice model of entrepreneurship, in which all agents are treated as homogeneous. The analysis is broadened out in section 2.3, where individuals are endowed with heterogeneous entrepreneurial abilities, and also in section 2.4, which endows them with heterogeneous risk attitudes. The theories contained in these two sections predict which individuals become entrepreneurs and which do not on the basis of differences in personal characteristics. In contrast, section 2.5 asks why a profit-making entrepreneurial occupational choice exists at all if entrepreneurs’ tasks can also be performed within existing firms. I also consider here the existence of non-profit-making ventures (social enterprises) and family firms. Section 2.6 asks what types of firm ‘spawn’ new enterprises. Spawned ventures can be either ‘dependent’ or ‘independent’ spin-offs. Dependent spinoffs are ventures formed in collaboration with an incumbent firm (sometimes termed ‘intrapreneurship’), whereas independent spin-offs are pursued entirely separately from an incumbent (‘entrepreneurship’). Section 2.7 traces out the implications of entrepreneurship as an occupational choice for economic growth and development. Section 2.8 briefly sketches out several multiple equilibrium models which carry implications for regional variations in entrepreneurship. Section 2.9 concludes. 31

32

Selection

This chapter provides an up-to-date account of theoretical perspectives on the economics of the entrepreneur. The discussion is deliberately kept as accessible as possible; technical details are relegated to the appendices located at the end of the chapter. 2.1

‘Early’ theories of entrepreneurship

My treatment of early theories of entrepreneurship will be fairly brief, since much of this literature has been summarised by other authors. I will group these theories by theme rather than chronologically, unlike Hébert and Link (2006), for example, in the updated version of their ‘classic’ enquiry into the identity of the entrepreneur. 1. Arbitrage and the bearing of risk and uncertainty. Richard Cantillon (1755) stressed the importance of the entrepreneur as an arbitrageur or speculator, who conducts all exchanges and bears risk as a result of buying at certain prices and selling at uncertain ones. Cantillon’s is a risk theory of profit: anyone who receives an uncertain income can essentially be regarded as an entrepreneur. According to Cantillon, successful entrepreneurs perform a key role in the economy by relieving the paralysis engendered by uncertainty, allowing production and exchange to occur and market equilibrium to be attained. Unsuccessful entrepreneurs go out of business: only the ‘fittest’ survive. Entrants appear when profits persist. Cantillon’s entrepreneur is not an innovator, nor does he change supply or demand. Instead, he is perceptive, intelligent and willing to take risks. His role is to bring the two sides of the market together, bearing all the risks involved in this process. Subsequent researchers have developed Cantillon’s thoughts in two separate and well-trodden directions. First, Kirzner (1973, 1985) emphasised the importance of the entrepreneur as a middleman or arbitrageur, who is alert to profitable opportunities that are in principle available to everyone. According to Kirzner, individuals have different access to, or views about, entrepreneurial opportunities, which can explain why some people become entrepreneurs while others do not. Successful entrepreneurs merely notice what others have overlooked and profit from their exceptional alertness. Kirzner neither explained where alertness comes from, nor whether individuals or government can deliberately cultivate it. More recently, Gifford (1998) has tried to endogenise alertness in a model of limited entrepreneurial attention. Entrepreneurs endowed with high levels of managerial ability optimally spend their time operating numerous projects: they therefore face a high opportunity cost of perceiving new innovative opportunities, which renders them less ‘alert’. Following Knight (1921), the second line of research highlights the importance of risk and uncertainty. According to Knight, entrepreneurs have limited information about the availability of natural resources, technological change and prices. Although factor prices are contractible, consumer demand and competitors’ actions can change, so output prices (and hence profits) fluctuate.1 Hence entrepreneurs need to possess particular characteristics such as self-confidence, judgement, a venturesome nature, foresight – and luck. These characteristics are not tradeable, and being complementary

Theories of entrepreneurship

33

to other productive assets, make it logical for entrepreneurs to own rather than lease those other assets and so form a firm. This provides a basic entrepreneurial theory of the firm (Foss et al., 2007). One of Frank Knight’s key contributions was to recognise that the decision to become a worker or an entrepreneur depends on the risk-adjusted relative rewards in each sector. In his own words: The labourer asks what he thinks the entrepreneur will be able to pay, and in any case will not accept less than he can get from some other entrepreneur, or by turning entrepreneur himself. In the same way the entrepreneur offers to any labourer what he thinks he must in order to secure his services. (Knight, 1921, p. 273).

Thus Knight viewed individuals not as born entrepreneurs or non-entrepreneurs, but as opportunists, who can turn their hand to entrepreneurship when the risk-adjusted returns there are relatively favourable – or alternatively to paid employment when they are not. The remainder of the chapter shows how modern economic research has followed directly in this tradition, making explicit the risk-adjusted returns Knight referred to. One aspect of Knight’s writing that has arguably generated more heat than light is the distinction between ‘risk’and ‘uncertainty’. Alvarez and Parker (2009) define risk as the case where individuals do not know in advance the (random) outcome of a draw from a given probability distribution, but do know the parameters and structure of the true underlying probability distribution. In contrast, uncertainty arises when individuals are not only ignorant of the outcomes of the random draw, but also of the true probability distribution. Some authors from the business studies tradition (e.g. Shane, 2003) claim that uncertainty precludes the use of economic optimisation methods. However, this claim is false because individuals can form subjective (albeit possibly incorrect) beliefs about the probability distribution; and optimisation methods can still be applied in this context (Alvarez and Parker, 2009). Alternatively, Knight might have meant that even the possible outcomes themselves are unknown, rather than just the probabilities (Langlois and Cosgel, 1993). Arguably, though, it is hard to imagine any practical situation where outcomes are completely unthinkable, rather than just very unlikely. For example, the concept of the Internet was anticipated even in the 1950s, as was space travel many decades before that; and these are very radical developments compared with the more mundane market uncertainties that entrepreneurs typically deal with. This makes Langlois and Cosgel’s (1993) argument difficult to sustain, quite apart from the fact that other researchers have reached very different verdicts about ‘what Frank Knight really meant’ (see, e.g., LeRoy and Singell, 1987). 2. Co-ordination of factors of production. According to Jean-Baptiste Say (1828), the chief contribution of the entrepreneur is to combine and co-ordinate factors of production. The entrepreneur stands at the centre of the economic system, directing and rewarding the various factors of production, and taking the residual as profit. Personal characteristics such as judgement, perseverance and experience required for successful entrepreneurship are in scarce supply, providing supernormal profits

34

Selection

to these entrepreneurs. Furthermore, all of these characteristics have to be present simultaneously in order for an entrepreneur to be successful. Entrepreneurs need to be resourceful, knowing how to overcome unexpected problems and to exploit (although not develop) existing knowledge. Although some have criticised Say’s view of the entrepreneur as just a superior kind of worker with managerial duties (e.g. Hébert and Link, 2006), others have offered modern re-statements and developments of Say’s perspective. For example: ‘an entrepreneur is someone who specialises in taking judgemental decisions about the co-ordination of scarce resources’ (Casson, 2003, p. 20).2 3. Innovation and creative destruction. According to Josef Schumpeter (1934, 1939), entrepreneurship entails innovation. The entrepreneur does not operate within conventional technological constraints, making small gradual changes to existing production methods. Instead, he or she develops new technologies or products that make discrete discontinuous changes which shift the paradigm altogether, breaking organisational routines and driving economic development (Santarelli and Pesciarelli, 1990). In Schumpeter’s words, the entrepreneur as innovator is responsible for ‘the doing of new things or the doing of things that are already being done in a new way’ (1947, p. 151). This could involve: (i) the creation of a new product; (ii) a new method of production; (iii) the opening of a new market; (iv) the capture of a new source of supply; or (v) a new organisation of industry. Similar to Say, the entrepreneur is an exploiter rather than an inventor of new knowledge. Schumpeter regarded entrepreneurial actions as the principal cause of business cycles and economic development. In his grand vision of ‘creative destruction’, a wave of entrepreneurial innovation hits the economy, displacing old products and production processes, followed by rapid imitation by new competitors. Ultimately stability is restored and entrepreneurship reaches a temporary cessation before the next wave occurs. Both entrepreneurial activity and the ensuing profits are temporary, unless the entrepreneur continues to innovate: ‘everyone is an entrepreneur only when he actually carries out new combinations and loses that character as soon as he has built up his business, when he settles down to running it as other people run their business’ (Schumpeter, 1934, p. 78). This argument can be seen as a rationale for defining entrepreneurship in terms of new venture creation, as discussed in the previous chapter. Schumpeter seems to have viewed the entrepreneur not as a calculating utility maximiser but as a rare, unusual creature driven by instinctive motives. Schumpeter regarded profit as a residual, not a return to the entrepreneur as a ‘factor of production’, and even claimed that ‘the entrepreneur is never a risk bearer’ – ‘capitalists’ (financiers) bearing the risk instead (1934, p. 137). However, the view that only capitalists and not entrepreneurs bear risks has been roundly and justly criticised by subsequent writers (e.g. Kanbur, 1980), for imposing an arbitrary distinction between ‘capitalists’ and ‘entrepreneurs’, and for ignoring entrepreneurs’ actual and opportunity costs in operating ventures that can (and often do) fail. Another limitation of Schumpeter’s thinking centres on his prediction of the eventual demise of the entrepreneur, caused by the rise of large monopolistic firms, which have

Theories of entrepreneurship

35

advantages at co-ordinating teams of workers to perform R&D. This prediction has not come to pass, partly because new and small firms continue to be important innovators (see chapter 10), and partly because the majority of US basic research is funded not by large firms but by the federal government (Schultz, 1980). Of all the early writers about entrepreneurship, Schumpeter is perhaps the most widely cited today. In several respects, this is puzzling. His insistence that entrepreneurs do not bear risk is frankly bizarre. And his gloomy prediction about the demise of entrepreneurship has been clearly disproved. Perhaps his ideas were overly influenced by the spread of socialism at the time he was writing. More generally, Schumpeter ignored the forces of competition which challenge and overturn incumbents in modern capitalist economies (Acs and Armington, 2006, chap. 1). Furthermore, arguably not one but several Schumpeters survive in his writings. His ideas underwent continual development, and it is hard to know whether it is this or the sometimes biblical opaqueness of his verbal argumentation which generates the confusion that has spawned a large if ultimately rather sterile academic industry interpreting ‘what Schumpeter really meant’. 4. Leadership and motivation. In stark contrast to Schumpeter, others have claimed that a defining feature of entrepreneurs is that they bring about changes of a gradual nature to existing products and processes, through a combination of leadership, motivation, the ability to resolve crises and risk-taking (Leibenstein, 1968). Some supportive if anecdotal evidence about the incremental nature of innovations, discussed by Bhide (2000), is discussed in chapter 10. 5. Personal or psychological traits. This line of thought relates entrepreneurship to the possession of special innate personal characteristics. It has rather fallen out of favour among entrepreneurship researchers, but it will be briefly discussed in chapter 4. While not exhaustive, the above list includes many of the most influential ‘traditional’ views about entrepreneurs. The brevity of this overview was deliberate. Much of this material has been discussed extensively before and, as others have noted, ‘a tome could be written on the connections and contradictions between these theories’(Barreto, 1989, p. 43; see also van Praag, 1999). One of the more interesting connections appears in a recent article by Langlois (2007), who identifies the ‘opportunity discovery, exploitation and evaluation’ definitions of entrepreneurship with Kirzner, Knight and Schumpeter, respectively. I close this discussion by drawing two conclusions. First, the early theories can be dichotomised between those in the neoclassical tradition (such as Knight, 1921; Marshall, 1930; Schultz, 1980), based on the notion that entrepreneurs lead markets into equilibrium, and those in the Austrian tradition (such as Kirzner, 1973; 1997) which sees entrepreneurs as part of an ongoing disequilibrium process of indefinite or infinite duration. On a related theme, Shane and Venkataraman (2000) argue that Kirzner’s and Schumpeter’s theories represent different types of opportunity. In Kirzner, opportunities embodied in existing information are discovered, by the most alert individuals (entrepreneurs); whereas in Schumpeter, opportunities are created as a result of new information and technology. Kirzner’s entrepreneurs move the market towards

36

Selection

an economic equilibrium consistent with existing information, whereas Schumpeter’s create the prospect of entirely new equilibria. Rosen (1997) attempted to find some common ground between the two sides of this dichotomy. Second, it seems self-evident that none of the above theories offers a complete account of entrepreneurship. To be sure, one can cite selectively from these theories to support a particular viewpoint, but none of them provides necessary or sufficient conditions for entrepreneurship. For example, construction site managers co-ordinate factors of production, while corporate R&D employees contribute to the development of innovations; yet neither is obviously an ‘entrepreneur’. This exemplifies the subtleties which make entrepreneurship such an elusive, and almost certainly multi-dimensional, concept. 2.2 The occupational choice model of entrepreneurship I: homogeneous agents

Modern economic theories of entrepreneurship differ in at least two important respects from those described above. Perhaps the most important distinction relates to the dominance of the utility maximising paradigm in modern economic research. Modern economic theories take as their starting point the Knightian premise that individuals do not have to be entrepreneurs. They can choose between entrepreneurship and some outside option (usually taken to be paid employment); and they choose the occupation that offers them the greatest expected utility. Most theories treat occupational choice as a discrete, rather than a continuous, decision. This follows Kanbur (1981), who noted the difficulty of viewing occupational choice as an adjustment at the margin of a continuous process, such as ‘engaging a “little bit” more in entrepreneurial activity’ (p. 163). However, as will be discussed below, some researchers have also analysed how individuals mix their time between different occupations, which resembles more of a continuous than a discrete choice. A second distinctive feature of modern economic theories of entrepreneurship is that they make their simplifying assumptions explicit. These assumptions usually include the existence of competitive product markets, known technology, and price-taking workers and entrepreneurs. In many cases these assumptions are inessential to the results, and merely simplify the analysis. The present section analyses entrepreneurship as an occupational choice when agents are homogeneous. It first analyses the simplest cases, and then introduces additional layers of complexity (and realism) by introducing risk and risk aversion. The simplest ‘static’ models, in which events take place in a single period, are discussed first. This is followed by a treatment of ‘dynamic’ models, in which events unfold over several periods. Before turning to these models, I will first define some of the terms that will be used extensively in what follows. While they can be found in many standard economics texts, it is convenient to group them together and establish a common notation. I will give simple explanations of these concepts in the text; more formal definitions are collected in the first part of the chapter appendix.

Theories of entrepreneurship

(a): Risk-averse

37

U(y) U(y)

Utility,

U

(b): Risk-loving

(c): Risk-neutral

U(y)

Income,

y

Figure 2.1 Utility functions

2.2.1

Definitions of risk aversion and risk

To commence, consider a utility function U (y) which is an increasing function of income, y. In principle, utility functions can take several shapes: three are illustrated in Figure 2.1. Utility function (a) is a concave function of income. While extra units of income increase utility, they increase it by progressively smaller amounts. This utility function is also said to embody risk aversion. Individuals with utility functions exhibiting a greater degree of curvature than in (a) are said to be more risk-averse. Technically, this utility function has a positive first and a negative second derivative with respect to y. More risk-averse people are willing to pay a higher insurance premium to avoid risk than their less risk-averse counterparts. Utility function (b) in Figure 2.1 is a convex function of income. Extra units of income increase utility by progressively greater amounts. This utility function is said to embody risk-loving preferences. Individuals with utility functions exhibiting a greater degree of curvature are said to be more risk-loving. Technically, this utility function has positive first and second derivatives with respect to y. Utility function (c) is linear in income. It embodies risk neutrality. As the name suggests, risk-neutral people are indifferent to risk. A wide array of evidence suggests that in the real world, most individuals, including entrepreneurs, are risk-averse.3 This debunks the popular view that entrepreneurs are gamblers, and is consistent with evidence that entrepreneurs’ behaviour is better described by moderate and calculated risk taking than outright gambling. Thus entrepreneurs ‘enjoy the excitement of a challenge, but they don’t gamble. Entrepreneurs avoid low-risk situations because there is a lack of challenge and avoid high-risk situations because they want to succeed. They like achievable challenges’ (Meredith et al., 1982, p. 25). Furthermore, individuals appear to exhibit decreasing absolute risk aversion (DARA: see the chapter appendix). This means that as they become wealthier, they

38

Selection

become less reluctant to gamble a fixed sum. DARA plays a key role in some of the models of entrepreneurship analysed throughout this book. It is also helpful to have a precise definition of ‘an increase in risk’. Two useful and general definitions are second-order stochastic dominance (SOSD) and meanpreserving spread (MPS). These concepts are defined formally in part 1 of the chapter appendix. Essentially, both concepts refer to a more dispersed distribution of risky returns. These are more general notions of an increase in risk than an increase in the variance of a random variable, as explained in the chapter appendix.

2.2.2 Simple static models The simplest static models treat an economy without risk, where individuals choose between working for a wage of w and producing output independently as an entrepreneur in return for profit, π. If π > w, workers switch into entrepreneurship. By the laws of supply and demand, the extra output decreases the price it is sold for, reducing π until it comes into equality with w. Conversely, π cannot be less than w because then entrepreneurs would quit, reducing aggregate output and thereby increasing the price until equality was restored. It follows immediately that w = π is an equilibrium condition in this simple case. Likewise, any exogenous increase in w (caused by technological change, for example) will decrease the equilibrium number of entrepreneurs (de Wit, 1993). Richer versions of this simple model of homogeneous agents introduce risk. Risk can emanate from various sources. Entrepreneurs may be unsure about the demand for their good, their ability to produce, or future costs of production (Wu and Knott, 2006). Although employees can also face some risk, through income variation and redundancy for example, entrepreneurs face more variable incomes than workers do as well as higher business closure rates (see chapters 13 and 14). Most researchers assume for simplicity that entrepreneurs face some kind of idiosyncratic risk to their profits, whereas employees all face a certain wage, w. The assumption of perfect certainty in paid employment is usually innocuous, and can be relaxed without affecting analytical predictions. It will also be assumed that entrepreneurs cannot completely diversify or sell their risk. This also appears to be a reasonable assumption. Markets for private unemployment, accident and sickness insurance are limited and prone to moral hazard problems. Few entrepreneurs have access to stock markets to share risk, and realworld capital markets are imperfect, undermining entrepreneurs’ efforts to smooth consumption in the face of income risk. It might be thought that, given risk aversion among entrepreneurs, an increase in risk in entrepreneurship would necessarily decrease the equilibrium number of entrepreneurs. In fact, the opposite turns out to be the case, if entrepreneurs can choose output once risky demand outcomes are revealed. The reason is that risk can provide upside potential as well as downside outcomes, making riskier markets more attractive and hence liable to market entry even by risk-averse entrepreneurs (Sheshinski and Drèze, 1976). The possibility that occupations offering high upside potential can attract able individuals seeking rare but highly profitable opportunities was recognised

Theories of entrepreneurship

39

over half a century ago (Roy, 1951). In practice, abler individuals are indeed more likely to try start-ups in risky markets like software development than in ‘safe’ markets like hairdressing (Bhide, 2000). Bhide contends that entrepreneurs who become very successful in sectors with highly skewed returns do not necessarily need to have special insights or a novel technological discovery. They might merely possess superior information and sell an already existing service or product more effectively than the competition. He concludes that ‘given…limited endowments, profitable start-ups tend to cluster in small, uncertain market niches’ (Bhide, 2000, p.113).4 I have so far considered the consequences of greater risk in the economy, perhaps caused by an increase in the volatility of trading conditions. But what if there is a general increase in risk aversion among individuals? That might reflect a change in tastes within an economy. Alternatively, it can be thought of as a device to analyse cross-country differences in risk attitudes. Kanbur (1979) studied the effects of greater risk aversion on the equilibrium number of entrepreneurs. Kanbur’s model generates two hypotheses: (i) if labour is hired after the outcome of the random shock is observed, an increase in absolute risk aversion decreases the equilibrium number of entrepreneurs, since risk-averse individuals avoid risky occupations; (ii) if labour is hired before the outcome of the random shock is observed, an increase in absolute risk aversion has an ambiguous effect on the number of entrepreneurs. Prediction (i) appears to give some theoretical backing for the popular view that Europe has less high-value-adding entrepreneurship than the USA because Europeans are more risk-averse than Americans are. However, there are at least two reasons to treat this argument with scepticism. First, most entrepreneurs who hire workers in practice do so continuously, i.e. before risk is resolved. That makes Kanbur’s hypothesis (ii) the relevant case – but this case is the theoretically ambiguous one. Second, the Kanbur model assumes that every individual is identical, and that all entrepreneurs hire workers. These assumptions appear unrealistic and are relaxed in some of the models discussed below.

2.2.3

Dynamic models

The models discussed so far assume costless switching between occupations. Thus if entrepreneurship becomes attractive relative to paid employment, workers are assumed to move immediately into entrepreneurship; the converse also applies. However, individuals might incur costs of switching occupation. These costs could be non-pecuniary involving, for example, the sudden loss of a pleasant compensating differential, disruption to an accustomed lifestyle, a feeling of rootlessness, stress from change, or stigma from failure (Gromb and Scharfstein, 2002; Landier, 2004). Or they could be economic in nature involving, for example, lost sector-specific experience, costs of raising start-up capital (if entering entrepreneurship), or retraining costs (if entering paid employment). Switching costs might also relate to exit barriers caused by incurring sunk costs of capital with limited resale value, prior commitments to customers, or a desire by entrepreneurs to avoid sending an adverse signal of ability by abandoning their ventures (Boot, 1992).

40

Selection

Dixit (1989) showed that risk together with sunk costs can give agents an option value of waiting before switching. This reduces the total amount of entry and exit that occurs – as conditions have to become very bad before entrepreneurs close their business and relinquish their sunk costs, or very favourable before they are willing to incur the risk of jeopardising their assets by entering the market. Risk generates an ‘option value’ of remaining in the present occupation and deferring a costly switch. Only when average incomes in entrepreneurship reach some upper ‘trigger point’ will people become entrepreneurs. And they will only leave entrepreneurship in the presence of the adjustment cost if incomes drop to some lower trigger point. Between these two trigger points individuals remain in their current occupation (Dixit and Rob, 1994). Consequently, there may be hysteresis (i.e. path-dependence) in occupational choice. Individuals may remain in entrepreneurship even if the returns there at a given instant are less than those available in an alternative occupation. It is rational to remain in the occupation not only because of the switching cost, but also because there is an option value to wait and see if conditions in the currently unfavourable occupation improve. Only if this option value becomes sufficiently small does switching become worthwhile. Dixit and Rob (1994) went on to show that a socially suboptimal amount of switching takes place in equilibrium. There might therefore be a role in principle for government to improve on the free-market equilibrium by speeding up the reallocation of labour between occupations, for example by subsidising training costs (see chapter 17). Although Dixit and Rob (1994) did not discuss this implication of their work, the hysteresis result is useful because it goes some way to explaining why there is relatively little voluntary switching between entrepreneurship and paid employment from year to year, despite the abundance of latent entrepreneurs (q.v. chapter 1). US estimates of the proportion switching into self-employment in any given year are only 2–3.5 per cent (see Parker, 2004, note 14, p. 67, for citations). The figures for switching out of selfemployment into paid employment are higher, amounting to 10–20 per cent in the USA and the UK (Fairlie, 1999; Henley, 2005). This reflects the high closure rates of small businesses, especially newly established ones (see chapter 14). A dynamic optimisation model proposed by Parker (1996, 1997a) explores the implications of introducing risk into paid employment as well as entrepreneurship. These models also allow individuals to mix work hours between occupations, enabling occupational choice to be analysed as a continuous, rather than just a discrete, choice. Incomes in the two occupations are assumed to evolve as independent Brownian motions.5 The key result from this model is that if expected income growth is higher in entrepreneurship than in paid employment, and if there is no income risk in paid employment, then greater income risk in entrepreneurship unambiguously reduces the optimal fraction of time spent in entrepreneurship. But if there is income risk in paid employment as well, greater income risk in entrepreneurship has ambiguous effects on the optimal fraction of time spent in entrepreneurship. At first sight, it might appear surprising that risk-averse individuals could respond to an increase in risk in entrepreneurship by spending even more time in it. But the logic is analogous to that applying in the twoasset portfolio problem in finance, where the risk of the overall portfolio is a convex

Theories of entrepreneurship

41

combination of the risk of each asset. Overall portfolio risk can sometimes be reduced by increasing the portfolio share of the riskiest asset (Parker, 1997a). 2.3 The occupational choice model II: heterogeneous ability – the Lucas (1978) model

In practice, it is likely that entrepreneurs differ from employees and among themselves in terms of their innate ‘entrepreneurial ability’. Lucas (1978) was one of the first researchers to trace out the economic implications of heterogeneous entrepreneurial ability. This section first explains the gist of the Lucas model before discussing some limitations and extensions. 2.3.1 The Lucas model Ability in entrepreneurship might derive from human capital (van Praag, 2005), idiosyncratic leadership qualities (Leibenstein, 1968) or judgement (Casson, 2003), among other possible sources. Whatever its provenance, to fix ideas and keep the exposition simple I will just assume that everybody, whether an entrepreneur or not, has some innate ability which describes how well they would perform in entrepreneurship were they to become an entrepreneur. Ability is measured as a single-dimensional quantity, x. The lowest ability in the population is x, while the highest is x. Denote the relative frequency of individuals with an entrepreneurial ability of x by f (x), and the cumulative relative frequency by F(x).6 It is also assumed that abilities are fixed and known with certainty by each individual. Jovanovic (1982) among others has relaxed this assumption, in a model where entrepreneurs learn about their x by observing their performance in entrepreneurship. Discussion of the Jovanovic model is deferred until chapter 11. Lucas (1978) assumed that x scales up an entrepreneur’s output of q to give net profit of π(x) = xq − c, where c is the cost of using capital and labour to produce q (similar results obtain if x scales down the entrepreneur’s costs, but it is simplest to work with the output assumption). The output price is normalised to unity and all people are taken to be risk-neutral. It follows directly that all and only individuals with x ≥ x˜ will become entrepreneurs, where x˜ is the identity of the ‘marginal entrepreneur’, defined as the person who is indifferent between the two occupations:

π(˜x) = w.

(2.1)

An implication of this ‘fundamental’ equation of entrepreneurial occupational choice is that there are a total of 1 − F(˜x) entrepreneurs; the remaining F(˜x) people work for the entrepreneurs as employees. This equilibrium is illustrated in Figure 2.2. A property of the Lucas production function xq is that entrepreneurs’ demands for labour and capital are greater among those with higher x. That is, abler entrepreneurs run larger firms, irrespective of whether size is defined in terms of employment or capital assets. This furnishes another reason why the ablest people become entrepreneurs.

42

Selection

Retur ns

Profits, p(x)

W age , w

Ability

,x x~

x Employees

x Entrepreneurs

Figure 2.2 Occupational choice with heterogeneous entrepreneurial ability

Operating a firm enables able people to spread their ability over a larger scale and so reap greater returns (see also Murphy et al., 1991). Lucas (1978) obtained further results by making a dynamic extension to his model. He asked how entrepreneurs adjust their demand for factors of production (capital and labour) when the stock of capital increases as the economy develops. A key parameter in this regard is the technical elasticity of substitution, σ , which describes the sensitivity of entrepreneurs’ chosen capital–labour ratios to changes in the relative prices of capital and labour. Assuming that firm growth rates are independent of firm size (known as Gibrat’s Law, explained in chapter 11), Lucas showed that if σ is less than (greater than) (equal to) unity, increases in per capita capital in the economy decrease (increase) (leave unchanged) the equilibrium number of entrepreneurs, and increase (decrease) (leave unchanged) average firm size. The intuition behind this result is as follows. If the supply of capital increases, (a) labour becomes more productive and (b) the price of capital decreases relative to that of labour (i.e. the wage). Effect (a) increases the demand for labour while effect (b) decreases it. But if σ < 1, effect (b) is modest relative to effect (a), since relative factor usage does not respond much following the change in their relative prices. So the demand for labour rises overall, increasing the wage, and so encouraging marginal entrepreneurs to become employees. That increases x˜ and thereby also the average size of firms.7 It is interesting to interpret Lucas’ dynamic result in terms of a prediction about future trends in the fraction of the workforce who are entrepreneurs. Empirical estimates consistently point to an elasticity of substitution of less than unity (see, e.g., Hamermesh, 1993, pp. 92–104). Given also that capital per head tends to grow over time (Maddison, 1991), Lucas’model therefore predicts that the fraction of entrepreneurs will inexorably

Theories of entrepreneurship

43

decline over time, while the average firm size and industrial concentration will both inexorably increase. Lucas (1978) concluded his article with an anecdote about a small, family-owned (‘entrepreneurial’) restaurant he visited on vacation. He conjectured that this organisational form would eventually come to be replaced by a large franchise outlet (‘paid employment’) as an inevitable consequence of rising real wages. More generally, one can observe the progressive disappearance of small, owner-managed convenience stores in Britain and America and the reallocation of their labour towards large supermarket chains. Consistent with these arguments, much of the observed productivity gains in the US retail sector in the 1990s involved the entry of new establishments associated with large multi-unit firms, and the exit of small independents (Haltiwanger, 2006). Crosssection comparisons between rich and poor nations bear this story out. For example, whereas fewer than 15 per cent of Ghanaian manufacturing workers are employed in establishments with more than ten workers, less than one per cent of total US valueadded in the manufacturing sector comes from establishments employing fewer than five employees (Gollin, 2008). Lucas (1978) also provided some evidence of his own that average firm size is positively related to the capital stock per head. He regressed the average number of employees per firm (as a proxy for average firm size) against per capita gross national product (as a proxy for the stock of capital per head). He estimated that a 1 per cent increase in GNP per capita is significantly associated with a 1 per cent increase in the average number of employees per firm. Gollin (2008) calibrates a ‘model economy’ based on a simple extension of Lucas’ model using Japanese data and also predicts a negative (‘L-shaped’) relationship between rates of entrepreneurship and GDP per capita. However, Lucas’ prediction that entrepreneurship is destined to inevitable decline as economies develop receives less support if the aggregate self-employment rate is used as a proxy for the size of the entrepreneurship sector. As shown in chapter 1, while the self-employment rate decreased steadily between the nineteenth and mid-twentieth centuries in most developed economies, the trend in some of them reversed in the last quarter of the twentieth century, at a time when real wages and the capital stock were both increasing. This leads one to ask whether something is amiss in Lucas’ model, or whether other factors are at work, overwhelming the mechanism proposed there. I consider these issues next. 2.3.2

Criticisms of the Lucas model Despite its elegance and profound influence on the economics of entrepreneurship, Lucas’ (1978) model can be criticised on several grounds. One technical objection is the assumption of Gibrat’s Law, which underlies the dynamic variant of Lucas’ model. Recent studies have cast doubt on the applicability of this ‘Law’, finding that firm growth rates are actually not invariant to firm size (see chapter 11). A more fundamental criticism of Lucas’ model is that it neglects technological change, which is arguably a more important source of macroeconomic growth than

44

Selection

changes in the capital stock. Innovation and technological change may have endowed new entrepreneurial ventures with a competitive advantage over their larger counterparts in some industries, such as ICT (Acs and Audretsch, 1991). To the extent that this is true, more rather than less entrepreneurship would be expected to emerge as economies develop. A related problem is that Lucas’ model is highly aggregated and simplified, glossing over important industry composition effects. This is likely to be important in view of the concentration of entrepreneurs in particular industry sectors (see chapter 4). In particular, rising levels of prosperity often translate into greater demand for services (e.g. personal and customised services) that entrepreneurs are particularly efficient at supplying. Furthermore, Lucas did not define entrepreneurial ability, x, precisely, taking it to be exogenous with an unexplained provenance. As noted above, it is common for researchers to interpret ability in terms of human capital, such as schooling and experience. Unfortunately, the empirical evidence does not point to an unambiguous relationship between human capital and selection into entrepreneurship (see chapter 4). Occasionally, ability is defined in a more specific manner, for example as the ability to predict and adjust to idiosyncratic changes in consumer tastes, as in Takii (2008). However, this definition is more difficult to operationalise empirically. Another problem is that if entrepreneurs learn over time, it may be inappropriate to treat entrepreneurial managerial ability as fixed and exogenous (Otani, 1996). A simple test of whether entrepreneurial ability is fixed or improves with learning can be performed by estimating how entrepreneurs’ productivities (a proxy for ability) vary with the number of firms in a locality. If abilities are fixed, then more firms in a locality implies lower average entrepreneurial ability.8 But if entrepreneurs can learn from each other, thereby enhancing their ability, then more firms in a locality might imply higher average entrepreneurial ability. Measuring ability as average firm productivity in a sample of Italian manufacturing firms, Guiso and Schivardi (2005) detected a significant positive relationship between average firm productivity and the number of firms in the entrepreneur’s locality, supporting the learning hypothesis and rejecting Lucas’ static ability hypothesis. These findings may imply that instead of trying to reduce entry barriers, governments should promote entrepreneurial clusters since these can efficiently facilitate entrepreneurial learning and productivity. 2.3.3

Variants and extensions of the Lucas model Numerous variants of Lucas’ model have been proposed.9 In each of these models, the ablest people select into entrepreneurship. This section reviews three major extensions to Lucas’ model. The Lucas model assumes that ability is one-dimensional; that entrepreneurs compete in only one industry sector; and that ability only affects returns in entrepreneurship. In the first set of extensions, ability is allowed to be two-dimensional. The second extension introduces additional industry sectors for entrepreneurs to work in. The third recognises that ability can also affect wages in the outside option of paid employment. I now discuss each of these extensions in turn.

Theories of entrepreneurship

45

One of the first models to make ability two-dimensional was Calvo and Wellisz (1980). Individuals are now distinguished by age as well as ability. Calvo and Wellisz showed that faster technological progress leads to an equilibrium outcome where older, inherently less able entrepreneurs are replaced by younger and inherently more able entrepreneurs. Gifford (1998) proposed a different two-dimensional model of ability, comprising ‘managerial ability’ (valuable for operating existing projects) and ‘entrepreneurial ability’ (valuable for innovating new projects). An implication of her model is that firms might do best passing up entrepreneurial opportunities because finite time and limited cognitive abilities enable them to pay only limited attention to new opportunities. That might help explain why new firms can exploit ideas which time-constrained incumbents choose not to develop themselves. The most influential model with multiple abilities is Lazear’s (2005). Suppose individuals differ among themselves by having different endowments of two skills, x1 and x2 . It is plausible that specialists in paid employment earn the maximum of x1 and x2 , i.e. max{x1 , x2 }. That is because they can choose to specialise in jobs where their highest skill is rewarded the most, while relying on other workers in the organisation to supply the other skill type. In contrast, solo entrepreneurs have to supply both skills themselves; and the performance of an entrepreneurial venture is only as good as its weakest link. Hence entrepreneurs earn λ min{x1 , x2 }, where λ > 1 is the market value of entrepreneurial ability. For any joint distribution F(x1 , x2 ), it follows that the more similar are x1 and x2 , the likelier any given individual is to do best by choosing entrepreneurship. Hence Lazear (2005) predicts that entrepreneurs are ‘jacks of all trades’, unlike employees who are specialists. Some evidence on this issue is reviewed in chapter 4. The second major extension of the Lucas model introduces multiple industry sectors. There are two principal models here. The first, by Murphy et al. (1991), departs from the Lucas model by showing that free occupational choice involving entrepreneurship can be inefficient. Murphy et al. (1991) consider two industrial sectors indexed by j ∈ {1, 2}, which use different technologies whose productivities at time t are denoted by Ajt . Productivity is assumed to evolve according to best practice (ability) in that sector: Ajt = Aj,t−1 max{xj,t−1 }, where {xj,t−1 } is the set of entrepreneurial abilities in industry j at time t − 1. That is, productivity in industry j at t is assumed to grow most in industries where the most talented entrepreneurs congregate. It is easy to think of several reasons why that might be so. For example, the most able individuals could pioneer radical new innovations specific to the industry which increase the productivity of every producer in that industry. In both industries, an entrepreneur’s output depends on the product of their own ability and the number of workers they hire. Suppose that the productivity of hired workers differs between the two industries, with less rapidly diminishing marginal returns in industry 1. This means that, all else equal, an extra worker in industry 1 will increase output by more than an extra worker in industry 2. Murphy et al. (1991) assume that entrepreneurs can freely choose which sector to enter to maximise their private returns. They then show that the ablest individuals, x ∈ [˜x1 , x], all become entrepreneurs

46

Selection

in industry 1; the least able individuals, x ∈ [x, x˜ 2 ] all become workers; and individuals of intermediate ability, x ∈ [˜x2 , x˜ 1 ], all become entrepreneurs in industry 2. In this model there are two marginal entrepreneurs, x˜ 1 and x˜ 2 < x˜ 1 . The ablest entrepreneurs all bunch in the sector with the lowest diminishing marginal returns because this enables them to run larger firms and spread their ability over a larger scale. Unfortunately, while these entrepreneurial occupational choices are privately optimal, they are socially inefficient, because bunching does not maximise aggregate productivity growth. Given the definition of Ajt above, the social optimum would have the ablest entrepreneur joining one industry while the next ablest joins the other. The problem is that each entrepreneur does not take into account the technological externality caused by their own occupational choice decision. Nocke (2006) analysed a second model of multiple sectors, where competition is more intense in larger sectors. Only the ablest entrepreneurs can effectively compete in large sectors, but relatively few entrepreneurs possess sufficient ability to make entry into those markets attractive. This can arguably help explain the puzzling real-world observation that many large markets contain relatively few entrepreneurs and therefore less product variety compared with smaller markets. Before turning to the third set of extensions to Lucas’ model, note that entrepreneurs can not only choose their industry sector, but also their mode of entry, for example whether: to start a new firm; to succeed their parents in a family firm (if this option is available); or to acquire an external firm. According to Dennis (1997), 78.5 per cent of new industry entrants in the USA are de novo (i.e. brand-new entrants), with the remainder acquiring a business through inheritance, takeover or franchising. Parker and van Praag (2006a) estimated that 83 per cent of Dutch entrepreneurs in 1994 started their venture de novo, while 10 per cent took over a family firm and 7 per cent took over an outside firm. Using a forward-looking rational choice model, Parker and van Praag (2006a) argue that individuals with a family-business background will optimally acquire less human capital. Having lower lifetime wealth from human capital these people are as a result more risk-averse (by DARA: see the first part of the chapter appendix) – and so are more likely to take over firms than to start risky new ones from scratch. Holmes and Schmitz (1990) explore a different angle of the ‘mode of entry’ issue: whether an entrepreneur should continue operating a venture, or transfer it to a possibly less able entrepreneur in order to release time and resources to explore new opportunities. Holmes and Schmitz (1990) predict that the least able types will take over and manage existing firms, while the most able will specialise in setting up new businesses. Those with intermediate ability optimally either manage the businesses they started, or replace them with higher-quality businesses purchased from abler entrepreneurs. The Holmes–Schmitz model incorporates and extends ideas from Schumpeter and Kirzner about opportunity recognition, and helps explain why some individuals are ‘portfolio’ entrepreneurs, running several businesses simultaneously, while others become ‘serial’ entrepreneurs, running a succession of businesses one after another (see chapter 1). This model can also explain why some entrepreneurs simply rely on buying businesses created by others.

Theories of entrepreneurship

47

The third set of extensions to Lucas’ model allows the outside wage w also to be a function of ability x: i.e. w = w(x).10 There are two possibilities (Jovanovic, 1994). One arises if more able individuals earn more in both occupations. Then π(x) and w(x) are both increasing functions of x. The marginal entrepreneur x˜ is defined by π(˜x) = w(˜x). Several outcomes can arise, including multiple marginal entrepreneurs (Figure 2.3(c)). Another outcome, illustrated in Figure 2.3(a), is that the least able individuals become entrepreneurs. Another possibility is that π(x) is an increasing function of ability x, while w(x) is a decreasing function of x. This could arise, for example, if x measures productive rebelliousness, which pays off in entrepreneurship but is penalised in team-based paid employment. In this case, Figure 2.3(b) is applicable and, as in Lucas’ model, only the highest-ability types become entrepreneurs. However, this case seems less realistic than the one where abilities increase in both occupations, but do so more steeply in entrepreneurship. As Frank Knight put it: It may well be true that able leaders are in general also more competent workers, or operatives, but the gain in superior direction is so much more important than that from superior concrete performance that undoubtedly the largest single source of the increased efficiency through organisation results from having work planned and directed by the exceptionally capable individuals, while the mass of the people follow instructions. (Knight, 1921, p. 17)

If w(x) does not depend on x, the outside return is simply a constant w, the case considered hitherto. This case might arise even if workers have heterogeneous abilities, provided that there is asymmetric information in the labour market, whereby employers cannot discover the abilities of their workers and workers cannot credibly signal their abilities to employers. Then employers have to offer all workers a single pooled wage, irrespective of their hidden ability. Interestingly, this can cause occupational choice to be inefficient. Asymmetric information gives able individuals an additional incentive to turn entrepreneur, since some of them who would choose paid employment under full information might now choose entrepreneurship under asymmetric information to avoid being penalised by a fixed wage in paid employment which does not reflect their above-average ability. The result is too many entrepreneurs in equilibrium, since some individuals choose entrepreneurship purely as a result of imperfect information about their ability (Laussel and Le Breton, 1995).11 Laussel and Le Breton (1995) speculate that their findings might have special implications for transition or developing economies, which lack institutions for screening workers efficiently, and which might therefore be burdened with too many (rather than too few) small-scale entrepreneurs. One problem with Laussel and Le Breton’s model is that workers and employers have incentives to engage in monitoring to reveal hidden abilities. Parker (2003c) allows for this possibility in a model in which individuals again possess private information about their abilities, and can choose freely between producing a good with either a safe or a risky technology. Under plausible conditions, it is shown that employers optimally deploy the safe technology and screen workers to reveal their abilities at the outset, whereas banks deploy the risky technology and monitor entrepreneurs only if and when

48

Selection

j

(a)

Expected retur n in occupation

j = PE

j = EN

~ x

0 Ability

x

Expected retur n in occupation

j

(b)

j = EN

j = PE

~ x

0

Ability

x

Expected retur n in occupation

j

(c)

j = EN

j = PE

0

~ x2

~ x1 Ability

x

Figure 2.3 Occupational choice with two occupations, entrepreneurship (EN) and paid employment (PE). (a) PE attracts the ablest entrepreneurs: x > x˜ enter PE (b) EN attracts the ablest entrepreneurs: x > x˜ enter EN (c) Multiple marginal entrepreneurs, x˜ 1 and x˜ 2

Theories of entrepreneurship

49

they default. The reason is that banks have to monitor entrepreneurs ex post anyway to guard against opportunistic defaults (see chapter 7); while efficient production in large firms requires accurate separation of worker types from the outset.12 A variety of possible occupational choice outcomes emerges from the Parker (2003c) model, including all of those illustrated in Figure 2.3. This model also has some implications for the efficiency of bank-financing of entrepreneurial investments – an issue followed up in chapter 7. 2.4 The occupational choice model III: heterogeneous risk attitudes – the Kihlstrom and Laffont (1073) model

What if individuals can choose freely between entrepreneurship and paid employment, as in the models just discussed, but face risk in entrepreneurship and have heterogeneous aversion to risk rather than heterogeneous entrepreneurial ability? The economic implications of this question have been analysed by Kihlstrom and Laffont (1979) (henceforth KL79). In Kihlstrom and Laffont’s own words, their model is ‘a formalisation, for a special case, of Knight’s discussion of the entrepreneur’ (1979, p. 745). This is because Knight proposed diversity among individuals with regard to confidence in their judgement to run firms: those who are ‘confident and venturesome “assume the risk” or “insure” the doubtful and timid by guaranteeing to the latter a specified income in return for an assignment of the actual results’ (Knight, 1921, p. 269). KL79 abstract from unequal managerial abilities: risk attitude is the only source of heterogeneity in their model. As noted earlier in the chapter, more risk-averse individuals are willing to pay a premium in order to insure themselves against risk. Put another way, faced with the choice between receiving a safe return in paid employment, w, and a risky return in entrepreneurship, a more risk-averse person is likelier to take the safe option. KL79 proved that if there is a continuum of agents differentiated only by their risk attitude, only the least risk-averse will become entrepreneurs. This can be understood within the usual occupational choice framework in terms of a marginal entrepreneur who is indifferent between risky entrepreneurship and safe paid employment. Everyone who is less risk-averse than the marginal entrepreneur becomes an entrepreneur, and everyone who is more risk-averse becomes an employee (see Figure 2.2, where x is now interpreted as an inverse measure of risk aversion). To be a viable occupation in this model, entrepreneurship must pay a risk premium. This is only attractive to the least risk-averse; it is insufficient to compensate more risk-averse individuals. The prediction that less risk-averse individuals are more likely to become entrepreneurs is intuitive and has received independent empirical support (see chapter 4). KL79 went on to derive several further results from their model. First, more riskaverse entrepreneurs are predicted to operate smaller firms, i.e. use less labour than less risk-averse entrepreneurs, under reasonably general conditions. Second, a general increase in individual risk aversion reduces the equilibrium wage. This is implied by the previous two results, because greater risk aversion decreases the demand for labour by

50

Selection

each entrepreneur and increases the equilibrium number of employees. Both changes reduce the aggregate demand for labour and hence w (cf. Kanbur, 1979). If all individuals were risk-neutral in the KL79 model, industry equilibrium would be Pareto efficient. That is, there would be no allocations that could make one individual better off without making another individual worse off. However, when some individuals are risk-averse, three types of inefficiency can arise. First, maximisation of aggregate output requires all firms to produce the same output when the production function is concave. However, entrepreneurs with heterogeneous risk aversion operate firms of different sizes, as noted above. Second, individuals could be made better off if risks were shared, but there is no mechanism for facilitating this. Third, in general the wrong number of individuals become entrepreneurs. On the one hand risk aversion causes too few individuals to become entrepreneurs (from the standpoint of efficiency), but on the other hand risk aversion causes too small a demand for labour, reducing w and so causing too many individuals to choose entrepreneurship. In general, the two effects will offset each other and the net effect on efficiency cannot be predicted without further information about tastes and technology. For example, in the special case where all individuals are equally risk-averse, it can be shown that there will be too many entrepreneurs in equilibrium. Another special case is constant returns-to-scale technology, under which only one firm is optimal, compared to the greater number that would emerge in KL’s competitive equilibrium.13 One way of enhancing efficiency would be to introduce a risk-sharing mechanism such as a stock market (Kihlstrom and Laffont, 1983; Grossman, 1984). For example, in the context of international trade with foreigners who have a comparative advantage in entrepreneurship-rich goods, Grossman (1984) argues that establishing risk-sharing mechanisms to stimulate domestic entrepreneurship is a better solution than imposing welfare-reducing tariffs or other trade restrictions on foreign entrepreneurs. However, a problem with the specific solution of a stock market to share risks is that it is likely to be impractical for small enterprises. The high fixed costs incurred by a stock market listing are likely to deter small firms from diversifying their risks in this way. One might also ask whether investors can write financial contracts to insure riskaverse entrepreneurs. In fact, standard principal-agent models in microeconomics predict that variable returns are generally required to elicit the high levels of discretionary non-contractible entrepreneurial effort needed to make ventures succeed. If individuals faced a fixed return irrespective of what they do, they would have incentives to economise on privately costly discretionary effort. This is a theme which will recur several times in this book. The implication is that it can be optimal for lenders to write contracts that avoid fully insuring entrepreneurs (Keuschnigg and Nielsen, 2004b; Rampini, 2004). With regard to partial insurance on the other hand, the results are more subtle and surprisingly far-reaching, as an important paper by Newman (2007) demonstrates. Newman envisages an insurance company offering entrepreneurs a predetermined and relatively smooth income stream in return for entrepreneurs’ risky stream of profits. Entrepreneurs still have to bear some risk, in order to induce them to supply effort;

Theories of entrepreneurship

51

but they receive partial insurance through the less risky income stream provided by the insurer. Newman (2007) contends that the conceptually modest extension of partial income insurance creates a serious conceptual problem for the KL79 model, since he shows it implies that the poorest, and hence the most risk-averse individuals (under DARA) become the entrepreneurs, while the wealthy work for them as their employees! The rationale underlying this counter-intuitive result is that it is easier for an insurance company to incentivise poor individuals to exert effort via an income-smoothing scheme than the wealthy. Poor individuals are more risk-averse, so it is cheaper for the insurance company to devise an income stream which provides the correct incentives for them to exert effort in entrepreneurship. In contrast, wealthy people need to bear a large amount of risk under partial insurance in order to induce them to supply effort – and that can make entrepreneurship expensive and unattractive for these agents relative to ‘safe’ paid employment. Newman (2007) claims that the preposterous prediction that wealthy people work for poor entrepreneurs calls into question any theory of entrepreneurship based on heterogeneous risk attitudes in which entrepreneurs’ primary role is to insure workers. As he put it, ‘The fragility of this theory’s [KL79’s] empirical predictions suggests that we probably should look elsewhere for explanations of the roles and causes of entrepreneurship’ (Newman, 2007, p. 12). However, there are two limitations to Newman’s contention. First, if risk attitudes are independent of wealth, then KL79’s analysis continues to apply regardless of whether entrepreneurs can be insured. And even if risk attitudes are not independent of wealth, a private market that insures entrepreneurs’ incomes is not guaranteed to exist. In practice, insurance for business owners tends to be against specific risks (e.g. loss or damage of business equipment, or travel insurance), and not of the form Newman studied, which has insurers claiming entrepreneurs’ payoffs in return for payouts of a partially smoothed stream of income. One reason why Newman’s insurance system is not observed is that there are obvious incentives for entrepreneurs to under-report their incomes to the insurers. If enough entrepreneurs did this, an income-transfer system of this kind would quickly become infeasible. With a missing market for income insurance, one immediately returns to the world of KL79.

2.5 The very existence of an entrepreneurial option

This section changes the emphasis from characteristics of individuals to characteristics of firms. I first ask why a profit-making entrepreneurial occupational choice exists at all, if entrepreneurs’ tasks can also be performed within incumbent firms. Some answers to this question are offered in the first part of this section and in section 2.6. In the second part of this section I then ask why enterprises which do not maximise profits, or which are less efficient than incumbents, can survive and even thrive in competitive markets. Such enterprises include small firms which lack economies of scale; family firms with objectives other than pure profit maximisation; and not-for-profit (social) enterprises.

52

2.5.1

Selection

Theory of the firm considerations Some industries and occupations are characterised by high levels of independent smallscale business ownership and self-employment. Examples include taxicab and truck driving, sales, draughtsmen services and freelance journalism, among others. Hitherto, we have asked why individuals might choose to become entrepreneurs. But one can also ask why entrepreneurship should ever be an option in the first place. Putting it another way, one can ask when it is more efficient for the owners of large firms (called ‘principals’) to unbundle their firm and buy in some parts of their production from independent contractors, rather than organising production activity within the firm. In the latter case, principals pay workers (called ‘agents’) a contracted wage. If only the most efficient producers can survive under competition, firm organisation can come to represent a binding constraint on the occupations available to individuals to choose from. An extreme articulation of this position might even claim that firm characteristics, rather than individual characteristics, lie at the heart of why some individuals are observed to be entrepreneurs. Essentially, this question goes back to the classic work of Coase (1937). Coase argued that firms exist because they economise on transaction costs of exchange compared with individual producers contracting in spot markets. But internal exchange can also be costly, leading to cases where individual contracting is preferred. Holmstrom and Milgrom (1994) refer to the choice between hiring employees or outside contractors as the ‘make or buy’ decision. For example, personal ownership of their own vehicle gives independent self-employed truck drivers greater incentives to care for and maintain it than if they are hired as employees driving the company’s vehicle. An employee’s lack of incentives to care for a company asset represents an agency problem. Given this (and the relative ease of measuring delivery performance), one might therefore expect logistics companies to hire independent self-employed truckers rather than to maintain an expensive payroll of employee truck drivers (Milgrom and Roberts, 1992, pp. 249, 311). In fact, contrary to this argument, independent contractors accounted for less than one-third of US trucking activity conducted by interstate trucking firms with $1 million or more in revenues in 1991 (Nickerson and Silverman, 2003). And in a study of the US taxicab industry, Sherer et al. (1998) found that taxicab companies make greater use of employee drivers than of independent contractors. To explore the forces at work here, consider a simple principal-agent problem in which employers (principals) can either hire individuals (agents) to produce output via an employment contract, or buy in the production from independent contractors. Contractor status is equated with independent entrepreneurship (self-employment or business ownership) in what follows. A simple way of conceiving this is to imagine employers ‘contracting out’employee work to the self-employed by changing the nature of their contract, as happens in some software companies, for example.14 The text below provides a verbal account of this simple ‘hire or outsource’ model; the second part of the chapter appendix presents it formally.15 An important consideration for employers wondering whether to produce output inhouse using an employment relationship or to buy in production from independent

Theories of entrepreneurship

53

entrepreneurs is the degree to which in-house production generates other ‘asset specificity’ benefits which would be lost if it were to be contracted out. For example, in-house production might generate knowledge or output spillovers which increase production elsewhere in the firm. If these asset-specificity benefits are valuable enough, employment (‘hire’) might dominate entrepreneurship (‘outsource’) and prevent the latter from ever emerging as a feasible occupational choice. But if these benefits are not very valuable, or absent altogether, employers might be able to do better by buying in the services of outside independent contractors. Independents might be cheaper than direct employees if the latter but not the former are covered by legally binding minimum wages, costly employment protection legislation and trades union collective bargaining (Parker, 2007b). Under these circumstances, an employer might be able to share the benefits of outsourcing by offering workers attractive contracts to become independent contractors. It is notable that by making the agent rather than the employer the residual claimant, the agent will always supply high levels of non-contractible effort. Thus the avoidance of agency costs provides a further reason why entrepreneurship might dominate paid employment and emerge as a feasible organisational form. It also highlights a potential efficiency benefit of entrepreneurship relative to paid employment, since it is not always possible for employers to incentivise employees to supply high levels of discretionary effort. The formal ‘hire or outsource’ model outlined in part 2 of the chapter appendix provides further details. Several examples of asset-specific benefits which encourage incumbents to keep work in-house arise in particular industries. For example, in the trucking industry asset-specificity benefits can be embodied in a logistic company’s reputation value of a good service record, or the value derived from co-ordinating interdependent freight deliveries (Nickerson and Silverman, 2003). These benefits might be lost if logistics companies outsourced trucking services to independent contractors, since the latter care little about the company’s reputation and have few incentives to take account of the company’s overlapping schedules with other deliveries. In the direct selling industry, the asset-specific benefit from keeping work in-house could be complementary with non-selling activities (Anderson and Schmittlein, 1984). In the specific context of the market for taxicab drivers, Sherer et al. (1998) highlighted the importance of training to safeguard product quality and the value of co-ordinating workers to ensure they engage in mutually co-operative behaviour. These considerations can also promote the use of employees rather than contractors. For example, contractors who compete against each other in an ‘eat what you kill’ environment might overuse scarce resources in an effort to gain a competitive advantage over rival contractors serving the same principal. And small firms and the self-employed are known to perform less training than large firms (see chapter 10), which can reduce the quality of outsourced work and hence the attractiveness of contracting out (Briscoe et al., 2000; Meager and Bates, 2004). More generally, Sherer et al. (1998) highlight an advantage of hiring employees over contractors based on the notion that employees can be more easily redeployed into other jobs in the firm when unforeseen contingencies arise. Employment contracts can therefore be thought of as incomplete contracts which are superior to spot

54

Selection

(entrepreneurial) contracts when some circumstances cannot be perfectly anticipated. Employment contracts can also avoid continual renegotiations, which entail search and transaction costs and are prone to opportunism, including hold-up problems (Grossman and Helpman, 2002). An example of opportunism arises in the trucking industry. Most independent contractors optimally buy trucks which are readily deployable, so there is a thin market for specialised trucks. To avoid hold-up problems, it pays logistics companies to buy specialised non-standard vehicles themselves. The evidence is consistent with this outcome (Nickerson and Silverman, 2003). Evidence also suggests that spot contracting is more common when local markets are thicker, not only in trucking (Hubbard, 2001), but also in business services (Ono, 2007), and across industries in general (Holmes, 1999). If asset-specific benefits from keeping work in-house are negative, entrepreneurship can be expected to dominate paid employment quite generally. This is likely to occur in vertically integrated firms which suffer from diseconomies of scale and scope, since it can pay these firms to focus on their core competencies in order to safeguard product quality. They can do so by accessing externally provided specialised services that are unavailable within the firm (Gramm and Schnell, 2001; Grossman and Helpman, 2002). In principle, specialised outside contractors can centralise expertise, reap economies of scale and exploit smoother production schedules, especially for intermediate goods or services which are only needed irregularly. The formal ‘hire or outsource’ model set out in part 2 of the chapter appendix is consistent with several aspects of Holmstrom and Milgrom’s (1994) theoretical analysis, as well as some independent empirical evidence. One prediction is that independent entrepreneurship is more likely to emerge when output is hard to measure. A second prediction is that entrepreneurship is more likely to emerge when it is costly to monitor employees. Garen’s (2006) analysis of US Dictionary of Occupational Titles data supports this argument. Garen (2006) showed that independent contractors tend to occupy jobs that require greater worker expertise and judgement (such as analysing data and undertaking various and complex duties) – for which worker monitoring is difficult and costly. Being more difficult to monitor, jobs with these characteristics can make independent entrepreneurship an optimal form of firm organisation, since being residual claimants, entrepreneurs have direct incentives to perform well.16 The first two columns of Table 2.1 summarise the firm-level determinants of the decision to ‘make or buy’ discussed in the foregoing. The evidence to date is scant but supports some of the above arguments. Harrison and Kelley (1993) surveyed plants in the US metalworking sector in 1986–7, and analysed the determinants of contracting out of a particular production process: metal machining or metal-cutting. Nearly 60 per cent of the firms which responded to the survey claimed that they regularly contracted out some machining work. Using logistic regression, Harrison and Kelley reported that the size of the parent company, the wages of the machinists and the diversity of the plant’s product mix were all significantly associated with a greater propensity to contract out machining. These findings are consistent with explanations 6 and 9 in column 1 of Table 2.1. A second study, by Abraham and Taylor

Theories of entrepreneurship

55

Table 2.1. Demand- and supply-side factors influencing the ‘make or buy’ decision Firm choice (Demand side)

Individual choice (Supply side)

Make (hire) employee

Buy from entrepreneur

Become an entrepreneur

Become an employee

1. Assets specific to firm; externalities

1. Assets not specific to firm; no externalities 2. Output hard to measure 3. Effort, care of assets and motivation unimportant 4. High monitoring costs 5. Thick markets

1. Autonomy (flexible schedules) desired 2. Income smoothing not desired 3. No borrowing constraints

1. Autonomy not desired

2. Output easy to measure 3. Effort, care of assets and motivation important 4. Low monitoring costs 5. Thin markets: hold-up problems 6. No diseconomies of scale or scope 7. Training needed to enhance quality and uphold routines 8. Gains from co-ordinating or redeploying workers 9. Low employment costs and restrictions

2. Income smoothing desired 3. Borrowing constraints

6. Diseconomies of scale or scope 7. Training and worker quality inessential

8. No gains from co-ordinating or redeploying workers 9. High employment costs and restrictions

(1996), using data from the US Bureau of Labor Statistics Industry Wage Surveys, found that smaller firms in urban locations and firms paying above-average wages were more likely to contract out services. Their interpretation was that small firms have greater need of, and urban firms have easier access to, specialised services provided by outside contractors. To complete the model’s treatment of the ‘make or buy’decision, the last two columns of Table 2.1 suggest motives from the worker’s perspective for choosing between independent contractor status (entrepreneurship) and paid employment. In the formal ‘hire or outsource’ model, autonomy (enabling flexible work schedules: row 1) can explain why individuals might be willing to forgo income smoothing provided by employers (row 2). On the other hand, highly risk-averse individuals would prefer income smoothing in paid employment over the income variability on offer in independent entrepreneurship.17 And borrowing constraints might render workers incapable of

56

Selection

raising the necessary funds to go independent in any case (row 3; this is T ∗ in the formal model; see also chapter 9 for an analysis of borrowing constraints). Finally, I briefly discuss how ‘theory of the firm’considerations might affect the structure and feasibility of ‘team’start-ups involving multiple founders. A natural question to pose is whether members of a team of founders are likely to continue to share ownership rights in the future, or whether one member will end up with exclusive ownership and control rights (the ‘entrepreneur’), employing the others (‘employees’) for a salary. In a model which explores this question, Barzel (1987) considered two founders in separate jobs, one of which (A, say) makes output that is intrinsically easy to monitor, while the other (B) makes a productive contribution that is intrinsically difficult to monitor. For example, A could be the producer of widgets, whereas B is the co-ordinator of contracts to supply clients, the keeper of accounts and records, and the searcher for new clients and cheaper raw materials. Barzel predicts that B will become the residual claimant (‘entrepreneur’), employing A using a standard wage contract. The reason is simply that, by the nature of her job, B finds it easier to shirk. But B has fewer incentives to shirk if she is given residual claimant status rather than a wage contract, since as a residual claimant she bears directly the costs of her own foregone effort (a similar point applies in the ‘hire or outsource’ model described above). Hence this arrangement maximises the joint product from collaboration and hence the payoffs of the two agents. An important point is that the identity of the entrepreneur is not fixed, but depends on the relative wages commanded by the two collaborators. In particular, if A’s productivity (and hence wage) increases relative to B’s, then A could eventually replace B as the entrepreneur. The reason is that the loss from shirking by A under a fixed wage contract may eventually become so great that it becomes imperative to eliminate it – which is achieved by shifting the residual claimant status from B to A. Barzel’s insight is consistent with a prediction from incomplete contract theory (Grossman and Hart, 1986) that individuals who objectively have the most to gain from ownership become the residual claimants and monitor other members of the production team (see also Alchian and Demsetz, 1972). Alvarez and Parker (2009) demonstrate that this prediction holds under conditions of risk, but not necessarily under ‘Knightian’ uncertainty, in the sense of the terms as defined earlier in this chapter. In practice, many founders of new ventures operate under conditions of uncertainty and often do not agree about the distribution of states of nature. In which case, Alvarez and Parker (2009) show that both the founders’ subjective beliefs and their respective productivities determine the identity of the residual claimant throughout the evolution of the firm. In contrast to incomplete contract theory, the residual claimant need not be the ablest entrepreneur with the most to gain objectively, but could instead be the most optimistic member of the founding team. Naturally, ventures founded on the basis of over-optimism are less likely to take objectively sound decisions and survive than those founded on the basis of a hard-headed evaluation of relative ability. Perhaps counter-intuitively in this context, Parker and Alvarez (2009) point out that extreme uncertainty can be a more benign venture-founding environment than moderate uncertainty. The reason is that highly uncertain environments force entrepreneurs to ‘keep an open mind’ and update their

Theories of entrepreneurship

57

beliefs using market data more rapidly than they would in less uncertain environments. In the long run, uncertainty eventually dissolves into risk as entrepreneurs learn the true distribution of states of nature. Then the predictions of incomplete contract theory once again prevail – provided the venture survives the consequences of early decisions made on the basis of possibly incorrect subjective beliefs under uncertainty.

2.5.2

Non-profit-maximising ventures Enterprise is frequently carried out in non-profit-maximising organisations as well as in conventional profit-maximising private sector ventures. Two obvious questions then arise: Why would any entrepreneur not want to maximise profits? And how can nonprofit-maximising, or inefficiently small ventures, survive in competition with efficient profit-maximising rivals who can be expected to undercut them in competitive markets? These questions will be answered in two specific contexts below: that of small (and often family-owned) firms and that of social enterprises.

Small and family-owned firms A fundamental question is why, when large firms enjoy scale economies in production, small firms which lack scale economies can survive in competition with them. Not only can large companies price below smaller enterprises, but their greater diversification enables them to take risks which smaller firms cannot, including investments which take a long time to generate returns. Yet we saw earlier in this chapter that both the Lucas (1978) and the Kihlstrom and Laffont (1979) models give rise to a size distribution of firms in which there is no intrinsic penalty to smallness. These models can therefore explain the coexistence of small entrepreneurial and large corporate firms without having to appeal to special factors that give small firms an advantage over large ones. Nevertheless, some special factors which do give small ventures a positive advantage over their larger counterparts are of intrinsic interest, to be sure, and are considered in various places throughout this book. Among these, it has been suggested that entrepreneurial firms enjoy superior innovation performance, being more efficient at supplying small batches of goods or customised goods and services than their larger rivals, who tend to have a comparative advantage at producing standardised products in large production runs. Small firms can also avoid diseconomies of scale (Williamson, 1985; Reid and Jacobsen, 1988). But a different and sometimes overlooked aspect of non-profit-maximising entrepreneurship is the existence of non-pecuniary benefits from business ownership. These benefits can explain why non-profit-maximisers can compete with or even undercut profit-maximisers, since the former may be willing to accept below-market incomes in return for non-pecuniary benefits. As one researcher has put it, ‘A casual look at neighbourhood pharmacies, groceries, small farming and independent truckers suggests that competitive pressures have not driven out small businessmen with preferences for self-employment, despite sometimes meagre returns’ (Feinberg, 1980, p. 1166). This could be another reason why Lucas’ (1978) argument

58

Selection

that capital accumulation progressively drives these types of firm to the wall might be incomplete.18 Even so, one can still ask why it is that in almost every industry large firms, with all their advantages of scale, do not supplant small enterprises on efficiency grounds. Wiggins (1995) proposed an answer to this question based on legal costs and incentive contracts. In circumstances when employees know that their employer has an incentive to renege on promised compensation for their effort, it becomes costly for employers to commit to strong incentives. When these costs outweigh the benefits of scale, the small independent firm can replace the large firm as the efficient form of organisation. Thus when costs of employers reneging are low, workers who anticipate employer reneging to be likely and who cannot afford the legal costs of tort can only guarantee being compensated for their work by becoming entrepreneurs themselves and owning the rights to the fruits of their own effort. This insight might possibly explain why some sectors of the economy are dominated by small firms rather than by large ones, including high-tech research-intensive enterprises with long development lead times which afford ample opportunities for employer reneging. Another question is how family firms can survive against non-family firms. Family firms might compromise efficiency by utilising narrow (within-family) pools of managerial expertise, or by pursuing objectives of family harmony rather than growth opportunities (Westhead and Howarth, 2006). But family firms might also enjoy efficiency advantages over conventional profit-maximising firms, for example by commanding greater loyalty and trust from their workforce. Perhaps reflecting these offsetting advantages and drawbacks, the balance of evidence about the relative performance of family compared with non-family firms is mixed (Bertrand and Schoar, 2006; Westhead and Howarth, 2006). Social enterprises According to a recent definition: The term ‘social enterprise’ describes organisations…that aim to achieve financial sustainability through trading for a social purpose. Any surplus they make is usually re-invested to further the approved mission of the organization, and not distributed for private gain. They are located mainly in the non-profit sector, although several, such as The Big Issue and the Grameen Bank, have been incorporated as for-profit social enterprises. (Haugh, 2006, p. 401)

Social enterprises are part of the not-for-profit (NP) sector, which plays an important role in many modern economies. According to Steuerle and Hodgkinson (1999, p. 77), NPs accounted for roughly 7 per cent of US GDP in the 1990s, and their number grew at a faster rate (5 per cent per annum) than those of for-profit (FP) enterprises (1.4 per cent). NPs enjoy relatively high survival rates, with about 60 per cent of the 130,000 new NPs formed between 1992 and 1996 in the USA still being in existence in 1996 (Cordes et al., 2004). They also employ about 4.4 per cent of the economically active population across the world (Haugh, 2006, p. 403). In many countries NPs are heavily concentrated in particular services, especially labour-intensive ones like childcare, medical care, education and care for the elderly (Rose-Ackerman, 1996; Haugh, 2006).

Theories of entrepreneurship

59

An important question is how NP enterprises can coexist with FP firms. Altruism has been proposed as one answer.19 Investors do not bear the costs of corporate social responsibility or other NP activities in the stocks that they hold, because in an efficient market these costs are already impounded in the prices of the shares when they purchase them. A lower share price simply increases the cost of capital to entrepreneurs, so NP entrepreneurs bear the cost, which they might be willing to do if they are altruistic (Baron, 2007). And altruism does not seem to prevent NP entrepreneurs from exploiting profit-making opportunities when they arise (James, 1998). It is less clear why non-altruistic entrepreneurs running FPs do not enter and drive NPs out of the market. One reason might be favourable tax treatment of NPs, which takes the form of tax exemptions on eligible income and tax deductibility on donations. In some countries such as the USA, there are also additional exemptions from sales and property taxes, and even postage subsidies. But tax benefits are not the essential characteristic of NP social enterprises (Glaeser and Shleifer, 2001). Many NPs were founded before tax benefits were introduced; and these benefits are limited or non-existent for noncharitable social enterprises (Weisbrod, 1988, pp. 4–5). All in all, some two-thirds of US NP do not qualify for tax deductions on contributions or subsidised postal rates. Their greatest benefit is probably the (non-deductible) effort input of volunteer workers (see below). An alternative explanation for NPs coexisting with FP enterprises is based on the notion of selfish behaviour combined with a Non-Distribution Constraint, NDC (Easley and O’Hara, 1983; Hansmann, 1987). An NDC is a legal constraint that prevents any surpluses generated by NPs being distributed to their owners in the form of stock, dividends or above-market salaries and perks to their employees. Instead, any surpluses must be reinvested in the venture. Easley and O’Hara (1983) derive the NDC as the outcome of a game between society and owner-managers, where the latter produce some benefit valued by society. When the benefit is not directly observable by members of society, it may be optimal to restrict owner-managers’ compensation to provide incentives for the desired output to be produced. In this case, the NP-NDC organisational form emerges. But when the social benefit is transparent, managerial compensation is left unrestricted in the optimal contract, which corresponds to an FP organisational form. More generally, an NDC may offer society benefits by protecting investments made by donors, volunteers, consumers and employees from ex post appropriation by rentseeking entrepreneurs. An NDC signals a credible commitment to outside stakeholders that an entrepreneur running a social enterprise will not exploit their donations by, for example, expropriating them and cutting back on other investments. Conventional FP firms cannot make this commitment, which puts them at a competitive disadvantage, at least for consumers for whom product quality matters and is hard to verify (Glaeser and Shleifer, 2001). Instead, FPs have to appeal to better-informed consumers. This can explain why many industries are characterised by mixed NP and FP enterprise provision. As an example, consider nursing care homes. FP care homes tend to use more sedatives – a cheap way of keeping patients calm – than NP care homes, which provide more

60

Selection

intensive patient care instead (Weisbrod, 1988, chap. 8). Sedatives are a cost-reducing strategy that adversely affects non-contractible quality; consumers anticipating such incentives in the FP sector demand NP care homes instead. The NDC in an NP eliminates the profit incentive to compromise on quality, which FPs lack. This may even enable NPs to command a higher market price than FPs (Glaeser and Shleifer, 2001).20 Other evidence supports this argument too. Weisbrod (1998, chap. 8) reports that NP healthcare providers offer higher levels of hard-to-detect quality than their FP counterparts do, including higher levels of consumer satisfaction and greater provision of information to customers. Rose-Ackerman (1996, sec. 5) cites evidence that quality in child daycare tends to be higher in NPs than in FPs, though this outcome often entails higher production costs. The NDC can also confer on NPs a competitive advantage by helping them to attract donations (Glaeser and Shleifer, 2001) and greater worker effort (Francois, 2003). That is because the NDC ensures that donations of money and labour effort will not simply be converted by the NP into profit, something that cannot be guaranteed by FPs. Volunteer labour turns out to be very important to many NPs, accounting for over 40 per cent of their workforces on average, compared with virtually none for FPs (Haugh, 2006, p. 403). FPs can be further out-competed if workers care about an NP’s social mission, since then NPs can attract the effort of paid employees with a smaller wage premium than FPs do. However, this particular argument is not supported by evidence from Mocan and Tekin’s (2003) analysis of data from 398 day care centres in four US states, which showed that workers in NPs receive substantial premiums in their wages and compensation compared with their counterparts in the FP sector. That might be because workers share in the proceeds, whose disbursement is restricted by the NDC. The discussion so far suggests that the NDC may be a major reason why social enterprises can compete with for-profit enterprises. There are other possible reasons as well. For instance, NPs can cater for heterogeneous demands. Governments are often under pressure to provide uniform levels of services on a majoritarian basis, leaving gaps that can only be filled by specialised NPs (Weisbrod, 1988). Some social enterprises furnish opportunities for ideologically motivated entrepreneurs to influence their environment, and connect with consumers with similar motivations. On the other hand, the NP organisational form is not a unique signal of quality. Reputation, competition and certification are others (Francois, 2003). And it should be borne in mind that social entrepreneurs might achieve their goals more effectively by running an FP and devoting the profits to charitable purposes. It is possible that some public goods would not be forthcoming at all without NDCs or NPs – which provides a possible social welfare justification for the (institutionally imposed) NDC. More controversial from a policy standpoint is the tax deductibility of NPs. Weisbrod (1988) observes a growing trend for NPs to engage in profit-making activities. This brings them into direct competition with FPs, possibly unfairly if they are tax-advantaged; and it may also divert NPs from their central purpose of enhancing social welfare. For example, the YMCA has been criticised for using its charitable status to help undercut private sector health and fitness clubs. The YMCA responds

Theories of entrepreneurship

61

by claiming it uses membership revenues in affluent areas to cross-subsidise provision in poor areas, which would not otherwise be forthcoming. Weisbrod (1988) recommends replacing targeted tax deductibility with neutral tax credits to address this problem. One can also argue that NDCs might put NPs at a competitive disadvantage by blunting the profit motive and hence incentives to be efficient. That can lead to shirking in NPs, unless NP managers are highly motivated by non-pecuniary factors. Lacking market discipline, and buoyed by favourable tax treatment and subsidies, NPs can survive longer than is efficient, and lack external scrutiny that can remove underperforming managers (Rose-Ackerman, 1996). Of course, the separation of ownership and control can also blunt efficiency in large FPs. Unfortunately, there are serious difficulties in comparing efficiency in FP and NP enterprises, owing to differing quality levels, different strategic objectives and disparate consumer needs (Weisbrod, 1988, chap. 2; Rose-Ackerman, 1996, sec. 5). It does seem, however, that NPs as well as FPs provide fewer pure charity services in more competitive markets (Rose-Ackerman, 1996, p. 720). Finally, we still know relatively little about the identity of the social entrepreneurs who operate NPs. Weisbrod (1988, pp. 32–3) cites some anecdotal evidence that NP managers attach greater importance to being cheerful, forgiving and helpful than FP managers. The latter state that financial prosperity and ambition are more important to them, and are likelier to claim that they are driven by a ‘need for power’. Other research shows that lawyers working for NP ‘public interest’ law firms earn less than their FP counterparts, but are likelier to declare themselves content with their career choice. Haugh’s (2006) survey of character trait studies finds that social entrepreneurs are unusually motivated to make a contribution to society; but then one would probably expect this. Based on a GEM-type survey, Harding and Cowling (2004) find social entrepreneurs in the UK to be older and better qualified, and to earn higher incomes than FP entrepreneurs. Harding and Cowling (2004) also report high levels of social entrepreneurial activity among disadvantaged groups, defined in terms of those on low incomes, the unemployed, women and ethnic minorities. These predictions tally in several respects with Parker’s (2008c) neoclassical life-cycle theory of social entrepreneurs, which predicts two dominant career types engaged in this form of entrepreneurship: idealistic individuals who operate social enterprises when young and who make little money over their lifetimes; and wealthy individuals who become social entrepreneurs later in life after they have achieved success either as an employee or as an entrepreneur in the for-profit sector. 2.6

Incumbents’ characteristics: entrepreneurial spawning

Workers often leave incumbent firms to start new ones. They can do so by forming independent start-ups, or new ventures that are related in some way to their employer (‘dependent start-ups’). Gompers et al. (2005) have called the process by which large firms create new independent ventures ‘entrepreneurial spawning’.

62

Selection

Employees of small firms are more likely to change their job than employees of large firms (Parker, 2009bb). In part, this reflects the higher wages and fringe benefits available in large firms. But what is less obvious is why, among workers who change jobs, those leaving small firms are much more likely to become entrepreneurs than those leaving large firms. The evidence on this point is consistent across a wide range of countries, including the UK, the USA and European countries.21 According to Parker (2009b), at least four different theories can explain this phenomenon. First, a ‘transmission theory’ predicts that small firms are particularly effective at transmitting useful business experience and pro-entrepreneurship attitudes to their workers, possibly via inspirational role models. Second, a ‘blocked mobility’ theory associates small firms with ‘bad’ jobs and proposes entrepreneurship as an escape route for frustrated employees trapped in the small-firm sector. Third, a ‘self-selection’ theory suggests that less risk-averse individuals sort themselves into both small firms and entrepreneurship at different stages of their lives (Parker, 2008d).Afourth theory is that bureaucracy imparts conservative attitudes, specialised skills, inward-facing experience and stable rewards, which are all antithetical to entrepreneurship. Bureaucracy is especially pervasive in large firms (Sørensen, 2007). Longitudinal matched employee–employer Danish data on labour market transitions over 1990–7 show that current or previous work experience in a large firm prompts workers to move into self-employment later rather than sooner (Sørensen, 2007).22 In contrast, an empirical analysis based on thirteen waves of BHPS (British Household Panel Survey) data suggests that the self-selection theory is most consistent with the evidence (Parker, 2009b). The analysis so far has focused on independent start-ups, whereby entrepreneurs develop ideas or technologies independently of their parent firm. These types of startup are more common than dependent starts, in which employees operate a new venture as a quasi-independent arm of the parent firm. Individuals engaged in dependent startups are sometimes referred to as ‘intrapreneurs’. Note that intrapreneurship is distinct from management buy-outs (MBOs), whereby employees take over the ownership and control of a firm. Parker (forthcoming) analyses the decision to practice intrapreneurship versus entrepreneurship. It has long been known that large incumbent firms play a central role in the creation of new ventures, especially in the high-tech sector. Rothwell (1984) observed this phenomenon in the semiconductor and computer-aided design industries, where several large incumbents were responsible for the initial technological breakthroughs and furnished both the entrepreneurs and the risk-capital required to found the new technology-based small firms which diffused these innovations into general use. Importantly, the process by which incumbents generate innovative ideas which become commercialised in new independent ventures is not observed only in high-tech firms (Bhide, 2000). In many industries, it is not uncommon for the ‘parent’ firm to maintain a close relationship with its offshoots, even several years after the date of start-up, either as a partial equity holder or as a customer for its products or technologies. A useful way of thinking about dependent starts is in terms of an employee, e.g. an R&D scientist, making a discovery. If the parent company (‘employer’) is unwilling

Theories of entrepreneurship

63

or unable to exploit the discovery, the employee may have to quit their job and found a new venture in order to exploit it. On other occasions, though, it can be worthwhile for an employee to exploit an opportunity in conjunction with their employer in a joint venture. The remainder of the chapter offers various reasons why entrepreneurs sometimes leave firms to pursue new opportunities instead of developing them inhouse with their employer. Following Klepper’s (2001) taxonomy, these theories relate to organisational limitations of incumbent firms; agency costs; and learning behaviour.

2.6.1 Organisational limitations of incumbent firms Why would incumbents be willing to pass up on valuable opportunities, lose talented employees and watch them create new firms, possibly competing in related markets to their own? Often, to be sure, some incumbents go out of their way to stop employees leaving to start new ventures – including, perhaps most famously, Intel (Jackson, 1998, pp. 211–338). But it is also clear that some incumbent firms play a benevolent, or at least not actively obstructive, role in the independent venture creation process. This can happen, for example, when a new business idea falls outside the firm’s core line of business and the firm does best by sticking to its core competencies. In dynamic industries, incumbents may simply face more opportunities than they can exploit. Pursuing every innovation opportunity could easily require more capital than can be raised, and could exceed the firm’s capacity to grow. If so, incumbents can do best by rejecting some projects in order to be ready to exploit more complementary opportunities later on (Cassiman and Ueda, 2006). Sometimes, incumbents are powerless to stop employees leaving to start new ventures, or cannot afford to retain them. In general, though, the potential mobility of scientific personnel need not be a problem for incumbents if the incumbent’s managers can design a labour contract ‘which only induces the scientist to leave and join a rival if the sum of the benefits to the two agents [incumbent and scientist] increases as a result of the scientist’s leaving’ (Pakes and Nitzan, 1983, p. 360). Another possibility, explored in the remainder of this subsection, is that incumbents lose skilled personnel because of organisational limitations. Large firms can be bureaucratic and inert, developing inflexible but valuable internal organisational ‘routines’ which discourage radical change.23 Employees who discover radical new opportunities within these firms are not permitted to develop them in-house because that would disrupt the firm’s routines (Freeman and Engel, 2007). Employees can then become frustrated and come into conflict with their employer (Garvin, 1983; Klepper, 2006). Ultimately, employees might have to leave and start up on their own if they are ever to commercialise their ideas, especially if other existing firms are also unwilling to take up the ideas. This logic can explain why, even though large bureaucratic firms generate plentiful ideas, they can also be associated with numerous independent breakaway starts (Klepper and Sleeper, 2005). Gompers et al. (2005) exemplified the large bureaucratic firm by Xerox, which refused to develop numerous valuable ideas discovered by its employees, including the

64

Selection

Ethernet, the graphical user interface, laser printing and personal distributed computing. Employees who left Xerox to exploit these opportunities founded numerous new companies, including Adobe Systems and 3Com. Somewhat ironically, Chester Carlsson originally started Xerox after his proposal to produce a (new) copy machine was rejected by Kodak, on the grounds that the new copy machine would not earn very much money, and in any case, Kodak was in a different line of business – photography. Entrepreneurial spawning sometimes spans several generations: ‘the ‘Traitorous Eight’ left Shockley Labs to create Fairchild Semiconductor, which itself saw its employees starting, among others, National Semiconductor, Intel, AMD and LSI Logic, which in turn became parents to Cypress, Zilog, Sierra Semiconductor, and many other semiconductor companies’(Hellmann, 2007b, p. 919). As another example, Christiansen (1993) observed that many founders of new companies in the US rigid disk drive industry over 1976–89 were members of an engineering team at a previous disk drive company who had developed a new disk drive architecture that their employer was unwilling to pursue. A particularly salient problem for large incumbent firms is the high levels of uncertainty often entailed by radical innovations. Responding to ‘short-termist’ investor pressure, large established incumbents often screen out ambiguous projects, preferring to invest in ones offering more predictable outcomes (Bhide, 2000; Freeman and Engel, 2007). Uncertainty also promotes divergent opinions between employees and employers about the value of a discovery. That makes it less likely they will agree on initiating a corporate venture, making independent starts more likely.24 Klepper (2006) develops a theory in which an incumbent firm’s managers agree about the firm’s current course until one manager comes into possession of a new piece of information about a valuable opportunity for the firm. This manager tries to persuade the other managers of the value of pursuing the opportunity. If this effort fails and the managers disagree, and if the opportunity’s expected value exceeds the cost of starting a new firm, then the well-informed manager may quit and start their own venture. This model generates an impressive number of empirical findings about spin-offs. One is that, because of his or her superior information, the manager’s spin-off venture will perform better than the parent firm and it will be socially beneficial for the spin-off to be formed. Another is that high-performing managers run high-performing new ventures. Furthermore, any event that decreases the decision-making influence of an informed manager (e.g. acquisition by an outside firm) is predicted to increase the chances of a spin-off. Another important consideration is that incumbents have incentives to avoid cannibalising their operations or risking their reputations on new ideas which may prove unsuccessful (Holmstrom, 1989; Freeman and Engel, 2007). These considerations are especially pronounced if customers prefer incremental to radical product changes (Christiansen, 1993; Christiansen and Bower, 1996). In addition, organisations undergoing crises are more likely to lose employees and spawn independent companies (Cooper, 1986; Klepper, 2006). The threat of job loss from a current employer, associated with closure or acquisition of the parent firm, or downsizing of management, can encourage employees to leave and start their own firms (Eriksson and Kuhn, 2006).

Theories of entrepreneurship

65

Market conditions characterised by numerous paradigm-shifting innovations can be another source of crisis, making incumbents more conservative while simultaneously creating positive opportunities for new independent ventures. This line of argument associates independent new starts with environments containing plentiful opportunities for radical innovation.

2.6.2 Agency cost theories Agency cost theories of entrepreneurial spawning centre on the incentive problems that occur when an employee discovers a significant invention and faces a choice between either leaving a firm to exploit it independently or remaining to develop it jointly with their employer. One source of conflict arises if the incumbent can exploit the opportunity more effectively owing to asset complementarity with the new innovation. Another emerges from asymmetric information problems which impede the ability of the employee to contract with the employer to develop an innovation jointly. Anton and Yao (1995) studied this problem in a setting where intellectual property rights (IPR) are weak. IPR is assumed to be weak both from the standpoint of the entrepreneur (whose idea can be expropriated by the employer) and from that of the employer (which cannot use trade-secret protection to prevent a private start-up). Anton and Yao (1995) showed that although joint profits are maximised if a joint venture outcome is chosen, imperfect IPR protection means that the incumbent sometimes does best by providing insufficient compensation to the entrepreneur, encouraging the latter to start up independently. Anton and Yao (1995) predict that the frequency of independent start-ups is greater:

• The more limited are the employees’ funds relative to the value of the invention, since limited funds prevent the employee from pre-committing to joint development by posting a bond; • The more similar are profits under an independent start-up to profits under a joint venture; • The greater the employer’s negotiating power to appropriate value from revealed inventions; • The less that innovations require distinctive complementary assets of existing businesses, and the more they rely on footloose human capital. Agency conflicts can come in other guises too. We saw earlier in this chapter that when firms cannot credibly commit to refrain from stealing workers’ compensation, the latter can be incentivised to start new firms (Wiggins, 1995). An implication of the Wiggins (1995) model is that workers are more likely to found new firms the longer they have to wait before output from their effort is observed. As noted earlier, an example is path-breaking innovations that take a long time to yield output, which might help explain why radical innovations are sometimes undertaken by small independent worker-owned firms backed with risk-capital, rather than by established firms.

66

Selection

Of course, incumbents wishing to discourage workers from quitting to found potentially competing start-ups need to design incentive contracts which deter workers from diverting effort into researching new opportunities in the first place. One solution to this moral hazard problem involves employers adjusting remuneration packages to deter quits and promote jointly developed spin-offs (Subramanian, 2005). Perhaps contrary to expectations, another solution is for employers actually to underbid for the rights to develop promising new ideas. By underbidding, the firm decreases workers’ incentives to divert effort from their main tasks. This can be optimal even if the firm possesses advantages over outside financiers such as venture capitalists owing to asset complementarities, or if underbidding runs the risk of losing employees to VCbacked start-ups (Bankman and Gilson, 1999). On other occasions, though, a potential entrepreneur might be more worried about an outside investor expropriating their idea, and might therefore prefer a joint corporate venture with their present employer. In joint ventures, employers presumably have fewer incentives to engage in opportunistic behaviour at the expense of entrepreneurs. On the other hand, entrepreneurs might have to cope with time-consuming internal ‘politics’ which impede development of their joint ventures. Hellmann (2007b) further explores the idea of effort diversion in a model where employees are paid to focus on a ‘main’ task but can privately divert unobservable effort into the development of new innovations. When the value of a new innovation is modest, Hellmann (2007b) shows that it can be optimal for employers to pre-commit not to develop any of them in-house, even though ex post it would seem to gain by reneging on these commitments and seizing the new opportunities. The strategy of reneging can be dominated by the employer’s policy of building a reputation for keeping promises in a repeated employment game, thereby focusing their employees most effectively on their main task. The cost to the firm is that it must resign itself to losing some employees who start up as entrepreneurs, if they hold the IPR. But if the value of new innovations is high enough, the firm might decide instead to break with its non-commercialisation policy and ‘go with the flow’ (assuming it holds the IPR) – choosing internal venturing (spinouts) instead. In other cases, entrepreneurial spin-offs can occur. Hellmann (2007b) provides a taxonomy of these cases. Klepper (2001) criticises the realism of agency theories which focus on strategic conflicts of interest between employers and employees. Klepper (2001) argues that although legal conflicts between parents and employee spin-offs do happen, they are relatively uncommon events compared with strategic disagreements, and are also less common than the simple dissatisfaction which employees experience owing to their employers refusing to develop new ideas. In any case, strategic conflicts of interest need not occur at all if the new firms compete in different markets from the incumbents, or if they supply the incumbents with intermediate inputs, as frequently is the case. Finally, asymmetric information in credit markets can help explain why intrapreneurship is more commonly observed in some industries and regions than in others. Suppose an investor can choose whether to fund a risky investment project within an existing firm (‘intrapreneurship’) or within a new firm (‘entrepreneurship’). The investor does

Theories of entrepreneurship

67

not know the individual abilities of recipients of the funds, but does know that if intrapreneurs fail in their investment projects, they will be redeployed within their firm if they are unlucky but talented, and will have to leave the firm immediately and apply for funds outside the firm if the firm realises that they are simply untalented. Gromb and Scharfstein (2002) show that two equilibria can emerge. In one equilibrium, there is a lot of intrapreneurship, so the pool of independents seeking finance is dominated by lots of low-ability failed intrapreneurs and entrepreneurs. This leads financiers to shun entrepreneurship and to back intrapreneurship as the preferred mode of commercialisation. In the other equilibrium, entrepreneurship is common and the pool of independents seeking finance is of high quality. This leads financiers to prefer entrepreneurship as the mode of commercialisation. 2.6.3 Learning theories According to these theories, employees use their work experience to acquire specific capabilities and learn about new venture opportunities (Buenstorf, 2007). This can facilitate independent start-ups in related technological areas. Parent firms might not find the costs of pre-empting these start-ups worthwhile if they occur rarely. In practice, the benefits of direct transfers of technology and specific capabilities from incumbents to employees appear to be less important than the value of general experience and knowledge gained from working in the parent firm, including those relating to internal firm ‘routines’ (Klepper, 2001). Gompers et al. (2005) had this idea in mind when they delineated a second type of spawning firm, exemplified by Fairchild Semiconductors, which (they argued) inadvertently prepared employees to be entrepreneurs by educating them about the entrepreneurial process and exposing them to a network of other entrepreneurs and venture capitalists. Klepper (2001) suggests that start-ups are well placed to adopt high-quality organisational routines from their parents, thereby disseminating good practice and technological progress across industries. This might explain why successful high-tech firms tend to spawn more rather than less successful firms. In particular, high-tech firms tend to have relatively complex production processes, so more of their employees are exposed to high-quality routines which they are able to replicate independently. This observation might in turn explain the high survival rates of spawned firms (see chapter 4). It might also explain why having multiple founders (who have presumably been exposed to numerous good routines) assists the survival of spawned firms, and why the product lines of spawned firms are often so similar to those of their parent firms, since this allows them to exploit their knowledge about similar routines.

2.7

Macroeconomic theories of entrepreneurship and growth

There are three broad macrotheoretical approaches to the study of entrepreneurship and economic growth. The first approach builds on wealth-based theories of entrepreneurship by proposing that as economies develop they accumulate wealth

68

Selection

through entrepreneurship, which promotes growth and further entrepreneurship. The role of entrepreneurs in this process, as capitalists who hire labour, borrow capital and produce output, is centre stage. Many of these models focus on the implications for economic development of entrepreneurial borrowing constraints. The second broad category of theories contains technology-based models, in which entrepreneurs respond to incentives created by technological change. The third category focuses on entrepreneurship as a means of exploiting knowledge spillovers, regarded by some researchers as the fundamental mechanism facilitating endogenous growth. Each of these broad approaches is now discussed in turn. 2.7.1 Wealth-based theories

Perhaps the best-known and most influential wealth-based theory of entrepreneurship and economic growth is that of Banerjee and Newman (1993). I will first sketch out the implications of this model before exploring other contributions to the literature. Banerjee and Newman (1993) proposed a model in which there are three occupations: ‘entrepreneurship’ (in which business owners pay a fixed investment cost to start up and hire a fixed number of workers), ‘self-employment’ (in which business owners operate small businesses with a smaller investment cost and hire no workers) and ‘paid employment’. Notice that the self-employed and entrepreneurs are defined as distinct occupational groups in this model. Not everyone can become self-employed or an entrepreneur: poor individuals lack the wealth needed to overcome borrowing constraints and so cannot raise the requisite investment cost (see also chapter 9). There are three social classes in the model, with workers being the poorest and entrepreneurs the wealthiest. The primary role of entrepreneurship in this model is to create wealth. But the distribution of wealth affects and is in turn affected by (constrained) occupational choices. The evolution of this dynamic feedback process traces out the development process. Every individual bequeaths a fraction of their lifetime wealth to their offspring, who similarly face constrained occupational choices. The offspring of the wealthy are also wealthy and can bypass borrowing constraints, giving rise to entrepreneurial dynasties. The existence of borrowing constraints injects sufficient friction into the model to ensure that the initial distribution of wealth has persistent effects on selection into entrepreneurship over the generations and on long-run economic development. Banerjee and Newman (1993) demonstrated that several possible development paths can emerge from their model. One is a ‘capitalist’development process, in which almost everyone is a worker and there are only a few very wealthy entrepreneurs. The large supply of workers depresses the average wage, which ensures that few workers in subsequent generations can overcome the borrowing constraints. That in turn ensures that wages remain low in the future. Entrepreneurs benefit from low wages since they entail lower production costs. This keeps them and their offspring wealthy, and perpetuates inegalitarian entrepreneurial dynasties as an equilibrium outcome.25 A different equilibrium in Banerjee and Newman’s (1993) model is associated with ‘artisanal’ development, in which self-employment is common and there are relatively

Theories of entrepreneurship

69

few workers. The small supply of workers keeps the average wage high, making entrepreneurship unattractive and perpetuates self-employment. The high wage also enables workers to accumulate enough resources to overcome the borrowing constraint and to join the ranks of the self-employed (but not of entrepreneurs). The equilibrium in this economy is one where self-employed artisans predominate. Banerjee and Newman (1993) suggest that the most beneficent equilibrium is one where there are not too many poor workers. Then high wages prevail, enabling subsequent generations of workers to rise up the wealth and social scales, become entrepreneurs in their turn, and making the economy increasingly prosperous and entrepreneurial. A poor economy can become rich if it does not start too far away from the rich one. But if it starts sufficiently poor it will be doomed to remain so, and entrepreneurship will stagnate. A policy implication of this line of thinking is that by relaxing borrowing constraints and allowing productive investment projects to be funded, financial development might be capable of boosting entrepreneurship (see also King and Levine, 1993). It should be borne in mind, however, that economic and financial development often go together, implying that financial development by itself is unlikely to be a panacea for slow economic development. Nevertheless, a recent calibration exercise suggests that financial intermediation enabling entrepreneurs to borrow to start businesses could be responsible for 7.3 per cent of the US economy’s total output, delivering an average welfare gain of 11.1 per cent relative to the case where financial intermediation is unavailable and people have to rely on their endowments of wealth to invest in new ventures (Bohác˘ ek, 2007). Some alternative wealth-based models of entrepreneurship and growth yield even more optimistic predictions about the prospects for long-run development. Aghion and Bolton (1997) proposed a model of moral hazard (outlined in greater detail in chapter 9) in which wealth incentivises effort. Economic development increases wealth and effort, thereby reducing moral hazard problems restricting lending, and so enabling more people to overcome borrowing constraints which formerly prevented them from becoming entrepreneurs. Eventually, everyone can become an entrepreneur and creditbased constraints cease to bind entirely as the economy passes a critical threshold. Two subsequent studies extended these wealth-based theories of entrepreneurship and growth by exploring how the distribution of heterogeneous entrepreneurial ability affects economic development. Lloyd-Ellis and Bernhardt (2000) show that two development paths can arise, depending on the distribution of entrepreneurial ability. If the latter is not too skewed, i.e. there is not a scarcity of innately talented entrepreneurs, wages as well as output, wealth and aggregate income continue to grow over time. The distributions of income and wealth eventually converge to unique limiting distributions which (unlike Banerjee and Newman, 1993) are independent of the initial distributions. Notably, entrepreneurial efficiency eventually replaces wealth as the primary determinant of occupational choice. The second development path is less benign, however. If most individuals have low entrepreneurial ability, i.e. the distribution of ability is positively skewed and efficient entrepreneurs are scarce, then a development path can

70

Selection

arise whereby the economy exhibits endogenous long-term cycles in production. Wages evolve pro-cyclically, so income inequality falls during booms and rises during recessions. The mechanism is as follows. Development-led wage increases transfer wealth from rich entrepreneurs (employers) to their less wealthy workers (employees). If there are few efficient entrepreneurs, wealthy entrepreneurs who lack much talent are easily persuaded to switch into paid employment as their wage costs rise and their profits fall. The increase in supply of workers then reduces the average wages. That stimulates the opposite process: having to pay lower wages, the supply of entrepreneurs increases – which in turn bids wages back up again. And so the cyclical process continues. In an interesting recent contribution, Acemoglu et al. (2006) propose a simple model of innovation, growth and entrepreneurial selection in which entrepreneurs reinvest their profits in their enterprises to bypass credit constraints. Economies can develop either by adopting existing technology or by innovating at the technology frontier. Acemoglu et al. (2006) argue that at early stages of economic development, re-investment enables mediocre entrepreneurs to predominate; and development proceeds by entrepreneurs adopting existing technologies. This is akin to wealth rather than entrepreneurial efficiency being the primary determinant of agents’ choices. But at later stages of development, as the economy approaches the technology frontier and credit constraints ease, able entrepreneurs replace mediocre entrepreneurs and carry out high-level innovation activities. Then entrepreneurial efficiency replaces wealth as the primary determinant of agents’ choices. On the basis of these predictions, Acemoglu et al. (2006) suggest that governments should choose institutional structures and policies which favour adopting ‘catch-up’ technology at early stages of development, followed by a switch into subsidising ‘high-tech’ innovation later on.

2.7.2 Technology-based theories Two broad theoretical perspectives will be discussed under this heading. One relates to the effects of technological progress on the occupational choice of entrepreneurship, while the other focuses on implications of technological progress for the prospects of small relative to large firms. The first set of theories is more formally based than the latter, which tend to be more conjectural in nature. Calvo and Wellisz (1980) and Schaffner (1993) propose formal models of firm-sizeneutral exogenous technological progress. Both articles predict a Lucas (1978)-type outcome – continually increasing numbers of workers and decreasing numbers of entrepreneurs. For example, Schaffner (1993) assumes that technological progress increases expected returns and risk for both employers and entrepreneurs. But by being able to costlessly smooth workers’ wages, risk-neutral employers gain an advantage over entrepreneurs who are assumed to lack opportunities to insure their incomes and so obtain a desired risk–return profile. This entices marginal entrepreneurs to quit and switch into paid employment. Furthermore, new (possibly de alio) risk-neutral employers can enter the market to exploit this advantage over entrepreneurs, bidding up wages and drawing even more individuals out of entrepreneurship and into paid employment.

Theories of entrepreneurship

71

Lazear’s (2005) jack-of-all-trades hypothesis suggests another reason why technological progress might erode the size of the entrepreneurial sector over time. If entrepreneurs need a mix of skills in production, then technological progress which demands additional skills requirements will decrease the number of suitably equipped individuals and therefore also the number of entrepreneurs. On the other hand, if a rapid pace of technological change is complementary with entrepreneurial ability, one would expect the number of entrepreneurs undertaking the most efficient projects to increase, with a reduction in the number of entrepreneurs undertaking the least efficient projects. ‘A higher rate of technological change is therefore likely to induce a cleansing effect on entrepreneurial activity and to alter the market perception of business creation’ (Crifo and Sami, 2008, p. 585). Another set of formal theories treats innovation and technological progress as endogenous outcomes. A literature on ‘patent races’ has emerged in which firms compete against each other to discover new patentable innovations that yield monopoly profits while rendering previous products obsolete à la Schumpeter.26 Profits are only temporary, however, because new innovations are eventually rendered obsolete, in Schumpeterian fashion, by future innovations. In some flavours of this model, imperfect competition is needed for Schumpeterian entrepreneurs to be willing to conduct growth-generating research (Aghion and Howitt, 1992). But subsequent extensions of this research programme suggest that under some circumstances competition and capital can be growth enhancing too (Aghion and Howitt, 1992; Howitt and Aghion, 1998). Arguably, however, this class of models does not bear closely on entrepreneurship. As Bianchi and Henrekson (2005) point out, these models primarily apply to large established firms which devote resources to large-scale ‘routinised’ R&D activities. Patent races therefore have little to do with entrepreneurial innovations and ignore issues of entrepreneurial choice by not really distinguishing between entrepreneurs and inventors. In contrast, several alternative approaches, analysed below, place entrepreneurs at the heart of models of endogenous technological progress and economic growth. Less formal theories in contrast speculate about the changing nature of production functions wrought by technological change, and their effects on the relative advantages of small relative to large firms. Some authors, from at least the time of Schumpeter, have claimed that technological progress tends to be biased in favour of large firms, making the latter’s economies of scale increasingly dominant over time, thereby squeezing out smaller producers and hence (by implication) entrepreneurs. This could be one factor behind the ongoing decline of the small independent producer in agriculture in particular; it has long been known for instance that the size share of the agricultural selfemployment sector and per capita GNP are strongly negatively correlated (Kuznets, 1966; Schultz, 1990). Garicano and Rossi-Hansberg (2006) propose another reason why independent entrepreneurship may decline with technological change. Technology reduces communication costs, working to the advantage of team-based production in large firms, which leverages high levels of communication between team members. Thus independent entrepreneurship can be progressively replaced by corporations on efficiency grounds.

72

Selection

On the other hand, it has been suggested that technological change can also work the other way, favouring smaller rather than larger businesses. Arguably, the advent of more flexible manufacturing techniques has begun to make scale economies less important, reducing minimum efficient scale and facilitating competition from smaller firms.27 These changes might also have endowed small firms with a comparative advantage at supplying markets with low or fluctuating levels of demand. The emergence of new technologies, including ICT and telecommunications, and the growing demand for labour-intensive services and ‘niche’ customised products might also have favoured smaller firms, by requiring less physical capital and more human capital (although see Bhide, 2000, chap. 14, for a sceptical view). Further theory-grounded quantitative research is needed to shed light on the extent to which technological change of this sort has impacted on entrepreneurship. 2.7.3

Knowledge-based theories Neoclassical theories of economic growth in the 1960s emphasised the role of savings and capital accumulation in promoting growth. Key to those theories was the notion of an aggregate production function q = Aq(k) linking per capita income q with capital per head k via an increasing concave function q(k), where A is the exogenous state of technology. For simplicity, these theories often also assumed a constant savings rate 0 < γ < 1 and constant rates of population growth g and capital depreciation δ > 0. According to the classic Solow growth model, economies will grow for as long as the supply of capital, γ Aq(k), exceeds the demand for capital, (g + δ)k. The concavity of q(·) and the straight-line nature of (g + δ)k ensures that the two must come into equality for some k > 0: this characterises the steady-state equilibrium. In this model, long-run growth is unaffected by the savings rate but instead is predicted to be driven by exogenous increases in A. By the late 1980s neoclassical growth theory had fallen out of favour. It was unable to explain the determinants of A (famously described as ‘falling like manna from heaven’) and hence long-run growth. And mounting evidence rejected the theory’s prediction that savings rates and economic growth are uncorrelated in steady-state. Romer (1986) offered a solution to these problems by proposing an endogenous theory of growth. This theory essentially replaces a concave q(·) (diminishing marginal returns to capital) with a straight-line function: q(k) ∝ k. Now it is possible for γ Aq(k) to permanently exceed (g + δ)k, leading to ongoing (‘permanent’) growth and a positive relationship between savings rates and growth rates – consistent with the evidence. The crucial contribution made by Romer is the replacement of a concave aggregate production function with a straight-line one. This makes no sense at all at the level of the individual firm, where it implies increasing returns-to-scale to all factors of production taken together, with the counter-intuitive implication that one single firm would end up dominating the entire economy. Of course, nothing like this ever remotely happens in reality. Romer’s insight was to propose that individual firms continue to operate with a concave production function, but that external spillovers give the economy as a whole the straight-line production function associated with continuous ongoing growth.

Theories of entrepreneurship

73

This raises the question about what the external spillovers are and where they come from. At a general level, it has been known for a long time (at least since Arrow, 1962) that knowledge and ideas are public goods, whose benefits are only partially captured by their creators. Knowledge creation has the tendency to carry within it the seeds of further knowledge creation, which might not be obvious to the creator but which can be recognised and developed by others. This is a positive externality which is commonly referred to as ‘knowledge spillovers’. Knowledge spillovers can ‘end up facilitating other agents’ innovation efforts (either unintentionally, as happens when inventions are imitated), or intentionally, as may happen when scientists divulge the results of their research’ (Breschi and Lissoni, 2001, p. 975). Economists have been fairly quick to recognise the role of entrepreneurs in promoting and exploiting knowledge spillovers. In an early contribution, Schmitz (1989) argued that entrepreneurs’ search and investment activities inadvertently add to the stock of knowledge which others can exploit. Consequently, entrepreneurship is a possible source of positive externalities at the level of the economy. An implication of positive externalities is that competitive equilibria are not welfare-maximising. There will in general be too little entrepreneurship relative to the social optimum. Hence policy-makers should actively promote technology-based entrepreneurship in order to stimulate economic growth. It is helpful to delve more deeply into the origins of knowledge spillovers. Two important sources are universities and private R&D laboratories. Both are important producers of ‘codified’ and ‘tacit’ knowledge. Codified knowledge is easily documented, transferable and reproducible (e.g. information published in a scientific article), and virtually costless in the age of the Internet to diffuse across local and national boundaries (Krugman, 1991). Such knowledge can be transmitted instantaneously through specialist interest groups such as online user communities. In contrast, tacit knowledge cannot be completely documented; it is often context-specific and best transmitted via face-to-face interactions and frequent repeated contacts (von Hippel, 1994). Costs of transmitting tacit knowledge rise with distance, so it tends to be locally concentrated. An example is new ideas developed through close discussion between a few experts. If tacit knowledge is a crucial ingredient of new commercial innovations, then geographical proximity is essential to access it or even to become aware of its existence. Propinquity also facilitates networking, trust, co-operation and social capital needed to exploit new opportunities arising from access to this type of knowledge (Saxenian, 1994; Thornton and Flynn, 2003). It remains to explain how entrepreneurs exploit knowledge spillovers. As noted in the previous section, one conduit is employees of firms which perform R&D but which do not develop the new opportunities in-house. That might be because of inflexible internal ‘routines’ or because the managers of R&D firms do not value the ideas’ potential as much as entrepreneurs do (Audretsch and Thurik, 2001a, p. 12). In which case, new venture creation can be the only way of diminishing the ‘knowledge filter’ between the creation and exploitation of knowledge. Naturally, entrepreneurship is not the only way of exploiting knowledge spillovers: others include

74

Selection

corporate venturing and opportunistic exploitation by existing firms which specialise in adapting new technologies originating in other firms (Cohen and Levinthal, 1989). But entrepreneurship is likely to play an important role in the process. In short, growth can be created by individual entrepreneurs who exploit knowledge via new ventures even though they do not contribute directly to the production of knowledge themselves. In contrast to the theories of patent races mentioned above, entrepreneurs and new venture creation play a central role in these theories of knowledge-based technological change and economic growth. Broadening out the implications of these ideas, Acs et al. (2004a) argue that entrepreneurship will become progressively more important in knowledge-based economies by providing one of the main channels through which new economic knowledge can be commercialised. As globalisation drives down wages and employment in traditional tradable sectors (especially manufacturing) in developed countries, and as multinational employers become increasingly footloose and stateless, entrepreneurship becomes increasingly important by enabling developed countries to exploit their comparative advantages in developing innovative knowledge-intensive goods. This has prompted Audretsch (2003) to argue that in an era of rapid globalisation, public policy ought to involve the ‘strategic management of places’ rather than the management of plants or firms. A fuller discussion of public policy relating to knowledge spillovers is deferred until chapter 16; evidence about the existence and importance of knowledge spillovers is reviewed in chapter 4. 2.8

Multiple equilibrium models

Some formal models attempt to explain why regions with similar or identical socioeconomic profiles can end up with markedly different qualities and quantities of entrepreneurship. This section briefly discusses three such models. The first is a model of serial entrepreneurship by Landier (2004), which predicts how similar regions can find themselves in different equilibria in terms of the quality of entrepreneurship. The second, by Ghatak et al. (2007), is a productivity-based explanation of differences in the quality of entrepreneurship. The third is a human capital model by Parker (2005b) which predicts how similar regions can end up with different quantities of entrepreneurship. The last two parts of the chapter appendix sketch out simplified technical versions of the first and third of these models for the interested reader. In Landier’s (2004) model, entrepreneurs possess either high or low ability. Able entrepreneurs periodically encounter attractive new business opportunities which they can only pursue by closing down their existing business and raising new capital. By assumption, new projects are more attractive to entrepreneurs than existing ones. At the same time, less able entrepreneurs can also apply for new capital for banks, which they secretly consume before defaulting. Banks face an asymmetric problem, because when confronted with a set of loan applicants they cannot distinguish the able serial entrepreneurs (who generate profit for the bank) from the less able serial entrepreneurs (who generate losses for the bank).

Theories of entrepreneurship

75

Two distinct equilibria arise in this model. In one, which Landier (2004) calls the ‘conservative’ equilibrium, the cost of capital is high. This deters able types from becoming serial entrepreneurs, as they have to cross-subsidise the less able types. So the borrower pool is swamped with less able types, justifying the high cost of capital and locking in this outcome as an equilibrium. The other equilibrium, called ‘experimental’, is associated with a low cost of capital. Now able types have incentives to exploit their profitable opportunities to become serial entrepreneurs. That creates a more favourable borrower pool from banks’ perspective, allowing them to break even with a lower interest rate. Neither of the equilibria in Landier’s model is unambiguously more efficient than the other, although the experimental equilibrium is more efficient in sectors close to the technological frontier. The conservative equilibrium generates too little serial entrepreneurship to exploit valuable opportunities efficiently in these sectors. Which initial conditions determine whether an economy finds itself in a conservative or an experimental equilibrium? Landier speculates that in some places the ‘stigma of failure’is high, which makes serial entrepreneurship unattractive. Then the conservative equilibrium is more likely to prevail. Landier identifies the USA with an experimental equilibrium and Europe with a conservative equilibrium, although one could also imagine multiple equilibria to arise within countries with non-integrated (i.e. regionally separate) markets. Landier’s (2004) model neatly analyses both multiple equilibria and serial entrepreneurship within one simple framework; but it is not the only or even necessarily the most convincing theoretical model of either multiple equilibria or serial entrepreneurship. Ghatak et al. (2007) point out that if a non-pecuniary benefit to being a successful entrepreneur is sufficient to attract very able entrepreneurs into the market despite low financial payoffs, high average entrepreneurial ability makes employees very productive, resulting in high wages (thereby confirming low entrepreneurial profits) and establishing an equilibrium with numerous well-paid workers. But if payoffs in entrepreneurship are high, this can attract less able people to become entrepreneurs, reducing the average productivity of the workforce and resulting in low wages (thereby confirming high entrepreneurial profits). The quality and quantity of the pool of entrepreneurs is very different in both equilibria, as are average living standards. In principle, either equilibrium can arise. Multiple equilibria in the quality of entrepreneurship also arise as a natural consequence of an observation, noted in the previous section, that technological progress increases the quality of entrepreneurial projects while being spurred in turn by improvements in the average quality of entrepreneurs (Crifo and Sami, 2008). In a ‘high-level’ equilibrium, technological progress and the probabilities of entrepreneurial success are high. Banks reward entrepreneurs with low interest rates, which encourages entrepreneurs to select high-quality technology-rich projects, which further adds to the technological progress underpinning this equilibrium. The ‘low-level’ equilibrium, in contrast, is characterised by low levels of technological progress, and the self-selection of less able entrepreneurs who perpetuate this state of affairs as an equilibrium outcome.

76

Selection

I now turn to the issue of multiple entrepreneurship equilibria in the quantity of entrepreneurship. In Parker (2005b), individuals have to decide how much human capital to acquire before learning about their idiosyncratic entrepreneurial ability. Human capital is assumed to be valuable in both entrepreneurship and paid employment, but generates a higher marginal return for the ablest people in entrepreneurship compared with paid employment (à la Lucas, 1978). To make optimal human capital choices, individuals must estimate expected returns in entrepreneurship in their region, which they can only do by observing local entrepreneurs. The fact that individuals make decisions based on observing local levels of entrepreneurship in their own region is what gives rise to multiple entrepreneurship equilibria. A ‘low-level’ entrepreneurship equilibrium arises when individuals deduce from the current occupational structure that they have a low probability of being able enough to prosper in entrepreneurship, and so acquire relatively little human capital. These choices predispose them to choose paid employment over entrepreneurship once their true ability is revealed, because having low human capital they cannot access high marginal returns in entrepreneurship while having to pay a fixed entry cost. But that perpetuates the initial occupational structure and locks in low levels of human capital and entrepreneurship as an equilibrium. Conversely, a ‘high-level’ entrepreneurship equilibrium can arise and persist for the opposite reason. Clearly an important assumption in this model is that entrepreneurs remain in their region and are not well informed about entrepreneurial ability in other locations. Evidence that entrepreneurs have low levels of geographic mobility lends some credence to this assumption (see chapter 4).28 In this model as in Landier’s (2004) and Ghatak et al. (2007), one does not need to appeal to exogenous behavioural, cultural or institutional differences to generate different equilibria in different regions. Instead, multiple equilibria arise from imperfect information and incomplete co-ordination of agents. However, Parker (2005b) concludes that the scope for practical policy interventions to shift entrepreneurship equilibria from low to high level outcomes is probably limited. This might account, for example, for the limited practical success of regional development agencies in changing regional profiles of entrepreneurship within the EU.

2.9

Conclusion

This chapter has reviewed a wide range of influential economic theories of entrepreneurship, both microeconomic and macroeconomic. Microeconomic theories predict what types of individual are likely to become entrepreneurs; why an entrepreneurial occupational choice exists at all; and the role of incumbent firms in the entrepreneurial entry process. Macroeconomic theories trace out the implications of these microeconomic choices for the broader economy and for aggregate rates of entrepreneurship in different countries and regions. Both sets of theories generate a range of hypotheses which can be tested and falsified in the accepted scientific tradition. Chapter 4 will review the salient evidence on these issues.

Theories of entrepreneurship

77

Modern economic theories of entrepreneurship clarify what can, and cannot, be said about the determinants of entrepreneurship. For example, we saw that greater market risk in entrepreneurship does not necessarily reduce the equilibrium number of entrepreneurs. One reason for this is that in general equilibrium, greater risk in entrepreneurship also affects employees indirectly by decreasing the equilibrium wage. The chapter also explained why rational individuals facing sunk-entry costs in a risky trading environment do not always switch between entrepreneurship and paid employment despite apparently being able to make clear-cut gains from doing so. As a further example, the theories reviewed in this chapter explained why some employees leave large incumbent firms to start up new ventures, while others do not. Nevertheless, numerous gaps in our knowledge remain. With regard to the famous Lucas (1978) model, for example, we still know relatively little about where entrepreneurial ability comes from, and to what extent it can be shaped by schooling and enterprise education. This relates to the age-old question of whether entrepreneurs are ‘born or made’. In addition, recent research about risk-sharing via partial income insurance questions the robustness of risk-attitude-based theories of entrepreneurship, notably Kihlstrom and Laffont’s (1979). It is presently unclear whether these models need to be amended or are intrinsically flawed. And theories about the boundaries of the firm and contracting out need to be developed further in future research, as they set limits on the extent to which an entrepreneurial occupational choice ever arises. To the extent that entrepreneurship is a viable occupational choice, the economics of entrepreneurship argues that individuals only become entrepreneurs if that occupation maximises their expected utility. This insight forms the starting point of empirical models of entrepreneurial choice reviewed in the next chapter. 2.10 Appendices 2.10.1

Technical definitions of risk aversion and risk Consider the case of risk-averse agents. Then Uyy (y) < 0 < Uy (y), where subscripts denote derivatives. The following definitions propose some useful ways of quantifying risk aversion.

Definition 1 Given a twice-differentiable utility function U (y), the Arrow–Pratt coefficient of absolute risk aversion at income y is defined as r A (y) = −Uyy (y)/ Uy (y) > 0. Definition 2 The utility function U (y) exhibits decreasing absolute risk aversion if r A (y) is a decreasing function of y. Definition 3 Given a twice-differentiable utility function U (y), the coefficient of relative risk aversion at income y is defined as r R (y) = −yUyy (y)/Uy (y). The concept of absolute risk aversion is useful for describing preferences over risky outcomes that involve absolute gains or losses of income. In contrast, relative risk aversion is more appropriate for risky situations where outcomes are percentage gains

78

Selection

or losses of income. Individuals whose preferences are described by decreasing absolute risk aversion (Definition 2) take more risks as they become better off. While this often yields economically reasonable results about risk-taking behaviour, it is sometimes too weak and is complemented by the stronger assumption of non-increasing relative risk aversion. This assumption states that individuals become more willing to risk fractions of their income as their income increases. It is a stronger assumption than decreasing absolute risk aversion because, by r R (y) = yr A (y), decreasing relative risk aversion implies decreasing absolute risk aversion, but the converse does not necessarily follow. It is sometimes convenient to consider the case where both apply, encapsulated in the following assumption: Assumption 1 The utility function U (y) exhibits decreasing absolute risk aversion and non-increasing relative risk aversion. The notions of ‘an increase in risk’ considered in the text, namely second-order stochastic dominance (SOSD) and mean-preserving spread (MPS), rank two return distributions, with distribution functions F(y) and G(y). Consider the following ranking:



U (y) dF(y) ≥

U (y) dG(y),

(2.2)

where U (·) does not necessarily have to be (though often is) the utility function defined earlier. Definition 4 (Second-order stochastic dominance) For any distributions F(·) and G(·) with the same mean, F(·) second-order stochastically dominates (is less risky than) G(·) if, for every non-decreasing function U (·), (2.2) holds. Definition 5 (Mean preserving spread) For any distributions F(·) and G(·) with the same mean, G(·) is a mean preserving spread of (i.e. is in this sense riskier than) F(·) if, for U (·) some concave function, (2.2) holds. SOSD evidently places less structure on U (·) than MPS, which is in turn a more general measure of ‘increase in risk’ than an increase in variance because it implies (but is not implied by) the latter. Under both definitions, every risk-averter prefers F(·) to G(·). For completeness, I also state a further useful definition: Definition 6 (First-order stochastic dominance) The distribution F(·) first-order stochastically dominates G(·) if, for every non-decreasing function U (·), (2.2) holds. Definition 6 implies that every expected utility maximiser who prefers more to less prefers F(·) to G(·). Equivalently, for any amount of money income y, the probability of getting at least y is higher under F(·) than under G(·).

Theories of entrepreneurship

79

2.10.2 A simple ‘hire or outsource’ model

Consider a principal, P, who wishes to elicit high levels of effort from an agent, A, who produces a good, S. If high effort is chosen (e = 1), A incurs a utility cost of 1 and produces a high level of output, S, with probability π1 > 0, and a low level of output S < S with probability 1 − π1 . If low effort is chosen (e = 0), A’s effort cost is zero, and the probabilities of high and low output are π0 > 0 and 1 − π0 respectively: π := π1 − π0 > 0. Thus high effort benefits P, but signals of effort are noisy: it is still possible that low effort yields high output, since π0 > 0. To elicit high effort from employee A, P contracts to pay A the wage t when S is observed, and the wage t < t when S is observed. The contracted values of (t, t) will be derived as solutions to P’s incentive problem described below. In entrepreneurship, A obtains total utility of U (t)+ξ −1 if they supply high effort, where U (t) is a concave function of remuneration t, and ξ > 0 is a non-pecuniary benefit from being an entrepreneur (see chapter 4). Let h = U −1 be the inverse function of U (·), which is increasing and convex (h
> 0,

h > 0). Hence writing U = U (t) and U = U (t) as the utilities of agents, we have t = h(U ) and t = h(U ). If P keeps the production of S in-house, her programming (moral hazard) problem is: max π1 [S − h(U )] + (1 − π1 )[S − h(U )] + φ

(2.3)

π1 u + (1 − π1 )U − 1 ≥ π0 U + (1 − π0 )U

(2.4)

(U ,U )

subject to

π1 U + (1 − π1 )U − 1 ≥ U (w),

(2.5)

where (2.4) is the incentive compatibility constraint designed to elicit high effort, and (2.5) is A’s participation constraint. The first two terms of (2.3) are P’s expected returns net of wage costs h(U ); the last term φ ≥ 0 is an ‘asset specificity’ value of producing S in-house, which is lost if S is made elsewhere and bought in by P. ∗ The solution to this standard principal-agent problem is a tuple of values (t , t ∗ ), where   (1 − π1 ) ∗ (2.6) t = h 1 + u(w) + π  π1  t ∗ = h 1 + u(w) − (2.7) π ∗

Notice that t ∗ < t ensures that the agent bears some risk: this is necessary for the agent to supply high levels of privately costly effort. By comparing P’s expected payoffs under high and low agent effort [which are given by (2.3) and π0 S + (1 − π0 )S + φ − w, respectively], it is only worthwhile for P to ∗ encourage high effort by offering ˜t ∗ ∈ {t ∗ , t } if the following condition holds: ∗

Sπ ≥ π1 t + (1 − π1 )t ∗ − w.

(2.8)

80

Selection

If it does not hold, P does best by settling for low worker effort, and paying A the outside wage w. ∗ An alternative strategy to paying ˜t ∗ ∈ {t ∗ , t } (or w) is for P to transfer to A the residual claimant’s rights to S, in return for a certain upfront payment, T ∗ . This changes A’s employment status from an employee to an independent residual claimant (‘entrepreneur’). Notice that entrepreneurs will always supply the first best levels of effort to produce S because they are the residual claimants and bear the full costs from supplying low effort (see Parker, 2007b, for a formal proof). If entrepreneur A supplies the good to P, then A is willing to accept the change of employment status (including the payment to P of T ∗ ) if π1 U (S − T ∗ ) + (1 − π1 )U (S − T ∗ ) + ξ − 1 ≥ U (w),

(2.9)

T∗

= T (S, S, π1 , ξ , w). which implicitly solves at equality for Clearly, P only agrees to buy in S from A – and thereby replace a paid employment relationship with an entrepreneurial relationship – if this makes P better off relative to the case of retaining direct control, i.e. if  ∗ if high effort is induced π1 (S − t ) + (1 − π1 )(S − t ∗ ) + φ ∗ T > (2.10) π0 S + (1 − π0 )S + φ − w if low effort is induced Note that (2.9) and (2.10) embody the principle of Pareto optimality, since A is no worse off and P becomes strictly better off under contracting out. To see how Pareto gains can be made, observe that as an employee, A benefits from partial insurance provided by P, while P benefits from employing A in-house because of asset-specificity. But by becoming an entrepreneur, A can obtain the nonpecuniary benefit ξ , which by (2.9) can be sufficiently strong to compensate for losing her insurance. Effectively, ξ increases the value that A is willing to pay for the rights to S, which is necessary in order to convince P to relinquish the asset-specificity value, φ. It is noteworthy in this regard that evidence discussed elsewhere in this book supports the notion that entrepreneurs are willing to pay premiums in the capital and labour markets in return for non-pecuniary benefits (see chapters 9 and 13). Regarding performance measurement, consider a hypothetical firm in which output is either insensitive to variations in worker effort, or is hard to measure accurately. Then S is too small for (2.8) to hold, so P does best by settling for low worker effort and paying A the flat employment wage w. This prediction is exactly in line with the argument of Holmstrom and Milgrom (1994) mentioned in the text, who expect nonincentivised employment contracts to appear when output is hard to measure. Note that by ensuring that high effort is forthcoming, P’s decision to unbundle the firm and make A independent can result in greater effort-based efficiency. This simple model also predicts that costs of monitoring workers encourage firms to outsource work to independent contractors. If the expected value of in-house production S˜ is reduced by high worker monitoring costs, the contracting-out condition (2.10) is more likely to hold for any given firm. These costs are irrelevant for contractors, because being residual claimants they have no incentive to shirk.

Theories of entrepreneurship

81

2.10.3

Landier’s serial entrepreneurship multiple equilibrium model Consider an economy with a unit mass of risk-neutral entrepreneurs. There are two periods and all entrepreneurs are active in both of them. Every entrepreneur borrows a unit of capital at the start of the first period, which is used to finance a project lasting two periods. Half of the agents are high quality or ‘good’ entrepreneurs, g, who end the second period able to repay principal plus interest of D > 0 to their lender. These entrepreneurs receive the good payoff Rg > D at the end of the second period with probability one and never default on their loans. However, at the start of the second period, a random proportion 0 < q < 1 of the g entrepreneurs are presented with an opportunity to close their first project, forego Rg and borrow another unit of capital to finance a second project, with a higher return R > Rg . If they take this opportunity, these entrepreneurs repay 2D at the end of the second period. The other half of the entrepreneurs are low quality or ‘bad’ entrepreneurs, b, who all operate a project generating a return of at least D if they are successful, with probability 1 − q. These (1 − q)/2 entrepreneurs repay the bank D at the end of the second period. If they are unsuccessful (with probability q), they secretly consume the income and borrow another unit of capital at the start of the second period to keep the project going. They dissemble to banks, claiming that they are g types with a profitable new project to invest in. For simplicity, suppose that even with this second round of finance all of these q/2 entrepreneurs end up defaulting at the end of the second period. All the bank can retrieve from these projects is Rb , where Rb < D. Banks are perfectly competitive and face an exogenous outside cost of capital, ρ > 1. Entrepreneur quality is private information and hidden from banks in both periods. There is therefore an asymmetric information problem (see chapter 7). Banks’ expected profits net of capital repayments if g types choose to become serial entrepreneurs, a case indexed by 1, are



π1B

1+q =D 2





1−q +D 2



q + Rb − ρ(1 + q) = 0, 2

where the first and second terms are expected repayments from successful g and b types, respectively; the third term is expected repayments from unsuccessful b types; and the last term is the cost of the 1 + q/2 + q/2 = 1 + q units of capital used. Setting π1B to zero yields the competitive repayment amount q D1 = ρ(1 + q) − Rb . 2

(2.11)

If the q/2 g types choose not to become serial entrepreneurs, a case indexed by 2, banks’ expected returns become π2B =

   1 q 1−q q D+D + Rb − ρ 1 + = 0, 2 2 2 2

82

Selection

yielding the competitive repayment D2 =

  ρ 1 + q2 − q2 Rb 1−

q 2

> D1 .

(2.12)

Which of D1 and D2 prevails in equilibrium determines whether the good entrepreneurs with the opportunity become serial entrepreneurs. The key point is that this in turn determines which of D1 and D2 prevails. This realisation lies at the heart of Landier’s multiple equilibrium result. To see this formally, note that the q g types in question choose between payoffs of R − 2D if they become serial entrepreneurs and Rg − D if not. All entrepreneurs are price-takers: they take the required repayments in competitive credit markets as given. Suppose Rg is such that the following condition holds: D2 > R − Rg > D1 . These inequalities imply respectively R − 2D2 < Rg − D2

and

R − 2D1 > Rg − D1 .

(2.13)

These inequalities respectively state that entrepreneurs optimally choose not to close their businesses and become serial entrepreneurs in locations where the cost of capital is high, but take them in locations where the cost of capital is low. These decisions change the mix of the borrower pool and determine the respective costs of capital in the different locations. That is, entrepreneurs who return to the credit market in locations where the cost of capital is high will disproportionately be low-quality types who failed in their first business. This generates losses for banks, which therefore have to raise the cost of capital to break even. Of course, this deters serial entrepreneurship and reinforces this equilibrium, which Landier calls ‘conservative’. In the other, ‘experimental’, equilibrium where the cost of capital is low, it becomes worthwhile for high-quality entrepreneurs to close their current businesses and start a new one in the hope that they will achieve a very high return. Once these better entrepreneurs join the borrower pool, competitive lenders can charge a lower cost of capital. Of course, this is the condition which encouraged serial entrepreneurship to occur in the first place, and establishes this second equilibrium. 2.10.4

Parker’s human capital multiple equilibrium model √ Consider individuals whose return to human capital, h, is either ψ(h) = x h in entrepreneurship (where x ∈ [0, 1] is uniformly distributed entrepreneurial ability) √ or w(h) = α h in paid employment. Here α is a parameter restricted to lie between zero and one in order for a mixture of occupations to arise in equilibrium. The population size is normalised to unity. All individuals have to choose their optimal human capital investment h∗ before their ability is revealed later on; the best they can do when choosing h is to observe average ability among the 0 < n < 1 individuals observed to be entrepreneurs in the region, i.e. the conditional expectation E(x|n). Hence their best guess of the probability that they will end up with sufficient ability to become an entrepreneur is n. The probability that they do not, and so end up choosing paid

Theories of entrepreneurship

83

employment, is 1 − n. If the marginal cost of human capital acquisition is c > 0, and if entrepreneurs pay an entry cost of D, then

√ √ h∗ = argmax n[E(x|n) h − D] + (1 − n)α h − ch The first term is the expected net return in entrepreneurship multiplied by the probability that individuals end up in it; the second term is analogous for paid employment. The first-order condition for this problem is nE(x|n) + (1 − n)α − c = 0. √ 2 h

(2.14)

Once individuals stop schooling they realise their entrepreneurial ability, x. They then choose their occupation, which is entrepreneurship for all and only those with x > x˜ , where the marginal entrepreneur x˜ is identified by √ J (˜x) := (˜x − α) h∗ − D = 0. (2.15) The uniform distribution implies that x˜ = 1 − n, and that E(x|n) = E(x|x ≥ x˜ ) = 1 − (n/2). Substituting this last expression into (2.14) yields   n(1 − α − n/2) + α 2 ∗ h = . 2c

(2.16)

Finally, putting (2.16) into (2.15) and noting that x˜ = 1 − n yields an occupational choice locus J (n) = EN (n) − PE(n) = 0, (2.17) where EN (n) := (1 − n)[n(1 − α − n/2) + α] − 2cD PE(n) := α[n(1 − α − n/2) + α]. The intersection of these two curves, EN (n) and PE(n), solves (2.17) and identifies up to two occupational choice equilibria, n ∈ {n1 , n2 }. Parker (2005b, pp. 838–9) provides necessary and sufficient conditions for both equilibria to exist, expressed in terms of the model parameters. One can as a result conceive of two regions identical in all respects except that one has high rates of entrepreneurship, n2 , and the other has low rates of entrepreneurship, n1 < n2 , where n1 = inf {n ∈ [0, 1]|J (n) = 0}

and

n2 = sup{n ∈ [0, 1]|J (n) = 0} .

Provided x2 < 1 − α (as seems to be the case in practice: Parker, 2005b, p. 837), the model also implies that chosen human capital is higher in the high-entrepreneurship region: h∗ |n2 > h∗ |n1 .

84

Selection

Notes 1. Knight viewed profits as a fluctuating residual, not as a return to the entrepreneurial factor of production. In contrast, Hawley (1907) regarded profit as a reward to risk-taking. According to Hawley, the entrepreneur needs to be the owner of an organisation, in order to exercise decisionmaking power. What sets the entrepreneur apart from managers is his or her willingness to be the ultimately responsible agent in the productive process, liable for ownership of output but also for losses. 2. Also, according to Klein and Cook (2006), T. W. Schultz’s conception of entrepreneurship – articulated in an unpublished 1979 paper – is the ability to adjust or reallocate resources in response to changing economic circumstances. 3. See Stiglitz (1970), Carroll (2002) and Charles and Hurst (2003). See also chapter 4, for evidence on the question of whether entrepreneurs are less risk-averse than non-entrepreneurs. 4. More ambiguous effects of risk are observed if output prices rather than quantities are random: see Appelbaum and Katz (1986). When the random price realisation is low, both entrepreneurial profits and employee wages decrease. 5. Alternative stochastic income processes tend to be intractable. See e.g. Li (2002) for such a model with Markov processes in both paid employment and entrepreneurship.  6. The cumulative distribution function, or cdf, is F(x) = xx f (χ) d χ . Hence F(x) = 0 and F(x) = 1.  7. The average size of firms is q[1 − F(˜x)]−1 x˜x xf (x) dx, which is increasing in x˜ . 8. This is because average ability is increasing in x˜ , as implied by the previous footnote. But a higher x˜ implies fewer entrepreneurs, who number 1 − F(˜x). 9. See e.g. Calvo and Wellisz (1980), Oi (1983), Blau (1985), Bond (1986), Brock and Evans (1986), Kuhn (1988), Murphy et al. (1991), de Wit and van Winden (1991), Laussel and Le Breton (1995), Yamada (1996), Lloyd-Ellis and Bernhardt (2000), Fonseca et al. (2001), Parker (2003c, 2005b), Fender (2005), Lazear (2005), Bohà˘cek (2006), Acemoglu (2008), Antunes et al. (2008) and Gollin (2008). 10. This possibility was recognised as early as Roy (1951), who considered the implications of free occupational choice for productivity when ability affects payoffs differently in two distinct occupations. 11. As will be seen in chapter 7, de Meza and Webb (1987) obtained a similar result in the credit rather than the labour market. 12. Alternatively, able workers demand ex ante screening to reveal their abilities to the general public and so gain utility from higher status (Parker and van Praag, 2009). In either case, it is assumed to be too costly for banks to screen entrepreneurs ex ante as well as ex post. 13. The KL79 model is based on an assumption of perfect competition. Welfare effects have also been studied for this model under imperfect competition. In that case there are entrepreneurial rents which are not competed away. This attracts more entrepreneurs, which increases social welfare relative to the case of perfect competition (Clemens, 2006). These rents partly offset the negative effects of limited entrepreneurial risk taking, though the latter is still associated with reduced economic growth (Clemens, 2008). 14. Of course, other types of contracting out also exist, e.g. making workers redundant and buying in their services from non-entrepreneurial firms. These sorts of arrangement are not the focus of interest here. 15. This model, drawn from Parker (2007b), can also be used to analyse a different aspect of the make-or-buy decision, namely the ‘retail or franchise’ decision, in which the principal chooses between paying employees to sell the product through company-owned retail outlets or paying franchisees to do so. Mazzeo (2004) analyses this issue and shows how the market environment affects the choice of independent or franchisee status in the US motel industry.

Theories of entrepreneurship

85

16. This provides an alternative reason to Casson’s (2003) association of entrepreneurship with the exercise of judgement, since judgement is likely to be needed in jobs which are costly to monitor and in which independent business ownership is the optimal organisational form. 17. In terms of the formal model, the contracting-out condition (2.9) would not hold for highly risk-averse workers. 18. Small scale and modest returns can make non-profit-maximisers unattractive acquisition targets for profit-maximisers, assisting their independent survival. Nevertheless, some not-for-profit enterprises protect themselves formally by forming foundations. 19. Drucker (1990), Rose-Ackerman (1996) and Baron (2007). 20. Proof: Suppose the entrepreneur sells a unit of a good at price P which embodies an innate quality q. The entrepreneur obtains utility βq from supplying a good of quality q, where β > 0 could reflect either altruism or a reputation advantage to quality. The cost of producing quality is c(q), where c(·) is a convex function. If the entrepreneur adopted FP status for his firm, he would choose q to maximise π = P − c(q) + bq, yielding a first-order condition c
(q) = b. Denote the quality that satisfies this condition by qf . If the entrepreneur adopts NP status, she cannot receive profits but receives perquisites instead, that are worth less to her than profits, namely d π, where d < 1. Hence her first-order condition is c
(q) = b/d . Let qn denote the quality that solves this equation. Clearly c
(qn ) = b/d > b = c
(qf ). By convexity of c(q), this implies qn > qf . Finally, if consumers value quality, the price P they are willing to pay is increasing in q, so P(qn ) > P(qf ). 21. See Storey (1994a), Boden (1996), Wagner (2004), Eriksson and Kuhn (2006), Mueller (2006b), Sørensen (2007), Hyytinen and Maliranta (2008) and Parker (2009b). Wagner (2004) and Sørensen (2007) also observe that young firms are associated with worker quits leading to entrepreneurship; but size seems to matter more than age (Sørensen, 2007). 22. However, Sørensen’s (2007) study uses a selected sample of people who enter entrepreneurship at some point in the sample frame. Hence sorting (self-selection) might still be an important part of the story, even if it cannot explain the entire story. 23. See Winter (1984), Henderson and Clark (1990) and Henderson (1993). 24. Audretsch (2001), Klepper and Sleeper (2005) and Klepper (2006). 25. A similar prediction of a persistent ‘class structure’ which separates people into rich employers and poor workers, who have no choice but to work for the employers for a low wage, emerges from the model of Matsuyama (2006). Endogenous investment decisions by entrepreneurs in Matsuyama’s (2006) model locks the class structure into place. 26. See e.g. Segerstrom et al. (1990), Segerstrom (1991), Aghion and Howitt (1992) and Peretto (1998). 27. See Piore and Sabel (1984), Dosi (1988), Carlsson (1989, 1992), Acs et al. (1990), Carlsson et al. (1994) and Wennekers and Thurik (1999). 28. The model also implies that all employees over-invest in human capital, while the ablest entrepreneurs under-invest in human capital relative to the first-best case of perfect information (Parker, 2005b, proposition 3).

3

Empirical methods in entrepreneurship research

The economics of entrepreneurship is notable for its careful use of econometric methods, and for eschewing the practice of asking entrepreneurs or other agents what they think they will do in various situations. Responses to these kinds of question are known to be prone to self-serving bias, or ‘cheap talk’, and respondents are liable to give answers which they think interviewers regard as ‘appropriate’. Instead, the ‘revealed preference’ principle trains economists to distrust individuals’ declared intentions, forcing them to undertake the harder but more objective task of inferring preferences and constraints from their actual behaviour. This often necessitates the use of advanced statistical techniques to overcome econometric problems that might otherwise vitiate empirical estimates. Examples of such problems include: • Sample selection bias, whereby sample membership is not random but instead is generated by some (at least partially) observable systematic process; • Unobserved heterogeneity, whereby important unmeasured idiosyncratic variables are omitted from a regression model; • Endogeneity, whereby an ‘independent’ variable is itself codetermined within the structural model of interest; and • Non-stationarity, whereby time-series variables follow unit root processes that violate a key assumption of the classical linear regression model and lead to invalid statistical inference. The present chapter outlines several ‘canonical’empirical models in the economics of entrepreneurship which address all of the problems on this list. Throughout, knowledge of regression analysis is taken as given; readers not versed in it are referred to Greene (2003) or one of innumerable alternative standard textbooks. It is worth noting at the outset that I will focus primarily on multivariate models. Such models enable the researcher to control for a potentially wide range of explanatory variables when conducting an empirical analysis. Correlations and (nonparametric) difference-in-means approaches are useful as simple descriptive tools and they continue to be deployed in some parts of the non-economics entrepreneurship literature. However, they are ill-suited for inference about structural relationships in entrepreneurship, because the omission of other conditioning variables generally creates 86

Empirical methods in entrepreneurship research

87

bias (Greene, 2003). For this reason, and also because they are widely taught, simple descriptive methods will not be discussed in this chapter either. The chapter is arranged according to the type of data available to the researcher: cross-section, time-series or panel. The first three sections deal with cross-section models, introducing sample selection and instrumental variables (IV) methods; binary occupational choice models; and several useful extensions of those models. The fourth section deals with issues which arise when the researcher possesses a sample of timeseries data. The fifth section discusses panel-data models. The sixth section focuses on entrepreneurial duration, or ‘survival’, models. Details of an especially technical nature are relegated to footnotes and the chapter appendix.

3.1

Cross-section regression models: sample selection bias and IV

Consider the simple regression model, written in matrix form as Yi = γ
Xi + vi

i = 1, . . . , n ,

(3.1)

where Yi is some continuous variable, Xi is a set of regressors (including a column of ones for an intercept), γ is a vector of parameters to be estimated, and vi is a random disturbance, assumed to be normally distributed. Each observation is denoted by i: the sample size is n. For example, Yi might be the financial performance of entrepreneur i at a given point in time, while Xi contains relevant individual- and venture-specific characteristics. Such models are often estimated by ordinary least squares (OLS). The objective of OLS estimation is to obtain unbiased, or at least consistent, estimates γˆ of the true but unknown parameter vector γ . If γˆ is consistent, any bias (defined as |γˆ − γ |) progressively vanishes and estimates become more precise (efficient) as the sample size n increases. The OLS estimator only yields consistent or unbiased estimates if several assumptions hold. Two important cases where they do not hold include nonrandom composition of the sample {Yi , Xi }; and non-independence of Xi from vi . These two cases will now be discussed in turn.

3.1.1

Sample selection bias Suppose, for the sake of argument, that a researcher is interested in estimating (3.1) in order to discover the determinants of entrepreneurs’ sales or profits, Yi , in terms of a set of explanatory variables, Xi . Unbiased or consistent estimates of γ will allow the researcher to predict accurately the effects of changes in those variables on entrepreneurial profits, provided the sample is representative of all individuals in the population. The danger, however, is that the sample is not representative, because individuals have unmeasured characteristics which predispose them to perform particularly well in entrepreneurship and hence select into the sample. In this case, we say that sample selection bias occurs. Without dealing with this bias, researchers are unable to predict for example whether non-entrepreneurs with particular characteristics

88

Selection

X could improve their lot by switching into entrepreneurship. A popular practical way of correcting for selection bias is Heckman’s (1979) method. Heckman’s method comprises two steps. The first step estimates a participation equation z i = ω
Wi + u i ,

(3.2)

where zi is an indicator variable equalling 1 if individual i is an entrepreneur and 0 otherwise; Wi is a set of explanatory variables, ω is a vector of coefficients, and ui is a disturbance term, with unit variance. Details about how to estimate a binary model of this sort appear in the next section. In the second step, the researcher computes fitted values zˆi from (3.2), and uses them to form a new variable called the ‘Inverse Mills Ratio’ λi = −φ(ˆzi )/(ˆzi ), where φ(·) and (·) are the density and cumulative distribution functions of the standard normal distribution. This variable is then added to the right-hand side of (3.1), yielding the augmented regression Yi = γ
Xi + αλi + i ,

(3.3)

where α > 0 implies positive self-selection into entrepreneurship and α < 0 implies negative selection. Equations (3.2) and (3.3) can be estimated jointly by maximum likelihood (ML) or in two separate steps. Most modern software packages do one or both automatically. The Heckman correction enables consistent estimates of the regression coefficients γ to be obtained. Applications of this method to earnings functions are discussed in section 3 of this chapter, while empirical estimates of α are reviewed in chapter 4. The Heckman correction method has been applied to such diverse topics as participation in public-sector entrepreneurship programmes (Wren and Storey, 2002) and models explaining entrepreneurs’ work hours (Parker, Belghitar and Barmby, 2005), among others. Arguably, however, it remains under-used in applied entrepreneurship research as a whole. 3.1.2

Endogeneity and IV If Xi is not independent of vi in (3.1), OLS no longer provides consistent estimates of γ . Specifically, if Xi includes endogenous, mis-measured, omitted or lagged dependent variables, OLS estimates of γ will generally be biased. As in Murray (2006), such variables will be termed ‘troublesome’ in what follows. Endogeneity biases in particular are especially widespread in the entrepreneurship field, for two principal reasons. First, researchers regularly explain performance in terms of variables whose values entrepreneurs choose themselves. But if choices are shaped by individual characteristics, these variables are likely to be correlated with unobserved characteristics captured within vi , rendering them endogenous. Second, several explanatory variables assumed to be causes of entrepreneurship are also consequences of it, such as wealth, income

Empirical methods in entrepreneurship research

89

and economic growth. Cases like these will be highlighted in future chapters as and when the problem arises. As a constructive solution to these problems, Instrumental Variables (IV) estimators consistently estimate coefficients γ even when ‘troublesome’ variables appear in Xi . To fix ideas, write Xi = [Mi , Wi ] in (3.1), where Mi is a set of troublesome variables, and Wi is a vector of non-troublesome variables. IV is made possible by a set of variables Zi that are (i) uncorrelated with the error term vi ; (ii) correlated with the troublesome variables Mi ; and (iii) not members of Xi . The elements of Zi are called ‘identifying instrumental variables’, or simply ‘identifying instruments’. The most frequently used IV estimator is two-stage least squares (2SLS). At the first stage of 2SLS, the researcher identifies troublesome variables and identifying instruments and uses OLS to estimate Mi = α
Zi + β
Xi + i ˆi (where i is a random error term). In the second stage of 2SLS, the fitted values M are substituted for Mi in the regression of interest, (3.1). The second-stage estimates   ˆ i , Wi ]
[M ˆ i , Wi ] −1 [M ˆ i , Wi ]
Yi are called 2SLS estimates. γˆ = [M 2SLS requires at least as many instruments Zi as there are troublesome explanatory variables Mi . When there are as many (resp., more) instruments as (resp., than) troublesome explanatory variables, we say the equation of interest is exactly (resp., over-) identified. In principle, the more instruments there are the more efficient the IV estimators will be. In practice, however, it can be hard to find instruments which satisfy all of the criteria (i)–(iii) above. Ideally the choice of instruments should be guided by theory; however this is not always straightforward and suitable empirical constructs are not always available in actual data sets. Murray (2006) discusses the serious problems that can arise when the researcher implements IV using invalid or weak instruments, e.g. instruments which are not uncorrelated with the error term or are only weakly correlated with the troublesome variables. In cases like these, the cure can be worse than the disease. To illustrate the IV estimator, Parker and van Praag (2006a) estimated the following ‘triangular’ model of entrepreneurial performance: π = β0 + β1 x1 + · · · + βJ xJ + βs sch + βc con + u

(3.4)

sch = η0 + η1 x1 + · · · + ηJ xJ + θ1 z1 + v

(3.5)

con = γ0 + γ1 x1 + · · · + γJ xJ + γs sch + θ2 z2 + ω

(3.6)

where π is gross log entrepreneurial profit; x1 through xJ are exogenous control variables (i.e. the Wi above); sch measures years of schooling; and con is a measure of how capital-constrained an entrepreneur was when they started up. These last two variables are potentially endogenous, so correspond to the Mi variables in the preceding discussion. Now let z1 and z2 be vectors of ‘identifying’ instrumental variables which do not appear anywhere in (3.4). To generate consistent estimates of the parameters of interest (chiefly βs , βc and γs , predicted by Parker and van Praag’s theoretical model

90

Selection

to be positive, negative and negative respectively), it is essential that z1 and z2 possess explanatory power and are genuinely independent of the disturbance terms u, v and ω. Based on a rationale explained in chapter 9, Parker and van Praag (2006a) chose father’s education and the number of siblings as instrumental variables z1 (yielding over-identification); and they chose capital intensity of the entrepreneur’s industry sector as the instrumental variable z2 (exact identification). Statistical tests described in Parker and van Praag (2006a) indicated that these identifying instruments were indeed exogenous. To demonstrate the practical importance of IV in terms of inference, Parker and van Praag (2006a) analysed a data sample of Dutch entrepreneurs taken from the mid1990s and obtained OLS estimates βˆs and βˆc (obtained by estimating (3.4) on its own) of 0.072 and −0.003 (with t statistics of 2.45 and 1.14 respectively). In contrast, IV estimates (obtained by estimating (3.4) through (3.6)) by 2SLS were 0.137 and −0.039 (with t statistics of 2.01 and 2.23 respectively). Thus IV implies larger and more significant effects from schooling and borrowing constraints on performance than OLS – illustrating the practical importance of correcting for endogeneity bias. Parker and van Praag (2006a) argue that these findings imply that government policies designed to promote entrepreneurship should seek to enhance the national stocks of both human and physical capital. There is growing use of IV estimation in the economics of entrepreneurship, as will become apparent from the discussion of various empirical findings throughout this book. However, applications of IV in ‘non-economics’ entrepreneurship research are still in their infancy; it is expected and hoped that many more applications will follow in the future. 3.2

Cross-section binary models of occupational choice

Binary limited dependent models are especially widely used in entrepreneurship research. We encountered one in the previous section, namely (3.2); below, the two most common binary models – probit and logit – are explained. Binary models are mainly, but not exclusively, used to model entrepreneurship as an occupational choice. Consider two occupations denoted by j: entrepreneurship, E, and paid employment, P. Each individual has a vector of observed characteristics Wi and derives utility Uij = U (Wi ; j) + uij if they work in occupation j, where U (·; ·) is utility that can be observed by the econometrician and uij is idiosyncratic unobserved utility. Denote by zi∗ a ‘latent’ variable which measures the relative utility advantage to i of being in occupation E relative to P. That is, zi∗ = U (Wi ; E) − U (Wi ; P) + uiE − uiP . If we assume that U (·; ·) is linear, taking the form U (Wi ; j) = βj
Wi , where βj are vectors of coefficients, then we can write zi∗ = α + β
Wi + vi ,

(3.7)

Empirical methods in entrepreneurship research

91

where β
:= βE
−βP
is another vector of coefficients; α := E[uiE −uiP ] is an intercept; and vi := uiE − uiP − α ∼ IIN (0, σ 2 ) is a disturbance term. Henceforth the intercept term is incorporated in Wi as a column of ones, so β will be treated as the complete set of coefficients. Individual i chooses entrepreneurship over paid employment if zi∗ ≥ 0. Hence as in (3.2) above, define the observable binary occupational indicator variable  1 if individual i is observed in E, i.e., if zi∗ ≥ 0 zi := 0 if individual i is observed in P, i.e., if zi∗ < 0 Therefore the probability that an individual is observed to be an entrepreneur in a representative sample, with characteristic vector Wi , is Pr(zi = 1) = Pr(zi∗ ≥ 0) .

(3.8)

The probit model assumes that the distribution of the disturbance term vi is normal. Hence Pr(zi = 1) = (β
Wi /σ ) and Pr(zi = 0) = 1 − (β
Wi /σ ), where (·) is the (cumulative) distribution function of the normal distribution. The model is estimated numerically by Maximum Likelihood (ML) and is programmed in most modern software packages.1 The logit model arises if the distribution function of vi is assumed to be that of the logistic distribution, in which case (3.8) becomes Pr(zi = 1) =

exp{β
Wi } . 1 + exp{β
Wi }

(3.9)

A likelihood function can be formed in a similar way as for the probit model and β estimated in the same manner (see the previous note). Logit models are routinely programmed in most software packages. Both logit and probit dominate OLS estimation of zi = β
Wi + vi (called the linear probability model), since OLS is an inefficient and heteroscedastic estimator in this context, and problematically can predict probabilities outside the unit interval (see e.g. Maddala, 1983). It makes little practical difference which of probit or logit is used, as estimates of β are relatively insensitive to this choice.2 Econometricians call the general approach outlined above ‘reduced form’. This is because the logit and probit empirical specifications are derived from a utility maximisation problem in which the indirect utility function U (Wi ; j) is simply assumed to depend on several observable covariates in a convenient way. In contrast, ‘structural’ occupational choice models relax this assumption and instead either estimate or calibrate the ‘deep’ parameters of agents’ objectives and constraints. A recent example of a structural occupational choice model is Paulson et al. (2006); but the structural approach in entrepreneurship research can be traced back at least as far as Brock and Evans (1986). There are at least three distinct applications of binary choice models to occupational choice. One (I) focuses on the probability that at a snapshot of time individuals

92

Selection

are entrepreneurs rather than employees. A second (II) asks a different question: what factors affect the decision to become an entrepreneur, as opposed to remaining in paid employment (or unemployment)? A third (III) investigates entrepreneurs’ decision to leave rather than to continue in entrepreneurship. All of these applications can be handled by binary choice models: they merely require different definitions of the dependent variable, zi . However, it is possible to dispute the relative merits of each. For example, Evans and Leighton (1989b) argued that II is preferable to I, on the grounds that I (unlike II) confounds entry and survival effects. This is because the probability of being an entrepreneur at time t depends on the probability of switching into entrepreneurship at some previous time and then remaining as an entrepreneur until t. On the other hand, as Wellington (2001) has pointed out, application II excludes people who are already successfully practising entrepreneurship, which is evidently a group of some interest. Low annual switching rates from paid employment to entrepreneurship (see chapter 2, section 2.2) also means that II sometimes suffers from small numbers of observations of entrepreneurs; and the characteristics of switchers may well be different from those of non-switchers. There are more serious methodological problems with application III, which will be discussed separately below. I would simply point out that all three applications can generate useful information and play an important role in applied entrepreneurship research. To conclude, a vast number of studies have implemented probit and logit binary choice models. Many of the findings based on them will be reviewed in chapter 4. 3.3

Extensions of the cross-section binary model

Five useful extensions of the cross-section binary occupational choice model are considered below: 1. 2. 3. 4. 5.

The inclusion of relative incomes Multiple occupational choices Multiple equation systems Non-binary occupational choices Heteroscedastic probit

3.3.1

The inclusion of relative incomes

Microeconomic theory teaches us that relative prices generally influence individual choices. If this precept is true for occupational choice, then one of the explanatory variables in the matrix W above should be relative income, or its logarithm: (ln yiE − ln yiP ). However, we know from section 3.1 that occupational incomes are potentially endogenous and prone to selection effects, so the binary choice model needs to be modified somehow in order to obtain efficient and unbiased estimates of the parameters of interest.3 The first stage of the probit model with relative incomes estimates selectivitycorrected earnings functions (see chapter 13) separately for entrepreneurs and

Empirical methods in entrepreneurship research

93

employees. Letting M denote the vector of (exogenous) explanatory variables used in the earnings functions, one estimates the equations z i = β
Wi + v i [ln yiE |zi = 1] = γE
Mi + ϑE λiE + uiE [ln yiP |zi = 0] =

γP
Mi

+ ϑP λiP + uiP

i ∈ {E, P}

(3.10)

i∈E

(3.11)

i ∈ P,

(3.12)

where λiE = −φ(ˆzi )/(ˆzi ) and λiP = φ(ˆzi )/[1 − (ˆzi )] are the Inverse Mills Ratios to correct for selectivity into each occupation. Eq. (3.10) is sometimes called the ‘reduced form probit’ and is not of direct interest: its principal role is to generate zˆi values used to correct for selection bias in the earnings functions (3.11) and (3.12). Stage two of the binary choice model with relative incomes generates the predicted
yiE log incomes from both occupations derived from (3.11) and (3.12), namely ln
and ln yiP . Predicted incomes must be obtained because individuals are usually only observed in one occupation at any point in time in a cross-section sample. The third stage estimates the misleadingly named (see the previous footnote) ‘structural probit’ model



zi = α[ln yiE − ln yiP ] + ω
Xi + v˜ i ,

(3.13)

where X is a further vector of explanatory variables, and where v˜ i , uiE , uiP and vi are all assumed to be normally distributed disturbance terms. If occupational choices do indeed depend on relative financial returns in each occupation, then α > 0. A t-statistic can therefore be used to test the hypothesis that an individual is more likely to be an entrepreneur the greater is their relative income in entrepreneurship. Estimation at each stage is easily accomplished using standard econometric software, taking note of the following four points. First, this model has a 2SLS structure, and is therefore an implementation of IV. Consequently identification issues arise. This model is only properly identified (i.e. the estimated parameters only relate to the model parameters of interest) if the researcher includes in X at least one variable not included in M or W , and vice-versa. For example, it is common practice to exclude variables relating to the presence of children in i’s household from Mi , but to include them in Wi and Xi , on the grounds that family size affects occupational choices but not earnings (see Hammarstedt, 2006; and chapter 4). It is conceptually harder to think of variables that affect earnings (Mi ) but not occupational choice (Wi and Xi ). Of course, any chosen restrictions should be tested for validity in empirical work, as observed in the discussion of IV above. Second, because the relative income variable in (3.13) has been generated from previous regressions (i.e. it is a ‘generated regressor’), Newey–West corrected standard errors should be used. Third, income under-reporting by entrepreneurs may bias yiE and hence αˆ in (3.13). This problem can be overcome by applying income underreporting corrections (if available) to the data at the outset (see chapter 13). Fourth, one should ensure that the disturbances of the earnings equations are normally distributed

94

Selection

(Bernhardt, 1994). This can be achieved by suitable transformations of the earnings variables in (3.11) and (3.12) (see Parker, 2003b). Unfortunately, previous applications of the probit model with relative incomes have tended to ignore most or all of these points, leading to biases of unknown magnitude. Chapter 4 will review estimates of the binary model with relative incomes, focusing on estimates of α in (3.13). 3.3.2

Multiple occupational choices Sometimes there may be more than two occupations to choose from. For example, Earle and Sakova (2000) studied the problem of choosing between employer selfemployment, own-account self-employment, paid employment, or unemployment. As another example, entrepreneurs might also be able to choose between several venture development paths. Thus Parker and Belghitar (2006) modeled nascent entrepreneurs’ decisions between launching a nascent venture, continuing to prepare their venture for eventual launch, and terminating the founding effort altogether. Multiple occupational choices can be handled by using multinomial choice models. Perhaps the most popular model within this class is the multinomial logit (MNL) model, which can be regarded as an extension of the simple logit model described above. In this model, individual i must choose between j = 1, . . . , J alternatives. Define zij as equal to one if i chooses j, and zero otherwise. Then the MNL model proposes that the probability i chooses j is

exp{βj
Wi + γ
Xj } . Pr(zij = 1|Wi , Xj ) =

j exp{βj Wi + γ Xj }

(3.14)

Here Wi is a vector of variables whose values vary across individuals, whereas Xj is a vector of variables whose values vary across occupations. The βj coefficients must vary across occupations or else they cannot be identified. Analogous to the genesis of the simple binary model discussed above, (3.14) is the probability that results from a choice problem in which individuals maximise utility across each alternative, where the utilities are given by Uij = βj
Wi + γ
Xj + uij (see Parker and Belghitar, 2006, for details and an illustration). The βj and γ coefficients can be estimated by maximum likelihood, a procedure that is routinely incorporated in most standard econometric software packages. 3.3.3

Multiple equation systems I will discuss here two useful applications of multiple equation systems for occupational choice in entrepreneurship. First, although individuals may wish to become entrepreneurs, they may not have the opportunity to do so. This motivates the use of the bivariate probit model, which separates individuals’ opportunities to become ∗ be a latent variable representing entrepreneurs from their willingness to do so. Let z1i ∗ represent willingness. Potentially different factors impinge on opportunities and let z2i

Empirical methods in entrepreneurship research

95

opportunity and willingness, suggesting the specifications ∗ z1i = β1
W1i + v1i

(3.15)

∗ z2i

(3.16)

=

β2
W2i

+ v2i ,

where (v1i , v2i ) are distributed as bivariate normal variates with correlation coefficient ρ: the joint distribution function is (·, ·; ρ).4 To identify this model, it is necessary that W1 = W2 .As with the IV and binary choice models with relative incomes, the researcher must impose testable identifying restrictions on the basis of a priori reasoning. For example, the height of industry entry barriers might affect entrepreneurs’ opportunities but not their willingness to enter. Applications of the bivariate probit model include van Praag and van Ophem (1995) in an analysis of occupational choice between entrepreneurship and paid employment; Henley (2007), who analysed determinants of aspirations to become an entrepreneur jointly with actual transitions to entrepreneurship; and Fairlie (2006a), who specified z1∗ and z2∗ in (3.15) and (3.16) as predispositions for business ownership and computer usage, respectively. Using matched CPS (Current Population Survey) and Computer and Internet Usage data from 1997–2001, Fairlie estimated ρ to be insignificant, implying that entrepreneurs do not have unmeasured characteristics in v1i that predispose them to use computers via v2i (or vice versa). A second extension generalises (3.15) and (3.16) to allow for interdependent occupational choices: ∗ ∗ z1i = β1
W1i + γ1 z2i + v1i

(3.17)

∗ ∗ z2i = β2
W2i + γ2 z1i + v2i .

(3.18)

The introduction of the γ terms in (3.17) and (3.18) allows, for example, a husband’s occupational choice to affect his wife’s, and vice versa. This type of structure appears to be well suited for explaining why a disproportionate number of entrepreneurs are married to other entrepreneurs (Parker, 2008a). The model (3.17) and (3.18) is identified if at least one variable, 1 say, appears in W1 but not W2 , and if at least one variable, 2 , appears in W2 but not W1 . The reduced form of this model is given in the first part of the chapter appendix. An important implication of this simultaneous equation model is that ignoring interdependence in entrepreneurial choices when it is actually present can generate seriously misleading inferences. This warning is of practical relevance because most empirical studies of business ownership do precisely that.5 There are two principal ways of estimating the simultaneous probit model (3.17) and (3.18). One is by a consistent two-step estimator (2SE) proposed by Maddala (1983, chap. 8.8).6 The second method is Full Information Maximum Likelihood (FIML). Parker (2008a) provides an application using a sample of Panel Study of Income Dynamics (PSID) data.

96

Selection

3.3.4 Non-binary occupational choices All of the models discussed so far assume that individuals choose to spend all of their time in either entrepreneurship or paid employment. However, many people have broken spells in entrepreneurship. Even among those with unbroken spells, some individuals mix their work time between multiple occupations. In particular, employees commonly start new businesses part-time while continuing to work for their employer (Carter et al., 1996; Delmar and Davidsson, 2000). A simplistic way of dealing with the problem of broken spells would be to measure involvement in entrepreneurship as the number of hours worked as an entrepreneur over a particular time period (e.g. Kolvereid and Isaksen, 2006). But this approach includes numerous non-entrepreneurs in the sample (e.g. professors doing consulting on the side) – while from a technical point of view it does not fully exploit the structure of the data. Recognising these limitations, Burke et al. (2008) propose the following model. Define a variable hE as the time spent in entrepreneurship over the last T years: 0 ≤ hE ≤ T . In a typical sample, hE values will be bimodal, with the dominant mode at hE = 0 (most sample respondents have no involvement at all in entrepreneurship) and a secondary mode at some hE > 0 (the modal work hours of entrepreneurs). Burke et al. (2008) recommend a two-stage estimator to first estimate the determinants of the outcomes hE = 0 rather than hE > 0 (using a binary model), and then to estimate the determinants of hE |hE > 0, i.e. of hE conditional on hE > 0.7 3.3.5

Heteroscedastic probit Binary choice models such as (3.2) will yield biased and inconsistent results if the variance of the error term is not constant across sample observations (i.e. is not homoscedastic). In such cases, the error term is said to be heteroscedastic. If the form of the heteroscedasticity is known, σ [in (3.31), note 1] should be replaced with σi , where σi is an estimated function of some covariates, Mi , which can be specified by the researcher. A convenient form is σi = exp{γ
Mi }. Fraser and Greene (2006) propose a simple model of Bayesian learning by entrepreneurs which gives rise to the heteroscedastic probit model. It is described in the second part of the chapter appendix. Fraser and Greene (2006) went on to estimate it using a sample of British Social Attitudes Survey (BSAS) data. The heteroscedastic component of the model turned out to be statistically and economically important in their analysis.

3.4 Time-series models

Consider again (3.1), but now for data observed for time periods t rather than individuals i. For example, Yt might be an aggregate measure of entrepreneurship in a country, such as the self-employment rate, which is related to a set of economic conditions in that country, Xt . One rationale for estimating a time-series model of entrepreneurship is that, unlike cross-section studies, it becomes possible to analyse trends in entrepreneurship. Also, the time-series approach can identify determinants of

Empirical methods in entrepreneurship research

97

entrepreneurship which are uniform for all or most members of a cross-section sample at a given point in time, e.g. income tax variables, interest rates and other macroeconomic variables. Prior to the 1990s, the preferred technique for estimating (3.1) using time-series data was ordinary or generalised least squares (OLS or GLS: see, e.g., Blau, 1987; Steinmetz and Wright, 1989). Since then, however, it has become known that least squares estimators are inappropriate when any of the variables in the set {Yt , Xt } are non-stationary. In simple terms, a non-stationary process is one in which there is no mechanism forcing values of a series to revert to its mean. The application of OLS or GLS to non-stationary data is known to generate misleading results in which genuinely independent variables spuriously appear to be related to each other, and in which classical inference is invalid (Phillips, 1986). The R2 goodness-of-fit measure is no longer informative and t and F statistics can no longer be used for hypothesis testing. The importance of this point is underlined by the fact that national self-employment rates in the USA, UK and most other OECD countries appear to be non-stationary in practice.8 Hence this point appears to be of considerable practical importance. When time-series variables are non-stationary, it is necessary to check whether they co-integrate, i.e. whether there exists at least one linear combination of the variables (called a co-integrating vector) that is stationary. If so, there is said to be a long-run (non-spurious) relationship between the variables. It is then possible to obtain consistent estimates of the coefficients of that relationship, and to perform appropriate hypothesis tests on the coefficients. It is also possible to examine a dynamic (‘short-run’) errorcorrection model that describes how agents behave out of equilibrium. The interested reader is referred to Greene (2003, chapter 20) for an introduction to co-integration analysis. Parker (1996) proposes the following ‘working guide’ for estimating and performing inference on the parameters of (3.1) using time-series data and co-integration methods: 1. Check each variable in (3.1) is non-stationary using unit root tests. 2. If at least two variables are non-stationary, use a multivariate co-integration estimator to identify the number of co-integration vectors. 3. If there is a unique co-integration vector, test if Yt is weakly exogenous; and perform significance tests on each element of γ . 4. Estimate an error-correction model using the co-integrating residuals vˆ t in order to determine the short-run determinants of entrepreneurship. Co-integration methods are slowly becoming more widely used in the economicsbased entrepreneurship literature (e.g. Torrini, 2005; Bruce and Mohsin, 2006). However, as a general rule, researchers in the business studies tradition seem to be less aware of these methods (see Shane, 1996, and Choi and Phan, 2006, as examples). Hopefully, this situation will change in the future. Also, as more years of data become available, results based on co-integration analysis gain greater credence, as the power

98

Selection

of unit root and co-integration tests tends to be low in small samples covering only a short time span. 3.5

Panel-data models

If time-series data are available for the same set of individuals (or ventures or countries) then a ‘panel’ of data is available. Panel data combines the advantages of the rich casespecific variation of cross-section data with the temporal variation of time-series data. It also enables the researcher to control for cohort- and unobserved person-specific effects – something which cannot be done with (repeated) cross-sections. Furthermore, panel data can increase the explanatory power of cross-section binary choice models which often lack predictive power (Reynolds, 1997). Suppose that a panel comprises n individuals i observed over T time periods t. A simple pooled regression model corresponding to (3.1) is Yit = γ
Xit + α + vit , where α is an intercept common to all individuals. A more general specification allows the intercept to vary across individuals, giving rise to the fixed-effects model: Yit = γ
Xit + αi + vit .

(3.19)

For example, consider an application in which Yit measures the rate of entrepreneurship in region i at time t. Then the αi can be interpreted as the net effects of unobserved (and time-invariant) cultural or institutional factors affecting entrepreneurship in region i (e.g. Grilo and Thurik, 2006; Freytag and Thurik, 2007). It should be noted that covariates Xit which are not time-varying cannot be included in fixed-effect models, as such covariates are indistinguishable from the fixed effects, αi . A solution to this problem is to use a random-effects estimator. The random-effects model is similar to (3.19), except that it assumes the intercepts are drawn from a common distribution with mean α and variance σα2 . Fixed- and random-effects estimators are both routinely implemented in modern software packages. Robson and Wren (1999) and OECD (2000) are two examples of panel-data models of entrepreneurship which take the cross-section units i to be countries rather than individuals. Letting  denote the difference operator (e.g. Xt ≡ Xt − Xt−1 ), both of these studies estimate the following dynamic specification of self-employment rates, Yit :  ln Yit = αi + β
Xit − γ ln Yi t−1 + ω
Xi t−1 + ϑt + vit ,

(3.20)

where X is a matrix of explanatory variables; αi is a country-specific intercept term; and ϑt is a set of time dummies. The β coefficients capture short-run effects of variables on self-employment rates, whereas the ω coefficients pick up their long-run effects. Other noteworthy examples of fixed-effect regression models include Audretsch and Thurik (2001a), Acs et al. (2004) and Mueller (2006a), who all analyse the relationship between entrepreneurship and cross-country economic growth; and Georgellis and

Empirical methods in entrepreneurship research

99

Wall (2006), who analyse entrepreneurship across US states and over time. Results of these studies are discussed in chapter 4. Random-effects models are more common in applications where the cross-section dimension relates to individuals rather than countries, partly because problems of time-invariant covariates (e.g. individuals’ genders and highest educational qualifications) are more prevalent in this case. When using panel-data models with a large time-series dimension, the researcher should remember that as in the time-series case, least squares estimates of panel models are biased if any of the variables in the model are non-stationary. Co-integrated panel techniques are appropriate in this case (see Parker and Robson, 2004, for details). In contrast, the remainder of this subsection turns its attention to the case where the time-series dimension of the panel is small (e.g. a few years) and the cross-section dimension is large (e.g. several thousand respondents). Examples of data sets with these characteristics include the US PSID, the UK BHPS and the European Household Panel Survey (EHPS). Panels with a large cross-section dimension allow researchers to generalise binary models like (3.7) in two especially useful ways. As noted above, one is by introducing fixed effects to control for unobserved heterogeneity. The other is by modelling statedependent choices, which captures the tendency for individuals to remain in the same occupation for many periods, perhaps because of switching costs (see chapter 2).9 In an application using BHPS data on self-employment participation over 1991–99, Henley (2004)foundthatfixedeffectsandstatedependencebothplayedamajorroleinexplaining self-employment propensities. First, the fixed effects αˆ i accounted for nearly 60 per cent of the unexplained variance in self-employment propensities, consistent with the view that many determinants of self-employment choices are idiosyncratic and not easily amenable to explanation by available socio-economic variables. Second, in terms of state dependence, Henley found that a Briton who was self-employed last year is 30 per cent more likely to be self-employed this year than someone who was in paid employment last year. Together with the finding that local unemployment rates explain little of selfemployment participation, the results suggest that persistence in self-employment is a structural rather than a cyclical phenomenon. That provides a possible explanation for the unit root processes in aggregate self-employment rates reported in the previous section. Strikingly, Henley (2004) reported that the significance of several ‘conventional’ explanatory variables declined sharply upon controlling for fixed-effects and state dependence. Hence findings based on simple cross-section models can generate misleading inferences and should be treated with caution. An important implication is that entrepreneurship researchers should whenever possible move away from conducting cross-section surveys and collect panel data instead. Of course, data limitations mean that this is not always possible. And panels can suffer from problems of their own, including sample attrition, whereby individuals drop out of the panel non-randomly, e.g. because they move away from the location where the survey takes place, or because they close their businesses. Parker and Belghitar (2006) discuss this issue in the context of the Panel Study of Entrepreneurial Dynamics (PSED) panel design and propose Heckman selectivity corrections to deal with possible bias caused by sample attrition.

100

Selection

The use of binary dependent variables gives rise to another problem with fixed effects, since this specification necessarily eliminates from the sample all individuals who do not change occupational status over the sample period – including the numerous interesting cases of ‘survivors’ who remain in business. This is because individuals who never change occupation have fixed zit = zi observations which are perfectly explained by their fixed effect αi . One solution is to use random-effects models instead (for example, see Hochguertel, 2005a). 3.6

Entrepreneurial duration models

Binary choice and hazard models are the most widely used techniques for quantifying the effects of individual- and firm-specific characteristics on the duration of spells in entrepreneurship. Binary choice models Consider again the binary choice models described in section 3.2. These models condition a binary variable zi on a vector of explanatory variables, Wi , where i indexes an individual observation. Suppose one now defines zi as equal to one if individual i remains in entrepreneurship after some time interval has elapsed, and zero if i has left entrepreneurship by this time. Alternatively, in applications where i indexes a particular venture from a sample of ventures, zi can be defined as equal to one if the venture is still in business and zero if the venture has exited the market. In both cases the marginal effects of an explanatory variable on the probability of survival are still calculated using (3.32) in note 2. In a variant of the binary model, Cooper et al. (1992) proposed the following ‘threshold’ specification to link financial performance of an entrepreneur with their exit or continuation decision. Define a variable z to take the value 1 for continuation if π ≥ π, where π is a minimum acceptable profit threshold in entrepreneurship. Hence z = 0 if π < π. Write π = β1
X + u1 and π = β2
X + u2 . A two-step procedure first estimates the simple probit model z = β3
X + u3 , where β3 = (β1 − β2 )/σ , where σ is the standard deviation of uˆ 3 . Second, estimate β1 from the π model, and infer β2 and its standard errors from the β1 and β3 estimates. This approach, which uncovers the determinants of both performance and thresholds, is valid under the assumption that cov(u1 , u2 ) = 0. Although they are widely used to model exit outcomes, binary choice models suffer from a serious drawback. They effectively assume that z = 1 cases, whether defined as entrepreneurs or ventures, survive when, in fact, they have only been observed not yet to fail. Statisticians refer to this as ‘right censoring’, a feature which should be incorporated explicitly in the likelihood function. By not doing so, binary models when applied to duration analysis are prone to bias. A superior approach is to use hazard models. Hazard models As well as addressing right-censoring problems, hazard

models (also referred to as ‘survival’ and ‘event history’ models) provide a direct way

Empirical methods in entrepreneurship research

101

of identifying the factors that determine how long (rather than whether) individuals remain in entrepreneurship, or how long ventures remain in the market. More generally, these methods can also be used to model business entry as well as other events in the entrepreneurial process (see e.g. Sørensen, 2007). For expositional clarity, consider the case of an individual remaining in entrepreneurship. At each discrete point in time t there is a probability (or hazard) that individual i, who has been observed in entrepreneurship for ai periods up to t, leaves entrepreneurship. The Cox proportional hazard model is described by

Hi (t) = H0 (t) exp[β
Xi (t)],

(3.21)

where H0 (t) is the so-called ‘baseline hazard’ at t; Xi is a vector of characteristics for individual i; and β is a vector of parameters to be estimated. This is called a ‘single risk’ hazard model, because there is only a single risk: that of leaving entrepreneurship. The single-risk model can be estimated without placing restrictions on the form of the baseline hazard in (3.21).10 Many applications report exponentiated values of the coefficients, called ‘hazard ratios’. Ratios greater (resp. less) than 1 indicate positive (resp. negative) impacts on duration: see Beckman, Burton and O’Reilly (2007) for an example in the context of the hazard of founding teams accessing venture capital finance. An ‘event’may take several forms. For example, individuals may leave entrepreneurship for either paid employment or unemployment. We might be interested to discover, for example, whether entrepreneurs who exit into unemployment have different survival characteristics from those who exit into paid employment. If so, a ‘competing risks’ model is needed.11 As an illustration of the usefulness of the distinction between single- and competing-risk formulations, Taylor (1999) distinguished between voluntary exit from self-employment into paid employment and involuntary exit in the form of bankruptcy. This distinction turned out to be an important one. For example, while limited personal wealth was associated with higher bankruptcy rates among British males, it had no effect on voluntary exits. Hazard models have been widely used in applied entrepreneurship research, especially in analyses of the survival of new ventures (see chapter 14). They continue to be enriched and extended. For example, a useful recent development incorporates fixed effects to control for unobserved heterogeneity (Strotmann, 2007). 3.7 Appendices 3.7.1 The reduced form of the two-equation simultaneous equation model Write W1 ≡ [, 1 ] and W2 ≡ [, 2 ], with parameter vectors β1 ≡ [β1ξ , β1⊥ ] and β2 ≡ [β2ξ , β2⊥ ]. Then the reduced form for the model is ∗ z1i =


i (β1ξ + γ1 β2ξ ) + 
1i β1⊥ + 
2i (γ1 β2⊥ ) ∗ + v1i 1 − γ1 γ2

∗ = Vi
1 + v1i

(3.22) (3.23)

102

Selection

∗ z2i =


i (β2ξ + γ2 β1ξ ) + 
2i β2⊥ + 
1i (γ2 β1⊥ ) ∗ + v2i 1 − γ1 γ2

∗ = Vi
2 + v2i

(3.24) (3.25)

where ∗ v1i :=

v1i + γ1 v2i 1 − γ1 γ2

and

∗ v2i :=

v2i + γ2 v1i 1 − γ1 γ2

(3.26)

are reduced form error terms and Vi := [1, i , 1i , 2i ]. 3.7.2

Fraser and Greene’s (2006) heteroscedastic probit model Suppose that an entrepreneur’s income at t is given by ytE = θ + t , where θ is the entrepreneur’s true but unknown ability, and t ∼ IIN (0, σ 2 ) is random noise. Entrepreneurs do not observe θ and t separately, but only their sum, which they use to learn about θ. They do so by Bayesian updating. An entrepreneur’s initial idiosyncratic prior belief about θ when they enter at t = 0 is assumed to be N (µ0 , σ02 ). With a non-pecuniary benefit in entrepreneurship of ψ, individuals choose to participate in entrepreneurship if

Et−1 (ytE + ψ − yP ) > 0,

(3.27)

where yP is certain income in paid employment, assumed for simplicity to be constant over time. To analyse occupational choice, one must derive the posterior distribution of ytE . Under Bayesian updating, this turns out to be N [Et−1 (θ ), Vt−1 (θ )], where Et−1 (θ) = W (t − 1)Et−2 (θ) + [1 − W (t − 1)]θˆ (t − 1) Vt−1 (θ) = σ02

t−1

W (i),

(3.28) (3.29)

i=0

where θˆ (t − 1) := (t − 1)−1  W (i) :=

t−1  (θ + i ) i=1

i + 2  i−1 σ02 h=0 W (h) σ 1

−1

1  i−1 2

σ0

h=0 W (h)

,

i≥1

W (0) = 1 Since ytE = θ + t , we can write Et−1 (ytE + ψ − yP ) = ytE + ψ − yP + B(t − 1), where B(t − 1) = [Et−1 (θ) − θ] is the entrepreneur’s optimistic bias. Clearly, learning implies plim B(t) = 0, i.e. entrepreneurs’ bias disappears with experience. This is an empirically testable prediction of the model.

Empirical methods in entrepreneurship research

103

Finally, since the posterior distribution is normal, and its mean (3.28) and variance (3.29) are known, (3.27) implies that the latent decision to participate in entrepreneurship is   E + ψ − y P + B(t − 1) y z∗ =  t . (3.30)  1/2 σ0 t−1 i=0 W (i) This is like a probit expression, but with the difference that it is heteroscedastic (since the denominator of (3.30) is time-varying). Also, because limt→∞ W (t) = 0, the variance is decreasing in the entrepreneur’s experience. Thus Bayesian learning implies a heteroscedastic binary probit model of occupational choice, in which the variance is conditioned (negatively) on the entrepreneur’s experience in entrepreneurship. The likelihood function for this model is  zi  1−zi   n

β
Xi β
Xi L=  , 1 −  exp{γ
Mi } exp{γ
Mi } i=1

where X are the variables deemed to affect the mean, and M are the variables deemed to affect the variance. Notes 1. ML estimation of probit maximises the likelihood function L=

n

1−zi  (β
Wi /σ )zi 1 − (β
Wi /σ ) .

(3.31)

i=1

Non-linear methods are needed to maximise (3.31) to estimate the β parameters (up to a scalar transformation since σ is unknown). It is standard to normalise σ 2 to unity without loss of generality. 2. The effects of changes in the kth explanatory variable of Wi , i.e. Wik , on the probability of entrepreneurship are given by ⎧ for the linear probability model ⎨ βk ∂ Pr(zi = 1) βk Pr(zi = 1)[1 − Pr(zi = 1)] for the logit model (3.32) := ⎩ ∂Wik βk φ(β
Wi ) for the probit model where φ(·) is the density function of the standard normal distribution. In practice, software packages usually evaluate these effects at sample means of the variables. 3. The probit model with relative incomes is sometimes referred to in the literature as the ‘structural probit model’ (including in the previous version of this book). However, it is not actually a structural model in the strict econometric sense mentioned above, so this terminology should probably be discouraged in future. 4. As before, define zi as an indicator variable, equal to 1 if individual i is an entrepreneur, and 0 otherwise. Evidently ∗ , z ∗ ) ≥ 0] = (β
W , β
W ; ρ), Pr(zi = 1) = Pr[min(z1i 2i 1 1i 2 2i

104

Selection and analogous to (3.31), the likelihood function is L=

n

 1−zi (β1
W1i , β2
W2i ; ρ)zi 1 − (β1
W1i , β2
W2i ; ρ) .

i=1

5. In terms of the first part of the chapter appendix, researchers should be careful when adopting the conventional approach of interpreting the coefficients in single-equation estimations of (3.23) and (3.25) as structural rather than reduced form parameters. To see how misleading this can be, suppose that γ1 γ2 > 1 (an empirically relevant case: Parker, 2008a). Because the denominators of both 1 and 2 are 1 − γ1 γ2 (see (3.22) and (3.24) above), the structural and reduced form parameters can take systematically opposite signs, leading to precisely the wrong (structural) interpretations. 6. This involves estimating the reduced forms (3.23) and (3.25) by single-equation probit ML at the first stage, and substituting the predicted latent values obtained from these estimates in place ∗ and z ∗ in (3.17) and (3.18). Estimation of the latter by ML generates consistent estimates of z2i 1i of all of the parameters, but requires a correction to the parameter variance-covariance matrix owing to the use of ‘generated regressors’ (see Maddala, 1983, pp. 246–7). 7. A different econometric model is needed to analyse the proportion of total available work hours ∗% an individual spends in entrepreneurship, denoted by h% E . Let hE be a latent variable relating the desired proportion of work hours in entrepreneurship to the regressors X . Consider the model ⎧ ⎪ ⎨ 1 h% = h∗% iE iE ⎪ ⎩ 0

if if if

h∗% iE ≥ 1 0 < h∗% iE < 1 h∗% ≤ 0 iE

(3.33)


h∗% iE = β Xi + ui .

This model can be estimated by double-limit tobit maximum likelihood (see Vijverberg, 1986, for details). However, a limitation of this model is its omission of relative occupational returns from the set of explanatory variables, X . Incorporating this extension into (3.33) complicates the estimation substantially. 8. Parker (1996), Cowling and Mitchell (1997), Parker and Robson (2004) and Bruce and Mohsin (2006). 9. The likelihood function for the probit model with fixed effects and state dependence is L=

n T



1−zit [(β
Wit + γ zit−1 + αi )/σ ]zit 1 − [(β
Wit + γ zit−1 + αi )/σ ] , (3.34)

i=1 t=1

where the αi are the individual fixed-effects; γ is a parameter measuring the importance of state dependence; and Wit are individual- and time-varying control variables. 10. The probability of an individual i’s spell being completed by time t + 1 given that i was still an entrepreneur at t is Li (t) = Pr[ai < t + 1|ai ≥ t] = F[γ (t) + β
Xi (t)], where F[·] is the cumulative distribution function of the Extreme Value distribution; and γ (t) is a set of dummy variables, one for each t, which captures time dependence. To estimate the parameters γ (t) and β, let di be the observed duration of i in entrepreneurship. This is either time completed or time censored (by the end of the sample): define ζi = 1 if the time is completed

Empirical methods in entrepreneurship research and ζi = 0 if it is censored. Then the log-likelihood function is ⎫ ⎧ di n ⎨ d i −1 ⎬   ln L = ln[1 − Li (t)] . ln[1 − Li (t)] + ζi ln Li (di ) + (1 − ζi ) ζi ⎭ ⎩ i=1

t=1

105

(3.35)

t=1

Equation (3.35) is maximised with respect to β and the γ s to obtain ML estimates. The method is implemented in many standard software packages. 11. Such a model posits a separate hazard function for each of the destinations, whose log likelihoods are given by (3.35) where ζi = 1 now denotes exit into the given destination, and ζi = 0 applies for other outcomes. The sum of the log likelihoods over all possible destinations gives the total log likelihood to be maximised.

4

Evidence about the determinants of entrepreneurship

Chapter 2 discussed several theories of entrepreneurship, the role of entrepreneurs and the factors influencing individuals to participate in entrepreneurship. All of these theories treat the decision to participate in entrepreneurship as an occupational choice. The present chapter considers a wide variety of empirical variables bearing on the decision to choose entrepreneurship. The focus of the chapter can be summarised by the following equation:   z ∗ = z π − w, Xhuc , Xsoc , Xrisk , Xpsy , Xdem , Xind , Xmac , Xemp ,

(4.1)

where z ∗ is the latent (unobserved) preference to be an entrepreneur, whose observed counterpart is usually a binary variable taking the value 1 for individuals who are entrepreneurs, and 0 for those who are not (see the previous chapter). The arguments of this equation are treated sequentially in the course of the present chapter. Thus the term π − w denotes (expected) profits in entrepreneurship relative to wage income in paid employment, and captures the relative financial return to entrepreneurship: one expects a positive effect on the propensity for entrepreneurship, z ∗ . The vector Xhuc includes variables measuring human capital; Xsoc measures social capital; and Xrisk measures risk factors, such as individuals’risk aversion or the degree of market risk. The vectors Xpsy and Xdem contain psychological and demographic characteristics, respectively; Xind and Xmac include industry-specific and macroeconomic factors; and Xemp includes characteristics of employers, relevant to employees contemplating a switch into entrepreneurship. Chapter 3 discussed the appropriate multivariate empirical methods for estimating a binary equation like (4.1), together with various extensions. The present chapter presents a comprehensive overview of results based on these estimation methods. Univariate differences between entrepreneurs and non-entrepreneurs, e.g. t tests of differences in location, are still sometimes cited and used as a basis for inference in some non-economics research (especially psychological factors: see section 4.4 below). But, as noted in the previous chapter, and repeated here for emphasis, univariate tests are prone to problems of misleading inference when elements of (4.1) are correlated among themselves (see Greene, 2003). Hence this chapter will devote little attention to results based on them. 106

Evidence about the determinants of entrepreneurship

107

Equations like (4.1), or more commonly special cases of it in which only some of the explanatory variables listed above are present, are used to help answer two fundamental questions in entrepreneurship research: (i) Who becomes an entrepreneur and why? (ii) What are the influences of particular personal characteristics and environmental factors? The promise of (4.1) is that it can examine different theories of entrepreneurship from a variety of disciplines, including economics, sociology, psychology, and business and management. This assists research efforts to build the new multi-disciplinary field called ‘entrepreneurship’. To this end, section 4.1 first reviews the evidence about the impact of pecuniary and non-pecuniary incentives to be an entrepreneur. Sections 4.2 and 4.3 explore the roles of human and social capital, respectively. Section 4.4 focuses on risk attitudes and conditions, Xrisk , and psychological characteristics, Xpsy , including over-optimism and other cognitive biases. Section 4.5 treats demographic and industry characteristics, while section 4.6 summarises evidence about the broader macroeconomic factors that affect entrepreneurship, including economic development, technological change, unemployment and regional factors. Up until this point, ‘entrepreneurs’ are taken to be those people who have successfully started a venture. Section 4.7 discusses whether any of the empirical findings differ for those individuals actively engaged in the process of start-up, i.e. ‘nascent entrepreneurs’. Section 4.8 then explores the role of spin-offs and employer characteristics, while section 4.9 supplies a brief conclusion. For convenient reference, the main empirical results of the chapter are summarised in Table 4.1. Results relating to psychological factors (other than risk attitudes) have been omitted because their diversity makes them difficult to summarise in this way. I refer only to published research results based on multivariate analysis, excluding studies citing simple bivariate correlations, since these are vulnerable to the most severe form of omitted variable bias. Positive and negative entries in Table 4.1 refer to measured effects on entrepreneurship that are significantly positive or negative; the zero entries refer to effects that are too imprecisely estimated to reach standard levels of statistical significance (usually 5 per cent). I now discuss the various determinants of entrepreneurship in detail. 4.1

Pecuniary and non-pecuniary incentives

The first argument of (4.1) is relative earnings. The idea here is that individuals become entrepreneurs in order to make more money. However, money is not the only or even necessarily the most important incentive for entrepreneurs. Schumpeter in particular highlighted the importance of non-economic motivations, writing of the ‘will to found a private kingdom. . .to conquer: the impulse to fight, to prove oneself superior to others, to succeed for the sake, not of the fruits of success, but of success itself. . . Finally there is the joy of creating, of getting things done, or simply of exercising one’s energy and ingenuity’ (1934, pp. 93–4). Further to this point, consider Table 4.2, which summarises reasons for being self-employed cited by respondents of the UK’s spring 2000 LFS (Labour Force

108

Selection

Table 4.1. Summary of determinants of entrepreneurship Explanatory variable 1. Income differential 2. Age 3. Experience 4. Education 5. Risk aversion 6. Married / working spouse 7. Number of children 8. Ill health / disability 9. Entrepreneur parent 10. Technological progress 11. Unemployment Cross-section Time series 12. Urban location 13. Immigrationa 14. Interest ratesb 15. Personal wealthc 16. Personal income tax ratesd

No. +

No. −

No. 0

8 83 24 69 0 52 16 5 40 4

2 6 1 21 11 9 2 4 2 4

4 14 2 27 3 8 3 6 2 2

22 33 7 5 1 40 12

14 5 7 1 9 2 5

18 2 4 0 3 4 1

Notes: +, − and 0 denote significantly positive, significantly negative and zero (insignificant) coefficients, respectively. Only multivariate studies (i.e. those including controls for other explanatory variables) are included; descriptive studies are excluded. For row 11, panel studies with large N and small T are classified as cross-section; those with large T and small N are classified as time-series. Sources:1 a Based on literature discussed in chapter 5. b Based on literature discussed in chapter 4, section 4.6. c Based on literature discussed in chapter 9. d Based on literature discussed in chapter 17.

Survey). These responses highlight the importance of non-pecuniary factors, especially independence, which is the single most important reason cited by the respondents (see below). The next most important reason cited in Table 4.2 is the nature of the occupation, perhaps reflecting the fact that self-employment is the chief or only mode of employment in some locations or occupations (e.g. forestry and construction). Both men and women give similar responses although women stress independence a little less than men, and family commitments substantially more (see also Hakim, 1989a). However, a caveat to interpreting self-declared motivations is that established entrepreneurs apparently attach much less importance to financial motivations than when they were in the process of starting their ventures (Cassar, 2007). This may be indicative of self-justifying response bias when ventures do not turn out as well as expected. So although researchers sometimes relate transitions into entrepreneurship to self-reported motivations

Evidence about the determinants of entrepreneurship

109

Table 4.2. Reasons given for becoming self-employed in the UK (per cent) Reason To be independent Wanted more money Better conditions of work Family commitments Capital, space, equipment opportunities Saw the demand Joined the family business Nature of occupation No jobs available locally Made redundant Other reasons No reason given No. valid responses (’000s) a

All

Men

Women

31 13 5 7

33 15 6 2

25 7 3 21

12 8 6 22 3 9 15 3

12 9 6 21 3 11 14 4

11 8 7 23 2 3 18 3

2, 960

2, 156

804

Note: Columns do not sum to 100 per cent because respondents can give up to four reasons. a Imputed percentages based on all those who gave a valid response to the ‘reasons for becoming self-employed’ questions. Source: LFS (2000).

(e.g. Henley, 2007) response bias may make them a misleading guide to actual behaviour. 4.1.1 Pecuniary incentives: relative earnings The most widely used tool for estimating the effects of relative earnings π −w on participation in entrepreneurship is the binary probit model with relative incomes described in chapter 3. Recall from (3.13) of that chapter that if α, ˆ the estimated coefficient on the earnings differential term in the probit model, is positive and significant, a relative earnings advantage in entrepreneurship increases the likelihood of participation in entrepreneurship. It is important to point out that methodologically this approach is superior to analysing the effects of just employee wages on the probability of being an entrepreneur.2 Omitting prospective entrepreneurial profits gives an incomplete account of the opportunity cost affecting individual behaviour, and should be avoided. Taking self-employment as a working definition of entrepreneurship, several British researchers have estimated (3.13). Most have reported positive α estimates. Estimates include 0.37 (Rees and Shah, 1986); 0.04 (Dolton and Makepeace, 1990); 0.09–0.13 (Clark and Drinkwater, 2000); and 0.60 (Taylor, 1996). But only the last two estimates were statistically significant. Subsequent work also provides weak support for the role of financial incentives: for example, Fraser and Greene (2006) reported positive and

110

Selection

significant effects using data from the 1980s but not from the 1990s. Parker (2003b) obtained mixed results (αˆ took varying signs) using several British data sets from the 1990s; his estimate of α was positive but insignificant for switchers between paid employment and self-employment. Data from other countries also paint a mixed picture: this is reflected in the varied findings documented in the first row of Table 4.1. In contrast, aggregate studies of new business starts or rates of entrepreneurship tend to find fairly uniformly positive effects from average business profits relative to average wages.3 However, aggregate data are ill suited to detecting the importance of individual incentives, since average entrepreneurial income is prone to measurement error and almost certainly endogenous, leading to upward-biased parameter estimates (Parker, Belghitar and Barmby, 2005). These considerations, together with the mixed evidence from individual-level data about the sign of the relationship between relative incomes and entrepreneurial choice, pose something of a puzzle, since they suggest that entrepreneurs do not respond robustly to pecuniary incentives. Perhaps previous researchers have not successfully isolated variables that affect earnings in the two sectors but not occupational choice itself – as is required to identify the econometric model (see chapter 3). And the poor quality of data used in many of these studies, especially relating to reported incomes, might also be partly responsible for the inconclusive results. Alternatively, these results may simply be telling us that pecuniary rewards are not the primary motive for being an entrepreneur (Amit et al., 2000). Participation in entrepreneurship might instead be motivated by lifestyle considerations, for instance as a means of being one’s own boss. These non-pecuniary factors are discussed further below. Another possibility is that entrepreneurs suffer from unrealistic optimism and continue to earn low incomes because they wrongly anticipate high future profits ‘which never come’. While there might be some truth in each of these conjectures, there is so much economic evidence that individuals adjust their behaviour in response to changes in relative prices that it would be puzzling if the same calculus ceased to apply entirely in the realm of entrepreneurship as an occupational choice. Indeed, some economic historians argue that American entrepreneurs have historically been very responsive to incentives, directing their attention to profitable innovations and the satiation of demand (Khan and Sokoloff, 1993). And there is some suggestive evidence that ‘postmaterialism’, i.e. the possession of non-materialistic life goals, is negatively related to entrepreneurship (Uhlaner and Thurik, 2007) – which might be taken to imply that entrepreneurship is largely about the pursuit of material gain. Hence we conclude that more careful research is needed to shed further light on entrepreneurs’ responsiveness to pecuniary incentives.

4.1.2

Desire for independence and job satisfaction It is commonly asserted that an attractive feature of entrepreneurship is independence in the workplace, something that is variously referred to as a desire for autonomy or

Evidence about the determinants of entrepreneurship

111

‘being one’s own boss’. This idea can be traced back to Knight (1921), and might explain why individuals remain in entrepreneurship despite the possibility that they earn less there and receive a lower risk-adjusted return on their assets (see chapters 13 and 9, respectively). Below, I review evidence about the desire for independence in entrepreneurship, before asking whether it feeds through into greater happiness among entrepreneurs relative to employees, in terms of both job satisfaction and contentment with work–life balance. Based on an analysis of 1991 BHPS data, Taylor (1996) found that fewer entrepreneurs than paid employees regarded pay and security as important aspects of their job. The proportions were 37 and 32 per cent compared with 48 and 57 per cent, respectively. However, greater proportions of entrepreneurs reported that initiative (51 per cent) and the enjoyment of work itself (57 per cent) were important job aspects, compared with 21 and 41 per cent of employees, respectively. These non-pecuniary factors were significantly associated with being self-employed, even after controlling for observable personal characteristics (see also Burke et al., 2000; and Hundley, 2001b).4 Recall from Table 4.2 that the self-employed claim that independence is a major reason for their occupational choice. Yet it is not clear that entrepreneurship offers much real independence, especially for those working long hours (see chapter 12) or in one of the ‘grey areas’ between paid employment and self-employment (see chapter 1). This might explain why sole proprietors, many of whom enjoy limited autonomy (especially the ‘dependent self-employed’) report lower levels of job satisfaction than entrepreneurs who hire employees – although both groups are more satisfied on average than employees are.5 Moreover, about one-third of GEM respondents in developed countries claim they are ‘necessity’ entrepreneurs, i.e. they become entrepreneurs because they have few other choices. Of course, these responses are all based on declared rather than revealed preferences, so they should be treated with commensurate caution. Nevertheless, the available evidence strongly supports the notion that entrepreneurs regard themselves as more satisfied with their jobs than non-entrepreneurs. For example, Blanchflower and Oswald (1998) reported that approximately 46 per cent of the self-employed claimed they were ‘very satisfied’ with their job, compared with only 29 per cent for employees. Because the self-employed tend to be older on average and are likelier to be married (both factors being associated with happiness: Hundley, 2001b; Benz and Frey, 2004), it is necessary to take account of these other factors when making cross-group comparisons. Several studies which control for job and personal characteristics, including income gained and hours worked, confirm the finding that entrepreneurs are more satisfied with their jobs on average than employees are.6 It is noteworthy that studies which control for socio-demographic and work factors find that higher incomes in entrepreneurship do not explain greater levels of job satisfaction. Nor is greater job satisfaction in entrepreneurship explained by panel-data fixed effects designed to control for unobserved heterogeneity and possible self-selection arising from it.7 Benz and Frey (2008) take this as evidence of ‘procedural utility’: people value not only outcomes, but also the processes leading to outcomes.

112

Selection

It appears that greater autonomy lies at the heart of entrepreneurs’ higher reported levels of job satisfaction. In a clever ‘natural experiment’, Benz and Frey (2004, 2008) measured self-reported job satisfaction of East Germans before and after the fall of the Berlin Wall in 1989, to explore whether entrepreneurship makes people happier or whether just intrinsically happier people tend to become entrepreneurs. Prior to the fall of the Berlin Wall (which can be regarded as a genuinely exogenous shock), East Germans were prohibited from becoming self-employed. Benz and Frey (2004, 2008) found that individuals who became self-employed after 1989 became significantly more satisfied than East Germans who remained as employees. This suggests that causality flows from entrepreneurship to happiness rather than the other way round. They attributed this to the greater autonomy that is inherent in self-employment. While both employees and entrepreneurs appear to value autonomy and freedom from subordination in a hierarchy, only entrepreneurs can meaningfully practise it.8 Benz and Frey (2004) left aside questions of whether entrepreneurs enjoy greater life satisfaction. Arguably, the flexibility of entrepreneurship makes it especially conducive to balancing work and family role responsibilities, leading to enhanced psychological well-being (Cromie, 1987; Loscocco, 1997). On the other hand, entrepreneurs work longer hours on average than employees do (chapter 12), while bearing direct responsibility for the success and survival of their businesses. That may result in greater work–family conflict for entrepreneurs than for employees. Evidence from the UK, USA and Canada shows that while entrepreneurs are significantly more satisfied on average with their jobs, especially with their flexibility and autonomy at work, they are less satisfied on average with their work hours than employees are, and experience higher levels of work–family conflict and lower levels of satisfaction with their family lives than employees do.9 These drawbacks to entrepreneurship appear to be more pronounced for men than for women. According to Finnish time-use data, the self-employed take significantly less time off on average for sickness or holidays; work under greater time pressure; and supply more work hours in the evenings than employees do (Hyytinen and Ruuskanen, 2007). Although entrepreneurs can and do interrupt their workday somewhat more and spend less time in their primary work location than employees, this might not compensate for these other drawbacks of business ownership. The issue of stress and anxiety in entrepreneurship is receiving growing attention from researchers. It is sometimes claimed that entrepreneurs are likely to suffer from greater average levels of stress, worry and burnout than employees (see e.g. Blanchflower, 2004; Jamal, 2007). On the other hand, entrepreneurs tend to be more autonomous and older on average than employees, and both of these factors are associated with lower levels of (self-reported) stress (Bluedorn and Martin, 2008). Perhaps reflecting these offsetting forces, group differences in satisfaction, stress and work– family conflict levels cease to be significantly different once researchers introduce controls for work and demographic characteristics (Prottas and Thompson, 2006). Note too that while entrepreneurship might not be conducive to reducing work–family conflicts for individuals who prefer to segment these two parts of their lives, it can be ideal for those who prefer to integrate, or ‘blur’, the two.

Evidence about the determinants of entrepreneurship

113

To conclude, it appears that for some people, the promise of autonomy can be a more compelling incentive to enter entrepreneurship than financial returns. Autonomy in entrepreneurship yields job satisfaction, but the long hours associated with entrepreneurship can create work–life conflicts which mitigate the accompanying increase in life satisfaction. Just as individuals differ in the way they trade off autonomy with financial returns, so they also differ in their tolerance of stress and the ways they seek to achieve a work–life balance. Only a minority of people seem to be sufficiently fond of autonomy, dedicated enough to work long hours, and able to cope with family conflicts and stress emanating from both work and home environments to become entrepreneurs. Perhaps this explains why so many more individuals claim to be ‘latent’ entrepreneurs than are actual entrepreneurs. 4.2

Human capital

This section analyses three main elements of the human capital vector Xhuc : age, experience and formal education. 4.2.1 Age One might expect older people to be likelier to become entrepreneurs, for the following reasons:

1. The human and physical capital requirements of entrepreneurship are often unavailable to younger workers. For example, older people are more likely to have received inheritances and to have accumulated capital which can be used to overcome borrowing constraints and set up a business (see chapter 9). There might also exist a particular type of human capital which is productive both in managing and in working for others, and which can be acquired most effectively by working initially as an employee. 2. Older individuals might choose entrepreneurship to avoid mandatory retirement provisions sometimes found in paid employment. 3. Older people have had time to build better social and business networks, and to have identified valuable opportunities in entrepreneurship, possibly through learning about the business environment. 4. As their own masters, entrepreneurs often possess greater control over the amount and pace of their work, making it sometimes better suited to older people who have lost their physical stamina, or for workers in poor health or with skills which are obsolete in paid employment. On the other hand (the retirement issue apart – discussed separately in chapter 12), it seems likely that beyond a certain age, entrepreneurship will become less attractive to individuals. Reasons include: 1. Older people are more risk-averse than the young; may have tastes that shift away from enterprise as they age (Lévesque et al., 2002); and are less capable of working the long hours undertaken by entrepreneurs (see chapter 12).

114

Selection

2. The returns to information gained about suitable job matches diminish with age. Thus, Miller’s (1984) ‘job-shopping’ theory predicts that workers try riskier occupations (like entrepreneurship) when they are younger, since these occupations provide the richest information about workers’ personal job-matching opportunities. 3. Older people’s human capital has a lower discounted value and so provides less of a hedge against risk, so making older individuals less willing to enter a risky occupation like entrepreneurship (Polkovnichenko, 2003). Conversely, if entrepreneurship is risky and requires payment of a sunk entry cost, younger entrepreneurs have more time over which to benefit from the higher returns and over which to amortise the entry cost (Hintermaier and Steinberger, 2005). Taken together, these arguments can be interpreted to suggest that individuals are increasingly likely to become entrepreneurs as they age, up to a certain point, after which the probability of becoming an entrepreneur declines with age (Lévesque and Minniti, 2006). In support of this argument, descriptive studies tend to find that entrepreneurship is concentrated among individuals in mid-career, i.e. between thirty-five and forty-four years of age.10 Recognising this, many empirical studies enter age as an explanatory variable in a quadratic specification, i.e. in levels and squares. The above arguments would lead one to expect the level term to enter a binary choice model of entrepreneurship with a positive coefficient and the squared term to enter with a negative coefficient. Age and experience are not synonymous, yet a common practice (often dictated by data limitations) is to measure ‘experience’ as current age minus school-leaving age. This measure is imperfect because it takes no account of breaks from labour force participation in individuals’ work histories. This may be a particularly salient consideration when analysing female entrepreneurship. It also runs into the problem of conflating cohort and experience effects. For instance, a secular increase in the rate of entrepreneurship over time shows up as a decline in the average age of entrepreneurs, irrespective of any experience effects.11 To separate cohort effects from genuine experience effects, either longitudinal data or accurate measures of years of actual work experience are required (see below). Despite these caveats, most econometric investigations have explored the effects of age on participation in entrepreneurship using cross-section data. Entrepreneurship is most commonly defined in these studies as self-employment or business ownership. Consistent with the theoretical arguments above, by far the most common empirical finding, obtained for a range of countries and time periods, is of a quadratic pattern, with the probability of being or becoming an entrepreneur increasing up to a maximum at some age in the forties or early fifties, before declining thereafter. For simplicity, these findings are reported as positive entries in the second row of Table 4.1. Digging deeper into the data, Evans and Leighton (1989b) report that rates of entry into self-employment do not increase as Americans age, but Americans remain in selfemployment for longer spells as they age. Uusitalo (2001) reports a similar pattern for

Evidence about the determinants of entrepreneurship

115

Finland, with a constant entry rate of 1.5 per cent up to age forty, declining thereafter to a constant 1 per cent until retirement. Age may have different effects on the willingness and opportunity to become an entrepreneur. For example, using the bivariate probit model described in chapter 3, van Praag and van Ophem (1995) found that the opportunity to become self-employed is significantly higher for older than for younger Americans. However, older workers are significantly less willing to become self-employed than younger workers (see also Blanchflower et al., 2001). Other work, based on the binary model with unobserved thresholds described in chapter 3, suggests that older entrepreneurs have worse outside options in the labour market, and so are more likely to remain in entrepreneurship once they are there (Cooper et al., 1992). This is of course consistent with Evans and Leighton’s (1989b) empirical findings.

4.2.2 Experience Another component of the human capital vector Xhuc is experience. As intimated earlier, experience captures more informatively than age does the productive impact of post-school training and skill acquisition. Greater experience might promote entrepreneurship for several reasons, some of which are related to age-related arguments reviewed in the previous subsection. But experience also explicitly embodies learning, for example about business opportunities, or how enterprises work in practice. Experience includes training for skills needed to exploit opportunities, such as selling, negotiating, leading, planning, decision-making, problem solving, organising and communicating (Shane, 2003, p. 75). Learning also generates information which reduces uncertainty about the value of exploiting entrepreneurial opportunities (Jovanovic, 1982; Parker, 2007c). The evidence, summarised in the third row of Table 4.1, points to a consistent positive relationship between experience (defined quite broadly) and entrepreneurship. Shane (2003) distinguishes between five types of experience: general business experience; functional experience (e.g. in marketing, product development and management); industry experience; start-up experience; and vicarious experience (e.g. through observing parents, friends and associates in business). One can also distinguish between different types of labour market experience, e.g. the importance of previous experience in entrepreneurship versus experience in paid employment or experience as a manager. The results here are scant but interesting. Evans and Leighton (1989a) estimated that previous self-employment experience had a positive and significant impact on the probability of white male Americans entering self-employment, whereas previous paid employment experience had no effect.12 This result is consistent with Jovanovic’s (1982) theory that entrepreneurs learn about their abilities over time, which they can only do from having engaged in entrepreneurship (see chapter 11). Other forms of business experience also promote entrepreneurship. Illegal experience of entrepreneurship has been associated with legitimate self-employment, both in the USA (experience from drug dealing: Fairlie, 2002) and in Eastern Europe (experience from owning

116

Selection

a proscribed business before the transition from socialism: Earle and Sakova, 2000; Aidis and van Praag, 2007). It is also noteworthy that individuals with greater past entrepreneurial experience and greater industry experience are more likely to start their own independent venture than to purchase a franchise (Kaufmann, 1999; Williams, 1999). That might be because franchisors furnish franchisees with industry knowledge and sell business models that embody pre-existing business expertise, so individuals with less of both types of experience are likelier to opt for the franchise route. General work experience can promote entrepreneurship if entrepreneurs start businesses related to their former occupations. Entrepreneurs often obtain ideas for new ventures from their previous jobs. For example, 71 per cent of Inc. 500 founders reported that they had replicated or modified an idea they had identified in previous employment, compared with only 4 per cent who claimed they had identified their idea through systematic search (Bhide, 2000). There are good reasons for this: Dunkelberg et al. (1987) observed that the ventures of founders who identified new venture ideas from prior jobs enjoyed higher than average rates of growth. This is suggestive of experience which takes the form of valuable direct knowledge transfers. Diversity of experience might also facilitate entrepreneurship, by conferring a broader set of skills and experience or by exposing potential entrepreneurs to a greater number of novel venture ideas. In which case, one would expect that individuals with more different previous jobs, past employers and job changes are more likely to be entrepreneurs.13 As noted in chapter 2, this has been formalised as the ‘jack-of-all-trades’ (JAT) hypothesis (Lazear, 2005). The hypothesis has been tested directly, by measuring breadth of experience in terms of the number of different work roles in previous jobs, or the number of fields of prior experience (Lazear, 2005; Wagner, 2003a, 2006a). These variables turn out to be significantly and positively associated with entrepreneurship status. Similar results have been recorded in Russia and China (Djankov et al., 2002). And there is evidence that top management teams with more diverse experience enjoy greater success in accessing venture capital and achieving an IPO (Beckman et al., 2007). However, Silva (2007) claims that these JAT results are not robust to the inclusion of panel-data fixed effects, suggesting that unobserved tastes and capabilities rather than diverse experience may simultaneously shape skill-accumulation strategies and choices to enter entrepreneurship. Most entrepreneurs who leave entrepreneurship transition into paid employment rather than unemployment (Cowling et al., 2004). So we can ask whether experience in entrepreneurship has long-term effects on what an individual can earn if they later switch to paid employment. On the one hand, the effect might be positive if entrepreneurship experience augments general human capital; on the other hand, it might be negative if entrepreneurship causes job-specific skills to rust and signals career instability to employers. Empirical investigations of this issue yield mixed and inconclusive results.14 Previous entrepreneurship experience has stronger effects on propensities to try entrepreneurship again in the future, where it is associated with superior sales, profitability and growth performance on average (Westhead and Wright,

Evidence about the determinants of entrepreneurship

117

1999; Ucbasaran et al., 2006). These particular findings are suggestive of productive effects of entrepreneurial experience, rather than just a taste for entrepreneurship.

4.2.3

Formal education

One can argue for either a positive or a negative relationship between entrepreneurship and formal education. On one hand, education might improve entrepreneurial judgement by providing people with analytical abilities, information about business opportunities, and an understanding of markets and the entrepreneurial process (Casson, 1995). Formal education is also associated with general search skills, foresight, imagination and computational and communication skills, as well as with specific skills and knowledge needed to run businesses in particular sectors, e.g. engineering, computer science and biochemistry. Even if skills and knowledge gained from formal education are unnecessary for starting a business, in an empirical context they might provide a proxy for social background, ambition and endurance. And there might also be a selection effect at work, if more educated workers select themselves into occupations in which entrepreneurship is more common, such as managerial occupations among professionals (Evans and Leighton, 1989b) or skilled craft jobs among manual workers (Form, 1985). On the other hand, the skills that make entrepreneurs successful are unlikely to be the same as those embodied in formal qualifications (Casson, 2003). In particular, one might hesitate to suggest education as a proxy for managerial ability in entrepreneurship à la Lucas (1978) (see chapter 2). Perhaps even more importantly, education increases the value of the outside option of paid employment, which can make entrepreneurship relatively less attractive to highly educated people at the margin (Le, 1999). Most econometric studies of the effects of education on entrepreneurship utilise cross-section data. In these studies, entrepreneurship is usually measured as selfemployment or business ownership; and educational attainment is usually measured either as years of education completed, or as a set of dummy variables registering the highest qualification achieved by survey respondents. As with age, the majority of studies find a positive relationship between educational attainment and the probability of being or becoming an entrepreneur, though this is far from being a uniform finding, with less than 60 per cent of studies finding a significant positive effect rather than an insignificant or significant negative effect (see row 4 of Table 4.1). These findings relate almost exclusively to developed economies. Developing economies seem to exhibit a different pattern. A meta-analysis of eighty-four journal articles, book chapters and working papers published between 1980 and 2002 conducted by van der Sluis et al. (2005) revealed a greater tendency for educated workers to select paid employment instead of self-employment – although the latter is preferred to farming. This selection effect was found to be stronger for women, urban residents and inhabitants of poor economies where agriculture predominates and literacy rates are low. Most econometric studies conceal some subtleties in the relationship between education and entrepreneurship. First, the quality of education matters. BHPS data indicate

118

Selection

that former pupils of high-quality British public-sector secondary schools (‘grammar schools’) are significantly more likely to become self-employed than pupils of other schools (Henley, 2004). In the USA, there is evidence that counties operating school voucher programmes have rates of youth entrepreneurship that are approximately one percentage point higher than counties operating traditional public-school delivery programmes (Sobel and King, 2008). Second, reflecting the two fat tails of the entrepreneurial income distribution (chapter 13), there is a bifurcation in the nature of the qualifications held by entrepreneurs compared with non-entrepreneurs, with disproportionate numbers of entrepreneurs being found among the least and most educated groups (Lohmann and Luber, 2004). Third, partly reflecting the occupational and industrial substructure of self-employment, there is growing evidence that vocational qualifications and apprenticeship training rather than purely academic qualifications affect participation in entrepreneurship.15 Fourth, the estimated effects of education on entrepreneurship depend on whether the researcher controls for industry sector and occupational group (Bates, 1995, 1997; Le, 1999). For example, Bates (1995, 1997) reported positive and significant effects of education on the probability of entering selfemployment in skilled services; negative and significant effects on the probability of entering self-employment in construction; and insignificant effects on the probability of entering self-employment in manufacturing and wholesaling. Bates concluded that the overall impact of education on entrepreneurship is obscured by aggregation across dissimilar industries. Fifth, there can be ethnic differences in the relationship between education and entrepreneurship, owing to different cultural traditions. For example, a positive relationship has been found among British whites and a negative relationship among British Indians (Borooah and Hart, 1999). Sixth, there are international differences in the education–entrepreneurship relationship. For example, pooled crosssection time-series micro-data regressions reveal positive effects from education on self-employment choices in the USA but negative effects in the EU (Blanchflower, 2004). The impact of human capital on entrepreneurship propensities can be amplified by externalities. At the regional level, there is evidence that start-ups are more common in locations with higher proportions of highly educated people.16 There are several rationales for a positive externality. As explained in chapter 2, Parker (2005b) proposes one explanation based on the idea that individuals possess imperfect information about their entrepreneurial ability at the time they invest in human capital. Average stocks of human capital in a region provide people with information about the distribution of ability, which informs their occupational choices. Iyigun and Owen (1998, 1999) propose another mechanism in the context of an intergenerational model in which individuals choose between acquiring human capital which is more or less productive in entrepreneurship, but do not take into account the effects of their choices on the production possibilities of the next generation. This leads to the possibility of a dynamically inefficient mix of human capital and occupational choices. For instance, if entrepreneurial experience is the key determinant of future production technologies, there can be too little entrepreneurship, and therefore too little technology,

Evidence about the determinants of entrepreneurship

119

leading to skilled wages and human capital stocks which are too low. This strand of research complements the theories about knowledge-based externalities, growth and entrepreneurship reviewed in chapter 2. Finally, entrepreneurship education (which is training and education specifically about entrepreneurship) provides another link between human capital and entrepreneurial entry. Bridge et al. (1998) argue that many individuals do not choose an entrepreneurial career simply because it has never occurred to them. Raising awareness can be done through entrepreneurship education programmes, possibly delivered in schools. Lee and Wong (2006) provide a descriptive overview of entrepreneurship education initiatives, including evidence relating to its efficacy and limitations. 4.3

Social capital

An increasingly widely used construct in empirical models of entrepreneurship is social capital, Xsoc . According to Davidsson and Honig (2003, p. 307), ‘social capital refers to the ability of actors to extract benefits from their social structures, networks and relationships’. Social capital can exist at the country level, for example in the degree of trust in government and other institutions, and at the community level, such as the quality of social networks within the locality. Social networks can involve the extended family, communities and organisational relationships. Abell et al. (2001) argue that social capital confers social legitimacy upon entrepreneurship; reveals information about opportunities, customers, suppliers and competitors; and facilitates access to resources such as cheap labour and capital while providing psychological aid, such as helping entrepreneurs to weather emotional stress and to keep their businesses afloat.17 In principle, social capital might be used to compensate for limited financial or human capital. Social capital can comprise ‘strong’ or ‘weak’ ties. Strong ties come from close relationships such as one’s direct family or close friends, who can leverage support and trust needed for resource acquisition (Brüderl and Preisendörfer, 1998). Weak ties are loose relationships with former business contacts, acquaintances and members of business networks such as trade associations or guilds (Parker, 2008b). A pronounced feature of entrepreneurial networks and start-up teams is homophily, which is the tendency for ‘birds of a feather to flock together’.18 Sociologists argue that an important source of homophily arises from strong ties and dense social networks, which constrain individuals to start ventures with people like themselves. While this can facilitate trust and knowledge sharing, it can also close off sources of diverse information which could benefit the entrepreneur (Kim and Aldrich, 2005). An alternative view is that entrepreneurs are over-optimists who work hard and so prefer to match with other over-optimists. Homophily then arises from free choice as a type of assortative matching outcome (Parker, 2009a). Despite the prevalence of homophilious new venture teams, heterophily apparently endows teams with key organisational and performance advantages. With regard to performance, Eisenhardt and Schoonhoven (1990) observed that semiconductor firms

120

Selection

Table 4.3. Observations about social capital and implications for entrepreneurship Observation 1 Social capital first rises, then falls with age 2 Social capital declines with expected mobility 3 Social capital is associated with higher returns to skills 4 Social capital is higher among homeowners 5 Social connections decline with social distance 6 Social and human capital investments are complements

Implications for entrepreneurship Another explanation for the empirical observation discussed in section 4.2.1 Entrepreneurs are less mobile than employees As residual claimants, entrepreneurs should obtain more social capital Higher owner-occupier rates among entrepreneurs Entrepreneurship is more common in urban than in rural locations Failure to control for social capital in entrepreneurship research is likely to overstate the role of human capital

with more heterophilious founding teams enjoy faster organisational growth, while in a study of 161 young high-tech firms in Silicon Valley, Beckman et al. (2007) show that diverse prior experience of top management team members is associated with superior access to venture capital financing and a greater likelihood of achieving an IPO. This poses a puzzle about why entrepreneurs prefer to form homophilious teams with lower average performance. Parker (2009a) proposes a solution based on the marriage of two cognitive biases which will be discussed further below: over-optimism and self-serving attribution biases. Glaeser et al. (2002) propose several testable economic propositions about social capital. To the extent that social capital is more productive in entrepreneurship than in paid employment (a debatable point), several of these propositions carry implications for entrepreneurship, summarised in Table 4.3. Some researchers claim that social capital promotes entry into entrepreneurship.19 In fact, there is relatively little hard evidence that it does (hence the absence of an entry for social capital in Table 4.1). Part of the reason is the elusiveness of a compelling empirical operationalisation of the social capital concept (see below). Instead, most research to date has focused on the impact of social capital on entrepreneurial prospects and performance. There is, for example, a general consensus that membership of networks such as clubs and societies, churches, trade associations and other formal and informal entrepreneur networks enhances entrepreneurs’ profits, growth performance and survival prospects.20 By way of illustration, an illuminating study of Gujarati motel owners in Texas found that poorer Gujarati entrepreneurs who owned unbranded motels survived significantly longer if they were located close to successful Gujarati entrepreneurs who owned branded motels (Kalnins and Chung, 2006). Locating close to either co-ethnics running unbranded motels or non-ethnics running branded motels made no significant

Evidence about the determinants of entrepreneurship

121

difference to survival, suggesting a direct role for social capital. And high occupational status and relevant industrial work experience have been linked with entrepreneurs’ propensities to use social network ties as opposed to market methods to access start-up capital for high-tech ventures in Singapore and Beijing (Zhang et al., 2008). Several problems beset research on social capital and entrepreneurship. One is the generally poor quality of empirical proxies for social capital, reflecting limited data and a lack of agreement about how to measure this construct. For example, there is still an ongoing debate about the distinction between social ties and social capital (Karlan, 2007).Another limitation arises from correlations between social capital, unobserved ability, human capital and financial capital. To date, relatively few studies have recognised that this can cause endogeneity bias necessitating Instrumental Variable estimation. For example, individuals living in areas with plentiful opportunities might find it easier to form social ties; yet they could feasibly owe their superior economic performance to a more opportunity-rich environment rather than to stronger social networks. As Licht and Siegel put it: ‘improved identification strategies are needed to better delineate the mechanism by which investments in social capital lead to sustainable competitive advantage’ (2006, p. 531). Evidently, much more robust empirical work remains to be done in this area. A start in this direction has been made in the context of micro-finance group-lending schemes in developing countries, in which social ties play a major role. Further discussion of this issue is deferred until chapter 8. 4.4

Risk attitudes, over-optimism and other psychological traits

The economics of entrepreneurship has devoted considerable attention to two particular psychological characteristics: risk attitudes and over-optimism. These are considered in the first two subsections below. The third one considers several other psychological trait variables Xpsy , which have traditionally been more widely used in the psychology and business studies literatures, but which are sometimes also studied by economists. 4.4.1

Risk attitudes and risk Chapter 2 distinguished between risk attitudes – namely risk aversion embodied in individuals’ utility functions – and the actual or perceived level of risk itself. Both of these factors can affect the decision to become an entrepreneur. I will first discuss empirical work which measures risk attitudes and relates them to entrepreneurial choices. This material bears directly on risk-based theories of entrepreneurship, such as that of Kihlstrom and Laffont (1979) discussed in chapter 2. The discussion is followed by a short critique pointing out limitations of this line of research. Finally, I discuss empirical work which measures risk and its effects on entrepreneurial choices. Risk attitudes can be measured in various ways. It is helpful to distinguish between three empirical strategies:

• Univariate comparisons between entrepreneurs and non-entrepreneurs based on hypothetical survey questions about risk preferences;

122

Selection

• Comparisons based on lifestyle choices deemed indicative of risk aversion; • Econometric estimates of binary choice models of entrepreneurship using hypothetical survey responses to measure risk attitudes. There is a long history of univariate comparisons between entrepreneurs and nonentrepreneurs based on survey questions asking respondents how they would choose between risky hypothetical situations. A famous example is the Choice Dilemma Questionnaire, which in Brockhaus (1980) famously revealed no significant differences between entrepreneurs’ and managers’ risk attitudes. Shaver and Scott (1991) survey the early literature on univariate comparisons and highlight the limitations of studies purporting to show that entrepreneurs are less risk-averse than non-entrepreneurs. Yet a more recent meta-analysis of univariate comparisons by Stewart and Roth (2001) reveals that entrepreneurs appear to be significantly less risk-averse than managers, the difference being greatest for entrepreneurs whose primary goal is venture growth rather than merely income generation. Subsequently, however, Miner and Raju (2004) considered other studies based mainly on the ‘Miner Sentence Completion Scale’ of risk avoidance and disputed Stewart and Roth’s findings; Miner and Raju (2004) conclude by declaring that the univariate evidence on this issue remains inconclusive. The second approach looks for differences between entrepreneurs’ and nonentrepreneurs’ risk attitudes using the revealed preference principle, i.e. by observing whether entrepreneurs adopt riskier lifestyles than non-entrepreneurs. The evidence of this kind is also mixed. Self-employed Scandinavians are apparently less likely to participate in lotteries than employees are (Lindh and Ohlsson, 1996; Uusitalo, 2001); but Dutch entrepreneurs are significantly more willing to take hypothetical gambles than employees (van Praag and Cramer, 2001). Using US SCF (Survey of Consumer Finances) data, Puri and Robinson (2005) find that although nearly twice as many entrepreneurs (32 per cent) claimed they were willing to take above-average hypothetical risks compared with employees (17 per cent), entrepreneurs temper this with stronger family ties (higher marriage rates and more children); good health practices (being less likely to smoke); and longer self-reported financial planning horizons. Puri and Robinson take this as evidence that entrepreneurs are ‘calculated risk-takers’. Brown, Farrell, Harris and Sessions’ (2006) study of UK FES (Family Expenditure Survey) data over 1997–2000 reveal that the self-employed are significantly less likely than employees to purchase financial insurance products, which is consistent with the idea that they are less risk-averse. Most multivariate econometric analyses of binary choice models – the third approach above – utilise survey instruments for risk attitudes based on responses to hypothetical gambles or harm avoidance. They mostly support the hypothesis that entrepreneurs are significantly less risk-averse than the average (see the fifth row of Table 4.1).21 However, in this third as in the first approach, reliance on hypothetical rather than actual risk scenarios is problematic. Entrepreneurs might give ‘risk-loving’ responses because they think interviewers expect this of them. Or there might be reverse causality

Evidence about the determinants of entrepreneurship

123

from entrepreneurship to risk attitudes. For instance, Taylor’s (1996) finding from the 1991 BHPS that self-employed respondents are significantly less likely than employees to regard job security as an important job aspect could simply reflect self-employees’ familiarity with working in risky or uncertain conditions, and/or ex post rationalisation of their prior choices. Responding to this objection, Ekelund et al. (2005) argued that reverse causality from entrepreneurship to risk attitudes is unlikely in their sample, since long-established entrepreneurs appear to be as risk-averse as newly minted entrepreneurs. A different measure of risk aversion, ‘fear of failure’, is measured by GEM. Several studies using GEM data claim to find a significant negative association between this variable and the propensity to be a nascent entrepreneur.22 However, it is unclear whether this variable measures risk aversion or something else, like anticipated social stigma, making it a questionable proxy of risk aversion. And interpretative strategies such as inferring risk aversion from decisions of numerous former drug-dealers to become self-employed later in life (Fairlie, 2002) are also problematic. Competing explanations of this observation include a desire for autonomy and a criminal record being a barrier to taking conventional jobs in paid employment (Aidis and van Praag, 2007). Measuring risk aversion using specially designed ‘laboratory’ experimental games might appear to be a more promising empirical strategy. The available evidence of this kind suggests that entrepreneurs are either risk-neutral or mildly risk-averse, and at any event less risk-averse than non-entrepreneurs (Elston et al., 2005). But this approach is also vulnerable to the charge of analysing behaviour in hypothetical scenarios. Perhaps the best way to circumvent these problems in applied research is to use the revealed preference principle to record actual examples of risk-avoiding or riskseeking behaviour among respondents, as per the earlier discussion. This still potentially faces the problem of conflating risk attitudes with over-optimism. There is, for example, some evidence that entrepreneurs frame and interpret the same business stimuli more favourably than non-entrepreneurs do (Palich and Bagby, 1995; Norton and Moore, 2002). Thus researchers could misconstrue adventurous actions based on overoptimistic expectations of outcomes as evidence of greater risk tolerance. In fact, Puri and Robinson (2005) found low correlations between their measures of risk attitudes and over-optimism (see below); and in experimental games of risky entrepreneurial investments, Coelho (2004) also found that the two phenomena appear to be orthogonal. However, over-optimism is not the only factor that can masquerade as risk aversion. Others include over-confidence, causing insufficient subjective uncertainty (Simon and Houghton, 2003); bounded rationality, whereby entrepreneurs do not know what they do not know (Cooper et al., 1995); ‘animal spirits’ (Marchionatti, 1999) and subjective feelings of control (Sarasvathy et al., 1998). Cognitive biases might also be at play, such as a tendency for entrepreneurs to have short-term planning horizons and to engage less in counterfactual thinking (Baron, 2000; Bluedorn and Martin, 2008) – and to be readier than employees to extrapolate from small samples of information when starting new ventures (Simon et al., 2000).

124

Selection

If it is hard to measure risk attitudes precisely, it has proven no less challenging to measure market risk itself in applied research. Cross-section data is especially ill suited to the measurement of risk, which by definition is a temporal phenomenon. For example, numerous authors, especially in the franchising literature, have measured risk as the cross-section variance of industry sales or profits. However, as Lafontaine and Bhattacharyya (1995) point out, there are at least two serious problems with this approach. It takes no account of variations in ventures’ inputs which also lead to variation in outputs yet which are unrelated to risk; and it is prone to an aggregation problem since the variance of aggregate industry performance also depends on the covariances between different ventures’ performances within an industry which are unrelated to risk. These problems can be expected to be even more pronounced at higher levels of aggregation, such as the entire economy, where others have proposed deviations from trend levels or growth rates (of output or investment) as aggregate measures of risk.23 Instead of the variance of sector sales or profits, Lafontaine and Bhattacharyya (1995) recommend the use of industry discontinuation rates as measures of industry (downside) risk. In fact, this proposal is questionable too, because as will be explained in chapter 14, much discontinuation is voluntary and is not necessarily a response to risk. Arguably, a better approach is to measure temporal variations in profits using individual-level panel-data (Parker, Belghitar and Barmby, 2005). Even this measure, though, has its limitations because it ignores risk in the form of involuntary venture closures – which can also create non-random attrition in the constructed panel-data risk variable. To conclude, it seems that there is not yet a completely ideal way to measure entrepreneurial risk in applied research. This is an important limitation in view of the theoretical importance of risk and risk aversion in the economics of entrepreneurship.

4.4.2

Over-optimism and over-confidence It has long been recognised that people have a tendency to be innately over-optimistic, especially about events which are only partially under their control. As Adam Smith observed, ‘The overweening conceit which the greater part of men have of their abilities is an ancient evil remarked by the philosophers and moralists of all ages. Their absurd presumption in their own good fortune has been less taken notice of [but is], if possible, still more universal. . . The chance of gain is by every man more or less overvalued and the chance of loss by most men undervalued and by scarce any man valued more than it is worth’(1776 [1937, p. 107]). Smith adduced support for this view from contemporary (eighteenth century) examples of economic and social behaviour, including the facts that most residential houses and ships at sea were uninsured; that young people gave little weight to downside risk in their career choices; and that wages appeared to be insufficiently high in certain kinds of dangerous jobs. The psychology literature offers several explanations about why entrepreneurs may be unusually prone to unrealistic over-optimism.24 It turns out that optimism tends to be highest when individuals have emotional commitments to outcomes; when they believe outcomes are under their control; and when there is relatively little hard evidence

Evidence about the determinants of entrepreneurship

125

about the likelihood of success of an endeavour. As Coelho et al. (2004) point out, most entrepreneurs’personal wealth is tied up in their businesses, creating obvious emotional commitments; and setting up a new business is uncharted territory and conducive to illusions of control, with little ex ante evidence being available to restrain entrepreneurs’ ‘unchecked fantasising’. These are not the only explanations for over-optimism. Schade and Koellinger (2007) refer to psychological studies suggesting that complex tasks which lack fast and clear feedback are also ripe grounds for optimism. Business scholars also argue that the environments in which entrepreneurs operate are typically so noisy and overloaded with information that cognitive biases like over-optimism may be necessary to cope with them.25 Hence innately over-optimistic people are likely to self-select into entrepreneurship (Busenitz and Barney, 1997). Because one aspect of over-optimism is a cognitive bias entailing excessive focus on one’s own ability at the cost of ignoring competition by others, there is likely to be excessive entry in markets where trading conditions are perceived to be easy (e.g. retail and restaurants), but insufficient entry in others where conditions are perceived to be difficult (Moore et al., 2007). Essentially, many entrepreneurs seem to overlook the fact that markets which they find easy to enter are also easy for others to enter, leading them to underestimate the degree of competition they will face. Psychologists term this type of behaviour ‘reference group neglect’. Evidence from experimental simultaneous market entry games supports these ideas. In these games, participants’ payoffs are designed to decrease with the number of entrants. Subjects choose to participate in the game based on their beliefs about their skill at playing it. Outcomes of these games reveal that entrants seem to believe that the total profit earned by all entrants will be negative but that their own profit will be positive (Camerer and Lovallo, 1999). Bolger et al. (2008) design an experiment in which absolute over-optimism in one’s own abilities, rather than relative over-optimism (i.e. relative to that of other participants), is associated with excess entry – consistent with Moore et al. (2007). One might still wonder though whether there might be a rational basis for optimistic self-selection. As van den Steen (2004) argues, it can be rational for individuals to compete for rewards with high perceived as well as actual probabilities of success, even if the former exceed the latter. In any case, it is important to distinguish between over-optimism and overconfidence. Loosely speaking, over-optimists over-estimate the probability of success and expected returns from a venture, while over-confident people under-estimate the degree of variation of outcomes.26 Despite the fact that over-optimism and overconfidence are distinct concepts, much of the literature confusingly conflates them. At the risk of sounding pedantic, this practice should be discouraged in future. A large body of evidence indicates that although most individuals are over-optimistic, entrepreneurs are even more over-optimistic than average. For example, using a sample of US NFIB (National Federation of Independent Businesses) data, Cooper et al. (1988) reported that 68 per cent of their respondents believed that the odds of their business succeeding were better than for others in the same sector, while only 5 per cent thought

126

Selection

that they were worse.27 However, while interesting, these findings can be no more than suggestive. While the interviewees displayed greater optimism on average than is objectively merited by statistics about average business performance, it is possible that these beliefs might nevertheless have been warranted in these particular cases. Notably, Cooper et al. (1988) did not compare their expectations with outcomes (for which longitudinal data are needed) or establish whether entrepreneurs are more optimistic than non-entrepreneurs. This task has been attempted by Arabsheibani et al. (2000). Using BHPS panel data over 1990–96 to compare expectations of future prosperity with actual outcomes, Arabsheibani et al. (2000) found that employees and self-employed Britons are both systematically over-optimistic; but that the self-employed are consistently and substantially the most over-optimistic. For example, 4.6 times as many self-employed people forecast an improvement in their prosperity but later experienced deterioration, as forecast a deterioration but later experienced an improvement. For employees the ratio was 2.9. This finding of a pronounced disparity between expectations and outcomes, which is robust to the inclusion of controls for other personal characteristics, has been replicated by subsequent researchers.28 French data show that entrepreneurs who start new firms are more optimistic than those who take over existing firms; optimism is also positively associated with education and a desire for autonomy (Landier and Thesmar, 2008). Landier and Thesmar (2008) further found that over-optimism is negatively related to business performance, as measured by sales, asset and employment growth rates, and returns on assets and equity – suggesting that optimism is not merely an artefact of high risk–high return projects. In contrast, Coelho (2004) and Reid and Smith (2000) report insignificant correlations between measures of over-optimism and performance. It is not just those who actually start or take over a new venture who display over-optimism. According to Astebro’s (2003) study of over one thousand Canadian inventions between 1976 and 1993, the chance of a new innovation reaching the market is only 7 per cent. Of these ‘lucky’ 7 per cent, some 60 per cent realise negative returns, and the average realised return among those who commercialise their inventions is −7 per cent, even ignoring the cost of the inventor’s often enormous efforts. Interestingly, one-half of the inventors persist with their idea even when external advisors recommend abandonment. This suggests that over-optimism is deeply ingrained. In a similar vein, Lowe and Ziedonis’ (2006) analysis of the commercialisation efforts of university inventions reveals that on average entrepreneurs continue with development efforts longer than established firms do. However, while this finding is consistent with over-optimism, it might also be consistent with rational behaviour if there are reasons why new ventures take longer to develop (e.g. lower endowments of expertise and complementary assets). These findings may have important implications for market efficiency. When factor markets are characterised by upward-sloping supply curves, over-optimists may overuse scarce resources in equilibrium and bid up prices that realists must pay. Manove (2000) cites Warren Buffett on this point: ‘It is optimism that is the enemy of the

Evidence about the determinants of entrepreneurship

127

rational buyer.’ Even if entrepreneurs have unbiased profit expectations on average, over-optimistic entrepreneurs will drive out realistic entrepreneurs from product markets. This rebuts Friedman’s (1953) long-standing argument that profit-maximising realists will drive over-optimists out of competitive markets in the long run. Realists would make positive profits in the absence of the over-optimists, but may be unable to do so when optimists are present because optimists produce excess output that reduces prices below the break-even price in the industry.29 Over-optimists therefore impose a negative externality on all other entrepreneurs, and the appropriate policy is to discourage entry, rather than to encourage it (de Meza, 2002). A similar policy prescription applies to credit markets, whose efficiency is also generally damaged by the presence of over-optimistic entrepreneurs. This argument is developed further in chapter 7. On the other hand, optimism can confer some advantages on individuals and society. First, over-optimistic entrepreneurs can hold their own against realists by working longer hours, retiring later and saving more, thereby compensating for mistakes induced by their over-optimism (Manove, 2000; Puri and Robinson, 2005). Over-optimistic entrepreneurs exert greater effort because they expect higher returns for any given marginal cost of effort (Landier and Thesmar, 2008; Parker, 2009a). Potentially, this might counteract underprovision of effort owing to moral hazard considerations. Furthermore, entrepreneurs often depend on the efforts of co-founders, employees and suppliers. By working harder, they can increase the marginal productivity of colleagues who then work harder as well, especially in teams where a free-rider incentive would otherwise lead to insufficient effort. Provided they are not too over-optimistic, this externality from other colleagues can even make over-optimistic entrepreneurs net beneficiaries from their over-investment of effort (Gervais and Goldstein, 2006). Overoptimism might confer an additional advantage by signalling high ability to outsiders, such as customers or financiers. On the other hand, realistic entrepreneurs might be better placed to exploit cognitive biases in customers and suppliers than over-optimists, thereby enhancing their business prospects (Bhide, 2000, chap. 3). Second, Bernardo and Welch (2001) claim that over-optimistic entrepreneurs are less likely to imitate their peers and are more likely to explore their environment. While realists can become trapped in an ‘informational cascade’, taking imitative actions that add little or no new information and so perpetuate socially inefficient outcomes (Bikhchandani, Hershleifer and Welch, 1998), bold over-optimistic entrepreneurs might buck the trend by backing their own perceived odds and discounting those of predecessors. This generates valuable informational benefits to the entrepreneurial group, enabling it to thrive in spite of the costs incurred by the entrepreneurs who obtained the information. As Bernardo and Welch put it: ‘Unknowingly, overconfident entrepreneurs behave altruistically, making irrational choices that are to their own detriment but help their groups.’ (Bernardo and Welch, 2001, p. 302). Bernardo and Welch go on to claim that this externality gives the entrepreneur group an evolutionary advantage which helps it to survive. However, this argument is based on the notion of group selection rather than individual selection, which is not in accordance with the received wisdom among evolutionary biologists.

128

Selection

De Meza and Southey (1996) claim that unrealistic optimism explains several wellknown features of entrepreneurship, including persistent business entry despite low incomes and high industry exit rates (see chapters 13 and 14). Over-optimism can also explain why purely self-financed entrepreneurs face higher business failure rates than debt-financed entrepreneurs. By overestimating their prospects of success, optimists prefer maximum self-finance of their projects – unlike realists, who prefer debt finance. But also unlike realists, optimists can tolerate making negative expected returns net of opportunity costs, leading to higher eventual exit rates for this group as their resources run out. The risk of failure is exacerbated by over-optimistic entrepreneurs underestimating the initial resource endowments their ventures require and overestimating their ability to obtain further resources. Over-optimistic entrepreneurs also run the risk of deploying their resources too precipitately and rejecting valuable venture liquidity opportunities by being unwilling to share equity – all making their ventures more prone to failure (Hayward et al., 2006). As a practical suggestion, entrepreneurs are well advised to counter the malign effects of over-optimism by forming relationships with outsiders, such as non-executive board members and professional advisors (Cooper et al., 1988). Outsiders have the objectivity and detachment to counteract unrealistic optimism – hopefully without extinguishing the essential spark of entrepreneurial zeal. Alternatively, Hayward et al. (2006) suggest that entrepreneurs should assemble diverse venture teams in which founders can more easily engage in counterfactual reasoning. However, as noted earlier in this chapter, homophily rather than heterophily generally regulates the composition of new venture teams in practice. Another practical problem is that entrepreneurs often ignore advice that runs counter to their will; and external advisors may also be prone to biases or have their own agendas that are not aligned with the entrepreneur’s interests. Furthermore, in contrast to theoretical work assuming that learning can eliminate entrepreneurial over-optimism (Parker, 2007c), evidence suggests that over-optimism is very persistent (Landier and Thesmar, 2008), something borne out by venture capitalists’ claims that serial entrepreneurs are if anything less able to recognise their own limitations than first-time entrepreneurs are (Wright et al., 1997). The combination of over-optimism with self-attribution bias can in theory explain persistent (and even escalating) over-optimism (Parker, 2009a). Clearly further research is needed on these issues, including the extent to which learning actually mitigates over-optimistic biases. Research is also needed to quantify the externalities caused by over-optimism discussed above, and to determine the appropriate public policy set when over-optimism is present.

4.4.3

Other psychological trait variables More generally, a long-standing research question is whether individuals possess special psychological traits, Xpsy , which predispose them to entrepreneurship. Early studies in this area usually deployed simple univariate statistical comparisons between entrepreneurs and non-entrepreneurs (see Begley and Boyd, 1987, for a representative

Evidence about the determinants of entrepreneurship

129

example). More recently, researchers have started to utilise multivariate methods to control for other individual characteristics and so obtain cleaner estimates of the role of psychological traits. In their review of the role of psychological factors in entrepreneurship research, Amit et al. (1993) singled out risk attitudes and three other traits that have attracted substantial research interest: 1. Need for achievement. In one of the first systematic attempts to provide a psychological profile of entrepreneurs, McClelland (1961) posited a ‘need for achievement’ (n-Ach) as the key characteristic of successful entrepreneurs. In McClelland’s words: ‘a society with a generally high level of n-Ach will produce more energetic entrepreneurs who, in turn, produce rapid economic development’ (1961, p. 205). McClelland claimed that entrepreneurs are: proactive and committed to others; like to take personal responsibility for their decisions; prefer decisions involving a moderate amount of risk; desire feedback on their performance; and dislike repetitive, routine work. A corollary of McClelland’s thesis is that the achievement motive can be deliberately inculcated through socialisation and training, although results from such efforts have drawn a mixed response (Chell et al., 1991). It is not clear that n-Ach can be ‘coached’, as has been claimed; and it is questionable whether entrepreneurship is the only vocation in which n-Ach can be expressed (Sexton and Bowman, 1985). 2. Internal locus of control. Another psychological trait is a person’s innate belief that their performance depends largely on their own actions, rather than external factors. Psychologists refer to this as a ‘high internal locus of control’. If entrepreneurship offers greater scope for individuals to exercise their own discretion at work than paid employment does, it follows that those with a high internal locus of control are likelier to become entrepreneurs. A psychological metric known as the Rotter Scale (Rotter, 1982) provides a basis for empirical tests of this hypothesis. The more a survey respondent believes a range of factors to be under her control, as opposed to outside her control, the lower her Rotter score in any given test. Evidence from multivariate binary choice models conditioning on Rotter scores is mixed, mirroring results obtained from univariate studies.30 The reasons for the mixed results are unclear. It could be that a high locus of control is not unique to entrepreneurs, being relevant for successful business managers and other employees too (Sexton and Bowman, 1985). Alternatively, negative empirical findings might reflect the inappropriate use by researchers of locus of control responses measured during respondents’ childhoods. These might understate the importance of this variable, since personalities tend to stabilise only later in life. 3. Tolerance of ambiguity. It is proposed that entrepreneurs have a greater capacity than employees for dealing with environments where the overall framework is illdefined, or ambiguous. Wennekers et al.’s (2005) conjecture that individuals dislike ambiguity, something they refer to as ‘uncertainty avoidance’. In his study of successful US Inc. 500 business owners, Bhide (2000) claimed that self-confidence lies at the root of ambiguity tolerance, and is distinct from risk tolerance.

130

Selection

The above list of traits is by no means exhaustive. Others have claimed that entrepreneurs: • Exhibit ‘Type A’behaviour, which is characterised by competitiveness, assertiveness, aggression, a striving for achievement, and impatience (Boyd, 1984; Caliendo and Kritikos, 2008); • Possess greater levels of ‘self-efficacy’ and ‘intuition’ than non-entrepreneurs, being more intelligent, perceptive, creative and less likely to identify risks than non-entrepreneurs (Shane, 2003, ch. 5); • Are less likely to exhibit ‘status quo bias’ (e.g. by repeating previous choices) than bankers are (Burmeister and Schade, 2007); • Achieve higher scores for ‘conscientiousness’ and ‘openness to experience’ than managers; lower scores for ‘neuroticism’ and ‘agreeableness’; and similar degrees of ‘extroversion’ (Zhao and Seibert, 2006); • Are misfits or displaced persons who live outside the mainstream of society, finding it hard to accept authority or to work with others, and prone to deviant and criminal behaviour, possibly emanating from a disrupted childhood (Shapero, 1975; Kets de Vries, 1977). Such people establish new enterprises in an act of ‘innovative rebelliousness’ to boost their self-esteem and acquire external approbation. Subsequent research has shown that criminals evince high levels of interest and participation in entrepreneurship (Light and Rosenstein, 1995; Fairlie, 2002), although few such studies control for the damage a criminal record can do for conventional employment prospects. In fact, representative data samples show that more entrepreneurs come from relatively stable work and family backgrounds than from an isolated or deviant fringe. This might reflect the need for entrepreneurs in modern economies to co-operate with others rather than strike out alone as ‘heroic solo entrepreneurs’ (Reynolds and White, 1997; Blanchflower and Oswald, 1998). Taking a biological perspective, one can ask whether psychological traits have a genetic origin. Recent research provides some tantalising clues. White et al. (2006, 2007) collected data from thirty-one male Canadian MBA students declaring significant prior involvement in new venture creation and from seventy-nine other male students declaring no prior start-up experience. White et al. (2006) reported that a physiological characteristic, testosterone, was significantly associated with a questionnaire-based measure of risk tolerance; and that both testosterone and risk tolerance were significantly associated with involvement in new venture creation. In their 2007 follow-up study, White et al. reported that the interaction of testosterone with a family business background was associated with involvement in new venture creation. It seems that individuals with higher testosterone are predisposed towards risky behaviour; and favourable family backgrounds channel this propensity in the particular direction of entrepreneurship. In contrast, less favourable family backgrounds might channel this propensity into socially deviant behaviour. As testosterone is over 80 per cent heritable, and had no independent effects on new venture creation in White et al.’s 2007 study, one possible inference is that entrepreneurs are ‘made as well as born’.

Evidence about the determinants of entrepreneurship

131

More generally, scientific research suggests that inherited traits account for approximately 50 per cent of the differences in human personalities, with the rest coming from environmental influences and through people maturing over their life span (Bouchard and Loehlin, 2001; Shirey, 2006). For example, the heritability of IQ is generally estimated at between 50 and 70 per cent, so if IQ is associated with entrepreneurship (as the Dutch evidence of de Wit and van Winden, 1989, and van Praag and Cramer, 2001, suggests), one can perhaps explain intergenerational transmission of entrepreneurship (see below) at least partly in terms of genetic traits. Using a sample of data on monozygotic British twins, Nicolaou et al. (2008) estimate that 48 per cent of the tendency to become self-employed can be attributed to genetic similarities. All of this evidence is consistent with a ‘balanced’ view that entrepreneurs are both born and made. Beginning in the 1980s, there have been growing objections to the whole psychological traits approach. It has still not been decisively established whether there is an ‘essential’ set of entrepreneurial characteristics, and if so what they are. Partly this reflects the small sizes and non-comparability of the samples used in previous research, and the conflicting results obtained from them. Brock and Evans (1986) argued that ‘the scientific validity of these studies, which are seldom based on random samples and often use ambiguous or overly inclusive definitions of an entrepreneur, is open to question’ (1986, n. 9, p. 190). A salient bias could be towards sampling only successful entrepreneurs, leading to the danger that observed traits are confused with entrepreneurial experience (Amit et al., 1993). It may be noteworthy too that psychological factors have a weaker effect on venture growth rates than on participation in entrepreneurship.31 Overall it seems unlikely that such a diverse group of individuals as entrepreneurs are amenable to glib generalisations in terms of their psychological characteristics. Traits are unlikely to be unique to entrepreneurs, raising a demarcation problem; and being unobservable ex ante, they are virtually impossible to separate ex post from luck and other extraneous factors (Amit et al., 1993). As Gartner (1988) put it: ‘the diversity among entrepreneurs is much larger than differences between entrepreneurs and non-entrepreneurs’. There is also, as Kaufmann and Dant have pointed out, ‘a tendency in this literature to personify entrepreneurs as embodiments of all that may be desirable in a business person, and almost deify entrepreneurs in the process’ (1998, pp. 7–8). The conclusion reached by several authors, including this one, is that psychological factors are neither necessary nor sufficient conditions for entrepreneurs or entrepreneurship. Nevertheless, this research programme continues, despite several attacks on it, including Gartner’s (1988) influential argument that it is the behaviours involved in creating new ventures, rather than the personalities of founders, which are fundamental to entrepreneurship. It seems that many people still find it hard to accept that individual psychological differences do not affect entrepreneurial behaviours – including writers in the popular press, practitioners who work with entrepreneurs and entrepreneurs themselves. In response, entrepreneurship researchers in the business studies tradition have shifted their attention to cognitive aspects of entrepreneurial behaviour. However, much of

132

Selection

this research effort focuses on entrepreneurial intentions and orientations – constructs which are hard to operationalise, veer towards tautology, and which in any case properly fall outside the remit of the economics of entrepreneurship. 4.5

Demographic and industry characteristics

This section discusses the influence of three prominent demographic characteristics, Xdem : marital status, health issues and family background, before turning to the role of industry characteristics, Xind . 4.5.1 Marital status One might expect a disproportionate number of married people to be entrepreneurs compared with single people, for the following reasons:

1. A spouse can help provide start-up capital. 2. Once in business, a spouse can provide labour at below-market rates (for example ‘doing the books’ or taking telephone bookings while their husband or wife is out on jobs). Spouses can make trustworthy workers, being less likely to shirk (Borjas, 1986). And spouses can offer valuable emotional support (Brüderl and Preisendörfer, 1998; Bosma et al., 2004). 3. Spouses can use their own income as insurance against the risky income of their entrepreneur husband or wife. 4. Spouses can share relevant information and knowledge about business ownership and business conditions easily and efficiently. Knowledge spillovers from a spouse who is already in business can therefore increase the likelihood that the other starts up a business of their own (Parker, 2008a). 5. Having a spouse may confer tax advantages, including income sharing to exploit personal tax allowances; introducing the spouse as a ‘sleeping partner’and allocating them a share of the enterprise’s profits; or, if trading through a limited company, providing them with benefits (such as a company car, private medical insurance or payments into a pension scheme). 6. Entrepreneurs are older on average, and older people are more likely to be married. Some of these arguments have prompted several researchers to regard marriage as a form of social capital (e.g. Davidsson and Honig, 2003). On the other hand, married people with children may be unwilling to take the risks associated with entrepreneurship. Cross-section econometric evidence from binary choice models tells a consistent story. People who are, or who become, entrepreneurs are likelier to be, or to have been, married, with dependent children – especially if their spouse is working (see rows 6 and 7 of Table 4.1). This finding appears to hold quite generally, with the possible exception of black Americans (Borjas, 1986) and some ethnic minority English people (Clark and Drinkwater, 2000) – possibly reflecting weaker family structures among these ethnic

Evidence about the determinants of entrepreneurship

133

groups. Because most of these findings control for individuals’ ages, they are probably capturing some kind of co-operative factor as in points 1–5 above rather than indirect effects from age itself (point 6). Point 2 receives some specific support from Brüderl and Preisendörfer (1998), whose regression estimates indicated that self-reported levels of emotional support from a spouse improves the survival prospects of new German business ventures. All of points 1 through 6 above are consistent with evidence that being married is associated with higher incomes in entrepreneurship.32 A different but related question is why so many married couples are both entrepreneurs (Brown, Farrell and Sessions, 2006; Parker, 2008a). Numerous studies have shown that having a husband who is an entrepreneur makes a woman significantly more likely to become one. The evidence on this point, which is reviewed in chapter 5, tends to be based on single-equation binary choice models of occupational choice. There is less evidence about whether women entrepreneurs make their husbands more likely to be entrepreneurs as well; and whether these choices are interdependent. Parker (2008a) estimated the simultaneous-equation binary choice model (3.17) and (3.18) in chapter 3 and detected strong positive interdependence in business ownership, for both genders. Parker (2008a) estimated that husbands whose wife is certain of being a business owner have on average a 17 percentage point higher probability of being a business owner than if they are married to a woman who is certain not to be a business owner. The corresponding figure for women is 19 percentage points. This compares with the unconditional probability of being a business owner in the PSID of 13.3 per cent for men and 6.8 per cent for women. These effects are not only large, but also appear to be consistent with knowledge spillovers between spouses. Parker (2008a) found little support for alternative explanations of interdependent occupational choices in entrepreneurship among spouses, including risk diversification, wealth transfers, role-model effects and assortative mating. 4.5.2

Health issues Entrepreneurship is often believed to offer greater flexibility than paid employment in terms of an individual’s discretion over the length, location and scheduling of their work time (Quinn, 1980). To the extent that people with poor health or disabilities need such flexibility, it might be expected that, all else equal, they have higher rates of entrepreneurship. In addition, entrepreneurship may offer a route out of employer discrimination against the disabled. However, entrepreneurship can also be a poor choice for individuals in poor health. Some jobs with high concentrations of self-employment, such as construction, are intrinsically less suited to those with disabilities – not to mention more dangerous, implying that entrepreneurship can also cause disability and ill health. As mentioned earlier in the chapter, work hours and stress are also greater on average in entrepreneurship. The evidence on the effects of health on entrepreneurship participation (in row 8 of Table 4.1) is mixed, mirroring the conflicting theoretical arguments. Some interesting empirical findings suggest that entrepreneurial performance can feed back positively into entrepreneurs’ state of health (Rau et al., 2008). On the other

134

Selection

hand, long work hours (see chapter 12) and the difficulty of detaching oneself from work in entrepreneurship (‘workaholism’), tend to be negatively related to health outcomes (Rau et al., 2008; Taris et al., 2008). An important policy issue relates to health insurance coverage. Whereas many employees receive subsidised health cover from their employers, entrepreneurs must provide their own. That might explain why self-employed Americans have lower average rates of health insurance coverage than employees do.33 For example, Perry and Rosen (2004) report a coverage rate of 69 per cent for self-employed Americans in 1996, compared with over 81 per cent among employees. This partly reflects high purchase prices of self-employed health insurance, and the more favourable tax treatment of health insurance for employees than the self-employed prior to 2003. Since 2003, self-employed Americans have been able to deduct their entire health premiums as business expenses from their tax liabilities. It might be thought that non-portable employer-based health insurance coverage would discourage employees from leaving their firms to start new ventures. Strikingly, however, the evidence does not support this notion (Holtz-Eakin et al., 1996; Bruce et al., 2000). This is surprising. It is possible that individuals who contemplate becoming entrepreneurs discount health insurance benefits, especially if they are over-optimistic. But we do not know the true reasons.

4.5.3

Family background Entrepreneurship runs in families. Calculations based on 1992 CBO data by Fairlie and Robb (2007a) indicate that over half (51.6 per cent) of all business owners had a self-employed family member prior to starting their business.34 According to Dunn and Holtz-Eakin (2000), a son’s probability of becoming self-employed doubles (from 0.015 to 0.031) when either of his parents is self-employed. According to a collection of papers from eleven nations assembled by Arum and Müller (2004), having a self-employed father raises the odds of a son transitioning into self-employment by a factor of between 1.3 and 2.2; the figure may be as high as 3.0 in France (Colombier and Masclet, 2008). Parental self-employment both increases the fraction of time that offspring spend in self-employment and reduces the age at which they enter it (Dunn and Holtz-Eakin, 2000; Niittykangas and Tervo, 2005). Fairlie and Robb (2007a) propose several reasons why having a parent who is an entrepreneur might increase the probability that a given individual eventually becomes an entrepreneur themselves:

1. Acquisition of general business or managerial experience obtained from proximity to a family-owned business (‘general business human capital’); 2. Acquisition of industry- or firm-specific business experience obtained from proximity to a family-owned business – possibly including business networks (‘specific business human capital’); 3. Inheritance of a business;

Evidence about the determinants of entrepreneurship

135

4. Provision of cheap finance by parents to help their offspring to overcome borrowing constraints; 5. Correlated preferences for entrepreneurial activities among family members, perhaps enhanced by favourable family role models. Evidence of strong intergenerational linkages in entrepreneurship between parents and children survives in multivariate settings, where plentiful control variables are used. In such settings, parental (especially fathers’) participation in entrepreneurship is frequently the most important determinant of entrepreneurship among offspring (especially sons). This is true in a wide range of countries, including European nations (Niittykangas and Tervo, 2005; Henley, 2007; Colombier and Masclet, 2008), the USA (Fairlie and Robb, 2007a) and China (Djankov et al., 2006). The balance of evidence is summarised in row 9 of Table 4.1. Instead of describing each study, my strategy below is to first consider what the evidence says about each of the five candidate explanations listed above, before discussing the limitations of empirical tests on this issue. Several studies lend support to explanations 1 and 2. Dunn and Holtz-Eakin (2000) observed that parental experience and parental business success in entrepreneurship (e.g. profitability) almost doubles the probability of a son entering self-employment, even after controlling for the individual’s and parent’s financial capital and the individual’s human capital. The importance of parental experience and success suggests that parents primarily transfer managerial skills to their offspring, rather than merely familiarity with or a taste for entrepreneurship (as suggested by explanation 5). For their part, Fairlie and Robb (2007a) uncovered a significant positive linkage between experience obtained from prior work in a family member’s business and the business owner’s own success, measured in terms of survival rates, employment creation, sales and profits. For example, having worked in a family member’s business reduces the probability of business closure by 0.042; increases the probability of substantial business profits and employment by 0.032 and 0.055 respectively; and increases sales by 40 per cent. In contrast, a dummy variable for merely having a family member who owned a business had a small and statistically insignificant effect. Both findings are consistent with the notion that experience of working in a parent’s venture transfers performance-enhancing general and specific business human capital, rather than intergenerational transfers of preferences or abilities. This evidence provides further support for explanations 1 and 2. So do findings from Kim et al. (2006) that parental business ownership does not affect start-up organising efforts (‘nascent entrepreneurship’), while it does affect actual entrepreneurship. This suggests that early survival in business is enhanced by parental entrepreneurship. However, Fairlie and Robb (2007a) observed that over one-half of small business owners who had a self-employed family member had not worked in that family member’s business before, suggesting that 1 and 2 alone cannot fully explain the data. And Danish evidence shows that parental self-employment significantly increases the likelihood of transitions to self-employment of offspring, regardless of whether the parental self-employment occurred during adolescence or childhood. Even if the parents

136

Selection

were no longer self-employed by the time their children entered the labour force, the consequences of early exposure were found to persist (Sørensen, 2006).35 Arguably, these findings militate against the ‘social network’ aspect of parental entrepreneurship transmission (explanation 2). There is a clear consensus that explanation 3 does not explain intergenerational entrepreneurship. According to Fairlie and Robb’s (2007a) exploration of 1992 CBO data, only 1.6 per cent of all small businesses are inherited and only 4 per cent of entrepreneurs obtain their businesses through transfers of ownership or gifts. Furthermore, both business heirs and non-heirs benefit in similar ways from having a self-employed parent, implying that business networks inherited from the transfer of a parent’s business are not needed to achieve good financial performance (Lentz and Laband, 1990). Evidence relating to effects of wealth, tastes and role models (explanations 4 and 5) is more opaque. Dunn and Holtz-Eakin (2000) found that self-employed parents are three times wealthier on average than non-self-employed parents. Yet a $10,000 increase in parental assets increases the probability of a son’s annual transition to selfemployment by only 0.0009. This effect seems modest compared with a probability of entry conditional on parental self-employment of 0.031. It is actually unclear whether wealthy entrepreneurial backgrounds chiefly clarify risk-return trade-offs in the minds of children, furnish favourable role models, or provide insurance against failure – all of which could promote entrepreneurship among offspring. Halaby (2003) and Sørensen (2007) observe that the offspring of entrepreneurs are more likely to stress the importance of variety in their jobs and are likelier as well to work for small rather than large firms. These findings accord with the taste transmission idea. Against explanation 4, only 6–8 per cent of business owners borrow capital from family members (Aldrich et al., 1998; Fairlie and Robb, 2007a). It should be recognised that entrepreneurial role models might extend beyond the boundaries of the traditional nuclear family, to include friends and neighbours as well. Several studies suggest that personally knowing someone who has recently set up a business significantly increases the probability of nascent entrepreneurship,36 while Djankov et al. (2006) point out that Chinese entrepreneurs are significantly more likely to have childhood friends who were also entrepreneurs. Further evidence that role models are broad, and not just confined to parents, comes from Stuart and Ding (2006), who show that exposure to colleagues who had previously made the transition to commercial science significantly increases scientists’ hazards of either founding or advising a successful biotechnology company. Other evidence is consistent with several of the proposed explanations, and cannot clearly distinguish between them. For example, having parents who are employers significantly increases the probability of self-employment in Britain, whereas having parents who are sole proprietors does not (Henley, 2004; Taylor, 2004). This could be because a child whose father employs others views entrepreneurship as a higher-status occupation than one whose parent is a sole trader (explanation 5). On the other hand, this finding could alternatively be capturing transfers of greater managerial experience

Evidence about the determinants of entrepreneurship

137

associated with employing others (explanation 1), or wealth effects if employers are wealthier than sole proprietors (explanation 4). To summarise so far, it seems that transfers of general and specific human capital play an important, but far from universal, role in shaping the strong intergenerational correlations in entrepreneurship that so many researchers have observed. However, two drawbacks of the current state of research on this topic should give us pause for thought. These are a lack of information about other aspects of parents’ occupations, and clean distinctions between general and specific human capital in the intergenerational transmission process. What we do know about these issues chiefly derives from research into ‘occupational following’, whereby children work in similar trades (e.g. plumbing) to their parents, which tend to be associated with self-employment. In fact, only a minority of entrepreneurs’ sons follow their parents’ occupation; and controlling for parents being in occupations with high self-employment rates does not remove the influence of parental entrepreneurship on the probability that their offspring are entrepreneurs.37 This evidence seems to highlight general rather than specific skill transmission. Sons are significantly more likely to become self-employed if their father is a manager, a professional or a retailer, and significantly less likely to do so if their father is unskilled.38 On the other hand, Meager and Bates (2004) report that parental transmission of selfemployment propensities are important for both skilled and unskilled self-employed parents in Britain, but not for professional self-employed parents, many of whom (e.g. doctors and lawyers) are not ‘real’ entrepreneurs at all. Staying with the human capital composition issue, French evidence indicates that people without self-employed parents obtain more formal human capital than those with self-employed parents (Colombier and Masclet, 2008). This implies that offspring substitute formal for informal human capital, where the latter is derived from working in a family business – consistent with theoretical arguments advanced by Parker and van Praag (2006a). Whatever the reasons for the intergenerational transmission of self-employment propensities, it appears not to take the same form for sons and daughters. Finnish evidence from Niittykangas and Tervo (2005) shows that sons are more likely to follow parents into self-employment than daughters are; but daughters obtain more education than sons. According to Dunn and Holtz-Eakin (2000), fathers’ self-employment experience has a stronger effect on the probability that sons become self-employed than mothers’ self-employment experience does. There is in contrast mixed evidence about the effects of mothers’ self-employment experience on entrepreneurship among their offspring.39 Generally, mothers’ self-employment experience is more important for daughters than for sons; but having two self-employed parents imparts the greatest overall effect. PSED data reveals that sons work more in their parents’ businesses on average than daughters do, although such work experience seems to have little impact on the probability of starting a business of one’s own, and so cannot explain the gender difference in this respect (Aldrich and Kim, 2006). With regard to ethnicity, African-Americans are less likely than white Americans to have self-employed fathers. And those who do have self-employed fathers receive

138

Selection

smaller impacts on average from parental self-employment on both their own probabilities of becoming entrepreneurs and their performance in entrepreneurship.40 For example, according to 1992 CBO data more than one-half of white business owners had a self-employed family member owner prior to starting their business, compared with approximately one-third of black business owners (Fairlie and Robb, 2007b). Higher rates of marital and family breakdown among blacks (who are 40 per cent less likely to be married than whites) itself no doubt reduces the scope for intergenerational transmission of business ownership for this group. Regarding performance, Fairlie and Robb (2007b) estimated that the lower rates of previous experience of working in a family member’s business among blacks explain between 5.6 and 11.6 per cent of the performance differential between black- and white-owned offspring businesses. In contrast, ethnic differences in simply having a self-employed family member or having inherited a business explain little of the ethnic difference in entrepreneurial performance. More important are ethnic differences in start-up capital, which can explain between 14.5 per cent and 43.2 per cent of ethnic differences in business performance (Fairlie and Robb, 2007b).

4.5.4

Industry characteristics Entrepreneurs are more likely to be found in some industries and occupations than in others. As well as providing some evidence on this point, the discussion below suggests some possible reasons for it. The available evidence is mainly derived from studies which define entrepreneurship in terms of self-employment. The pronounced heterogeneity of entrepreneurs means that one has to exercise caution when discussing industry and occupational dimensions of entrepreneurship. For example, many developed countries in the 1980s and 1990s witnessed pronounced growth in both professional/managerial and low-skill self-employment, at the expense of traditional ‘artisan’self-employment (Müller and Arum, 2004). Nevertheless, several common factors can be elucidated. Entrepreneurs are likelier to enter service rather than manufacturing industries in part because entry barriers and minimum efficient scale are lower in the former than the latter, making sustainable entry easier.41 Entrepreneurs also tend to be found in larger, more profitable and more segmented industries and markets (Hause and Du Rietz, 1984). Entrepreneurs naturally prefer to enter larger markets, but they also quit these markets more frequently owing to stronger competitive pressures (Nocke, 2006). It is scarcely surprising that entrepreneurs prefer to enter industries with lower costs and lower levels of industrial concentration.42 Looking at industry sectors and occupations in greater detail, male entrepreneurs in North America tend to be concentrated in construction, services and retail trades; and in the sales, agriculture, hotels, repairs, craft, managerial and professional occupations. The predominance of service sector self-employment in America has a long history. It pre-dates the recent increases in aggregate self-employment rates in many developed economies and so cannot explain these increases (Aronson, 1991). The role of entrepreneurs in each of the agricultural, transportation, communications and retail

Evidence about the determinants of entrepreneurship

139

sectors has steadily declined over time (Steinmetz and Wright, 1989; Aronson, 1991). Similar industry variations and trends in entrepreneurship are observed in the UK and other OECD countries. Certainly international differences in rates of entrepreneurship cannot be explained by appealing to differences in nations’ industry structures (Parker and Robson, 2004; Torrini, 2005). In particular, OECD countries share high concentrations of self-employment in construction, distribution, hotel and repairs, banking and financial services (Loufti, 1992). According to 1991 LFS data, these sectors accounted for 62 per cent of all UK self-employment, although construction in particular is highly cyclical (Harvey, 1995; Meager and Bates, 2004). Agriculture, construction, distribution and finance witness the highest entrepreneurial entry rates in Britain (Taylor, 1996), corresponding to high concentrations of entrepreneurs in both professional and unskilled occupations (Henley, 2004). However, there has been a striking trend decrease in the proportion of self-employees engaged in distribution, hotels and restaurants, from 29.9 per cent in 1984 to 18.7 per cent in 2000 (Meager and Bates, 2004). On the face of it, the last strand of evidence bears out Lucas’ (1978) prediction that small family-owned businesses in these sectors will be progressively replaced by large hospitality and catering chains as capital accumulates in the economy (see chapter 2). To conclude, while entrepreneurship is often concentrated in particular industries, different industry structures can explain neither changes in rates of entrepreneurship over time nor why some places are persistently more entrepreneurial than others.

4.6

Macroeconomic factors

Three broad macroeconomic theories of entrepreneurship and growth were outlined in chapter 2. These were wealth-based, technology-based and knowledge-based theories. Wealth-based theories have attracted an enormous empirical research effort connected to the literature on borrowing constraints, which is reviewed separately in chapter 9. The first two subsections below survey the evidence under the other two headings of technology- and knowledge-based economic development. The third subsection analyses entrepreneurship and the business cycle. Evidence relating to an important macroeconomic policy variable, the interest rate, is also discussed here. The fourth subsection treats a related issue: the well-researched relationship between entrepreneurship and the unemployment rate. The final subsection examines entrepreneurship at an intermediate level of aggregation: the regional level. 4.6.1

Technology as a determinant of entrepreneurship

Improvements in technology are known to drive growth. Such improvements are often measured in terms of their impact on ‘total factor productivity’ (TFP). Growth in TFP is the amount by which output increases as a result of improvements in methods of production, with all inputs unchanged. Blau (1987) was one of the first researchers to explore the role of TFP in entrepreneurship. Blau claimed that changes in TFP ratios

140

Selection

favouring industries in which self-employment is common was one of the principal causes of rising US self-employment between 1972 and 1982. Blau (1987) went on to suggest that the spread of computers might have been the source of technological change. A problem with this explanation is that self-employment rates have declined steadily in France and Japan over the same period, despite these countries experiencing similar productivity improvements. And personal computers were not widespread even by 1982, the last year of Blau’s sample. Furthermore, Fairlie and Meyer (2000) provide separate evidence questioning the ability of TFP to explain trends in US self-employment rates. More recent evidence points to a stronger link between computers and entrepreneurship, though. Computers are widespread in small businesses, with more than 75 per cent of small US businesses using them in the 1990s according to the 1998 Survey of Small Business Finances (Bitler et al., 2001). Utilising matched CPS and Computer and Internet Usage data from 1997–2001, Fairlie (2006a) showed that computer usage increases the probability that American adults become business owners, not only in the IT sector but across a range of industries. Fairlie (2006a) detected a stronger effect of computer usage on business ownership among women than men. IT might create new opportunities for entrepreneurial firms to outsource goods and services and/or to supply goods and services outsourced by other firms (see chapter 2). Thus evidence that investment in IT is associated with subsequent decreases in the average size of firms is ‘consistent with the hypothesis that IT is facilitating the ‘decoupling’ of existing vertically integrated firms and the supplanting of existing firms by value-adding networks of new, smaller firms . . . the current downsizing of firms, the popularity of outsourcing, and the rise of value-adding partnerships is not simply a management fad, but rather may have a technological and theoretical basis’ (Brynjolfsson et al., 1994, pp. 1640, 1642). More generally, it has been proposed that the relationship between entrepreneurship and exogenous technological change is U-shaped (Acs, Audretsch and Evans, 1994). At early stages of a country’s development, technological change shifts output away from agriculture and small-scale manufacturing towards large-scale manufacturing, with a consequent reduction in self-employment rates. At later stages of development, technology changes again such that manufacturing gives way to services and self-employment rates recover. It is possible to test this hypothesis by regressing cross-country selfemployment rates on measures of value-added in manufacturing and services as a proportion of GNP. However, empirical estimates disagree about the role of these proxy measures (see row 10 of Table 4.1). 4.6.2

Knowledge spillovers and growth Some evidence is consistent with the knowledge spillover theory of entrepreneurship articulated in chapter 2. That theory predicts that entrepreneurs will locate close to sources of knowledge spillovers, forming new organisations in order to exploit them and thereby generating economic growth.

Evidence about the determinants of entrepreneurship

141

Specifically, there is evidence that knowledge- and technology-based new ventures tend to locate close to universities and corporate research laboratories.43 For example, Audretsch, Keilbach and Lehmann (2006) observe that publicly held German SMEs are attracted by access to universities’ supplies of graduates and their production of tacit knowledge. At the same time, proximity to universities increases the speed with which new ventures progress from start-ups to stock market listings. Although new ventures undertake negligible amounts of R&D themselves, they appear to be able to generate innovative output by exploiting knowledge spillovers generated by other institutions. Precisely how they do so is less clear, although Almeida and Kogut’s (1997) analysis of patent data from the US semiconductor industry reveals that ideas (and presumably knowledge spillovers) are often spread via the mobility of key scientific personnel. Founders can replicate or modify ideas encountered through previous employment (Bhide, 1994, p. 151). This kind of mobility is facilitated by high-tech firms clustering closely together, especially in industries where new economic knowledge plays a major role, such as biotechnology (Prevenzer, 1997). Mobility can however be hindered by legal systems that enforce post-employment non-compete covenants which restrict workers’ freedom to exploit elsewhere ideas derived from a previous spell of employment. Gilson (1999) partly attributes to laxer enforcement of non-compete provisions in California compared with Massachusetts the greater success of Silicon Valley as a high-tech cluster than Route 128. Some of the knowledge spillover theory’s implications appear to carry through to more aggregated industry and regional settings. Several researchers have found that new firm formation rates are higher in regions where higher proportions of adults hold college degrees and where local densities of firms and personnel are greater.44 This could be suggestive of human capital-based knowledge spillovers. Other evidence suggests that industries and regions with greater investments in knowledge creation (e.g. higher shares of the labour force employed as scientists or engineers) also possess aboveaverage venture start-up rates, especially those where small firms undertake a larger share of the investment.45 Stuart and Sorenson (2003b) report that IPOs of biotechnology firms located contiguous to or within a locality characterised by knowledge creation accelerate venture founding rates in that locality. It is interesting in this regard to speculate about the relationship between innovation and R&D. This is known to be weak or non-existent at the level of the individual firm, but becomes progressively stronger as the level of aggregation rises to that of the industry and nation (Audretsch, 2003). These findings can be reconciled if numerous small firms innovate without conducting R&D but instead exploit external (aggregate) knowledge spillovers, e.g. from universities. It has been observed that industries and patent classes associated with higher levels of university research exhibit higher average start-up rates (Audretsch and Acs, 1994; Shane, 2001). However, new venture creation is known to affect knowledge creation via patent activity (Acs and Varga, 2005) – implying reverse causality and possible endogeneity. Although self-reinforcing feedback from new venture creation to knowledge creation is consistent with a central tenet of the knowledge spillover theory,

142

Selection

it is troubling that the above findings might simply reflect the existence of greater entrepreneurial opportunities in some industries than in others, which stimulate both patents and new firm entry without there necessarily being a (knowledge spillover) link between them. The knowledge spillover concept has also been criticised by Breschi and Lissoni (2001, p. 976) as being ‘no more than a ‘black box”. Breschi and Lissoni’s (2001) literature survey led them to question the empirical veracity of the knowledge spillover concept, on the grounds that empirical tests of it are indirect (see above) and do not explore precisely the ways that knowledge is actually transferred among agents located in the same geographic area. Breschi and Lissoni (2001) also contend that what appears to be inadvertent knowledge spillovers might actually be well-regulated knowledge flows between universities and firms, or groups of firms, which have more to do with speeding up the development phases of new products and processes than inadvertently providing new innovation opportunities. Consistent with this critique, Bania et al. (1993) detected only a small effect of university research funding on the high-tech startup rate, while Zucker et al. (1998) identified the intellectual capital of star scientists rather than knowledge spillovers as the key factor affecting the location and timing of new biotechnology firms.

4.6.3

Entrepreneurship and the business cycle

In theory, the greater output that entrepreneurs produce in return for bearing greater risk implies that changes in the number of entrepreneurs can have a major impact on output, by amplifying aggregate shocks. That suggests a potentially important macroeconomic role for entrepreneurship. In this vein, Rampini (2004) proposed a risk-based reason why the number of entrepreneurs is likely to be pro-cyclical. When shocks to the economy are favourable, productivity and wealth in entrepreneurship increase, making agents more willing to bear risk (via DARA) and become entrepreneurs. In addition, anticipating greater returns in favourable states, entrepreneurs supply higher levels of effort, reducing moral hazard problems and making lenders more willing to fund risky investment projects. However, when shocks are unfavourable, the opposite process occurs: wealth, investment and entrepreneurship all decline. A dynamic externality inherent in innovation provides another reason why entrepreneurship and aggregate economic activity exhibit similar cycles over time. Radical innovations increase economic activity directly, and frequently indirectly create opportunities for other, subsequent innovations, further increasing opportunities for entrepreneurship and greater economic activity. Because entrepreneurs do not internalise this dynamic externality when making their decisions to innovate and invest, the result is excessive volatility and pro-cyclicality of innovation and economic growth (Barlevy, 2007). US evidence suggests that, at the aggregate level at least, new firm formation is indeed pro-cyclical, with venture formation rates and individual transitions into entrepreneurship rising in booms and falling in recessions.46 However, to set this in perspective,

Evidence about the determinants of entrepreneurship

143

pronounced state-dependence (i.e. persistence) in entrepreneurship at the individual level has been observed to dominate temporary or cyclical variations (Henley, 2004). And counter-cyclical forces may be at work too. It has been argued that recessions might have a ‘cleansing’effect, removing low-quality enterprises, and thereby increasing average entrepreneur quality (Caballero and Hammour, 1994). Offsetting this effect, though, wages fall in recessions, causing low-ability marginal types to enter entrepreneurship and reduce the average quality of the entrepreneur pool (Ghatak et al., 2007). This might explain the common emergence of worker co-operatives and other ‘marginal’ enterprises in recessions, which dissolve in economic recoveries when conventional employment opportunities become more readily available (Ben-Ner, 1988; Pérotin, 2006). The interest rate is an important macroeconomic variable which responds to the state of the business cycle, usually in a pro-cyclical manner. When economies are growing above-trend, interest rates tend to rise to control inflation. Obviously, higher interest rates increase the cost of financing and operating a business, leading to lower rates of entrepreneurship. On the other hand, if banks offer long-term loans at fixed interest rates, then entrepreneurship measured at time t might be relatively insensitive to interest rates also measured at time t, depending instead on interest rates at t − 1, t − 2, etc. Overall, though, UK and US time-series evidence suggests that interest rates do have a significant negative effect on entrepreneurship (see row 14 of Table 4.1).

4.6.4 Unemployment There is now an extensive literature on the relationship between entrepreneurship and unemployment. One of the motivations for studying this topic is the policy interest in promoting entrepreneurship as a way of reducing unemployment. There are two channels through which this could occur. First, there is the direct effect of removing a formerly unemployed person from the official unemployment register. Second, there is the indirect effect of eventual job creation by entrepreneurs who hire outside labour. Contrary to the impression given by some authors (e.g. Shane, 2003, chap. 4), there is actually no clear-cut theoretical or empirical relationship between entrepreneurship and unemployment. Unemployment may affect entrepreneurship in two opposite ways: via ‘recession-push’ and ‘prosperity-pull’ effects. According to the ‘recession-push’ hypothesis, unemployment reduces the opportunities of gaining paid employment and the expected gains from job search, which ‘pushes’ people into entrepreneurship. A secondary and complementary effect is that, as firms close down in recessions, the availability and affordability of second-hand capital equipment increases, reducing barriers to entry (Binks and Jennings, 1986). Both effects imply a positive relationship between entrepreneurship and unemployment. According to the ‘prosperity-pull’ hypothesis, on the other hand, at times of high unemployment the products and services of entrepreneurs face a lower market demand. This reduces incomes in entrepreneurship and possibly also the availability of capital, while increasing the risk of bankruptcy. Then individuals are ‘pulled’ out of

144

Selection

entrepreneurship. At the same time, entrepreneurship may become riskier because if the venture fails, it is less likely that the entrepreneur can fall back on a job in paid employment. Furthermore, at the level of the individual, unemployment could be associated with low levels of human and financial capital needed for successful entry into entrepreneurship. In contrast to the ‘recession-push’ hypothesis, these factors imply a negative relationship between entrepreneurship and unemployment. Empirical estimates of the entrepreneurship/unemployment relationship invariably confound the above two effects, capturing a ‘net’ effect. However, the size and direction of the net effect is still of policy interest. It is helpful to sort the large number of results on this topic by empirical method. Results obtained using cross-section binary choice models are reviewed first, followed by results from time-series and panel-data models. In the cases discussed below, entrepreneurship is measured as self-employment unless otherwise stated. Cross-section evidence In my previous book, I wrote that the bulk of crosssection evidence pointed to a negative relationship between entrepreneurship and unemployment (see Parker, 2004, Table 3.3, row 10, p. 104). It is striking that recent studies have detected more positive than negative relationships – culminating in the ‘muddier’ picture given in row 11 of Table 4.1. It is therefore no longer possible to conclude, as I did in Parker (2004), that ‘overall the econometric evidence from cross-section studies supports the ‘prosperity-pull’ hypothesis’. Why are the newer cross-section findings so different from the older ones? One reason might be that more recent research includes more controls for demand conditions, identifying the opportunity cost mechanism in the ‘recession-push’ argument more precisely. Another possible reason is that more recent research uses data from time periods where chronic unemployment is not observed, so confounding demand effects are more muted. Digging into the data reveals interesting information about the numbers and transitions from unemployment to entrepreneurship, and the impact of worker quality, unemployment experience and unemployment duration on these transitions. The following stylised facts describe transitions from unemployment to self-employment in developed countries:

• Most transitions into self-employment are made from paid employment, reflecting a greater stock of employees than unemployed (Storey, 1982); • Nevertheless, a higher proportion of unemployed than employed people transition to self-employment, being twice as likely to do so according to US evidence (Evans and Leighton, 1989b, 1990);47 • Most unemployed people remain unemployed or transition to paid employment rather than self-employment (Cowling and Taylor, 2001). Of the unemployed, those with unstable work histories (including periods of past unemployment) are significantly more likely to become entrepreneurs.48 In fact, it appears to be a history of job changes, rather than unemployment per se, which increases

Evidence about the determinants of entrepreneurship

145

the willingness of workers to become self-employed (van Praag and van Ophem, 1995). One possible interpretation is that many entrepreneurs are former workers who made low-quality job matches in paid employment. Evidence that unemployed people who become entrepreneurs experience a larger drop in their earnings than either the unemployed who return to wage work, or employees who enter entrepreneurship, bears out this idea (see Evans and Leighton, 1989b; Schiller and Crewson, 1997). So does evidence about the duration of unemployment spells, which are related to low-quality job matches and inferior skill sets. This evidence indicates that longer unemployment durations are associated with significantly higher probabilities of transitions into selfemployment.49 For example, Alba-Ramirez (1994) estimates that a 1 per cent increase in unemployment duration increases the probability of a switch to self-employment by 0.15 percentage points in the USA and by 0.17 percentage points in Spain. Cowling and Mitchell (2001) argue that unemployment duration rather than the stock of unemployed is what drives changes in aggregate rates of entrepreneurship. In contrast, there is no reason to believe that employee quits are systematically related to worker quality. According to US evidence, employees who lost their jobs in the previous three years (‘job losers’) are significantly less likely to become entrepreneurs (by 3 percentage points) than ‘non-losers’ – even after controlling for observable characteristics.50 It is useful to distinguish between voluntary quits on one hand and involuntary layoffs and redundancy on the other. The latter are more common for low-skilled workers and are associated with higher rates of entrepreneurship.51 Nevertheless, redundancies actually result in more transitions to paid employment than self-employment; and few new firms set up by redundant workers hire many external workers.52 Finally, there can also be a gender dimension to unemployment–self-employment transitions. Kuhn and Schuetze (2001) report that most of the increase in female Canadian self-employment in the 1980s and 1990s was attributable to an increase in retention rates in self-employment owing to better work prospects for women, whereas for men most of the increase in self-employment was attributable to a decrease in stability in paid employment and inflows from unemployment, owing to deteriorating opportunities for them. Time-series and panel-data evidence In contrast to results from cross-section studies, time-series studies tell a relatively consistent story. As row 11 of Table 4.1 shows, the overwhelming majority of time-series studies uncover significant positive effects of national and regional unemployment rates on national self-employment and new firm formation rates.53 This appears to support the ‘recession-push’ hypothesis. Most researchers estimate a linear relationship between measures of entrepreneurship and unemployment rates, although a few have explored the possibility of non-linearity (Hamilton, 1989; Georgellis and Wall, 2000). Just like cross-section analyses, time-series data have been used to link job layoffs and redundancy payments with regional and national rates of entrepreneurship.54 Paneldata evidence also supports the ‘recession-push’ hypothesis for relatively advantaged

146

Selection

workers possessing advanced academic qualifications, as well as for people with self-employed parents.55 Two types of panel study can be distinguished: one utilising a relatively large cross-section dimension and one utilising a relatively large time-series dimension.56 Panel-data estimates of the entrepreneurship–unemployment relationship generate mixed results, which vary across countries (Blanchflower, 2000). Unfortunately, significant coefficients are sometimes observed to vanish once other covariates are added to econometric specifications (Henley, 2004; Parker and Robson, 2004). As always, causality is a tricky issue. Thurik et al. (2008) point out that in a dynamic setting, an increase in the unemployment rate is likely to feed into higher rates of entrepreneurship later on (‘recession-push’). Yet higher rates of entrepreneurship are likely to generate more benign economic conditions which serve to reduce subsequent unemployment rates (termed the ‘entrepreneurial effect’). Thurik et al. (2008) embed both of these causal mechanisms in a vector autoregression (VAR) model and detect evidence of both mechanisms using data from twenty-three OECD countries over 1974–2002. The entrepreneurial effect apparently dominates the ‘recession-push’ effect, suggesting that entrepreneurship increases employment growth overall. More will be said on this issue in chapter 11. Conclusion Greater use of cross-section data on transitions into entrepreneurship from unemployment is beginning to generate findings which complement those from time-series data. The upshot is a growing consensus about a zero or positive relationship between unemployment and entrepreneurship. This contrasts with the pre-2000 literature where cross-section studies usually suggested a negative relationship, in contrast to positive relationships identified in time-series studies – and which prompted several researchers to propose sometimes ingenious ways of reconciling these disparate findings.57 It should be borne in mind, however, that the unemployment rate is only an imperfect proxy for the underlying factors inducing workers to enter, or leave, entrepreneurship. Undergraduate macroeconomics students learn that exogenous technological progress need not cause unemployment if wages are sufficiently flexible and workers are sufficiently adaptable to switch from declining sectors into those growing as a direct or indirect result of technological progress. Thus if technological change is the cause of labour-saving changes in production methods, unemployment is only the symptom of slow adjustments in the labour market. If changing costs or technologies make entrepreneurship a more efficient and attractive form of productive organisation, one would expect to see transitions between paid employment and entrepreneurship without necessarily observing any impact via unemployment. So there is really no economic reason why unemployment and entrepreneurship need to be related at all. The extent to which they are related probably reflects rigidities in the economy. These rigidities are likely to diminish over time as governments adjust their tax-benefit systems to make their economies more flexible. Chapter 17 explores some of the policy issues surrounding the entrepreneurship–unemployment nexus.

Evidence about the determinants of entrepreneurship

147

Table 4.4. Self-employment rates in the British regions, 1970 and 2000

Region North Yorkshire & Humberside East Midlands East Anglia South East South West West Midlands North West Wales Scotland

1970 rate (%)

2000 rate (%)

Rank in 1970

Rank in 2000

Human capital 2000 (%)

6.02

7.96

8

10

9.1

6.45 7.00 10.11 7.28 11.71 5.74 6.54 10.37 5.68

10.42 10.71 13.35 13.47 14.98 10.29 9.87 12.43 9.31

7 5 3 4 1 9 6 2 10

6 5 3 2 1 7 8 4 9

12.2 13.1 14.5 17.8 14.6 12.9 12.8 12.1 12.9

Note: The self-employment rate is defined as the number of self-employed jobs (male plus female) in a region divided by the region’s labour force. ‘Human capital’ is defined as the percentage of the economically active workforce in a region with a degree-level qualification. Source: Parker (2005b).

Research in the economics of entrepreneurship over the last decade has largely shifted away from the unemployment issue towards other topics. Partly this reflects the end of the era of high unemployment rates in many developed economies, which peaked in the 1980s. But it might also reflect growing recognition among entrepreneurship scholars that there is relatively little more to say about this issue which is either original or exciting (Parker, 2006b, p. 455). 4.6.5

Regional factors

All major economies exhibit regional differences in rates of entrepreneurship. This is true for administrative regions, labour market areas, cities and neighbourhoods. The most entrepreneurial regions within a country can have new firm formation rates which are between two and four times higher than those observed in the least entrepreneurial regions (Reynolds et al., 1994). As Table 4.4 shows for the case of Britain, regional differences are not only pronounced but are also persistent. The Spearman correlation coefficient for Table 4.4’s rank orderings of British self-employment rates in 1970 compared with 2000 is 0.87. Similar patterns have been observed for regional rates of new firm formation in the USA.58 Why are there regional differences in entrepreneurship and why do they persist? I will consider two broad theoretical perspectives below. The first attributes persistent differences in regional rates of entrepreneurship to persistent differences in regional

148

Selection

characteristics conducive to entrepreneurship. Because this argument has been shown to have only limited explanatory power, recent research has shifted towards explaining a second set of alternative explanations based on externalities and multiple (regional) equilibria. Regions experience different economic conditions, so it is natural to propose these as the causes of regional differences in rates of entrepreneurship. Regions with low levels of demand, poorly educated workers, or high concentrations of capital-intensive industries (which can sustain effective barriers to entry) are less attractive for entrepreneurs to enter. At the other end of the spectrum, with respect to high-tech entrepreneurship venture capitalists’ expertise tends to be concentrated within tight geographic and industry spaces (Sorenson and Stuart, 2001). Previous research shows that the highest new firm formation rates are found in regions with the highest proportions of employment in small firms, despite the risk that colocation enables technologically similar firms to expropriate the ideas of new entrants (Reynolds et al., 1994; Rosenthal and Strange, 2003). Other important determinants of regional new firm formation rates, highlighted in the now classic study of Reynolds et al. (1994), include high rates of in-migration; rapid growth in regional incomes and populations; and high levels of employment specialisation and population densities. These findings have been generally confirmed in subsequent research.59 Interestingly, local government expenditures and assistance programmes appear to have only limited effects on regional firm birth rates (Reynolds et al., 1994). Other factors associated with high levels of regional entrepreneurship include housing wealth (Robson, 1998a); plenty of college drop-outs and well-educated, older citizens (Acs andArmington, 2006; Glaeser, 2007); industrial structures characterised by low capital intensities and factor costs (Fritsch and Falck, 2007); and plentiful access to finance (Naudé et al., 2008). While this approach of identifying different regional characteristics has generated some interesting insights, it suffers from both conceptual and empirical drawbacks. On the conceptual front, less entrepreneurial regions tend to have low stocks of human capital (Table 4.4) and hence wages. Hence one might expect entrepreneurs would seize profitable opportunities by moving from expensive high-entrepreneurship regions to start new firms in low-entrepreneurship regions where there is less competition and a lower cost base. That this does not generally happen implies that some other factors are at work. Indeed, American and Italian entrepreneurs turn out to be less geographically mobile than employees are, especially in more financially developed regions (Michelacci and Silva, 2007).60 One plausible explanation is that local entrepreneurs have greater access to finance than non-locals do, and benefit from ‘home bias’exhibited by investors located in their region (Parwada, 2008). Another is that most entrepreneurs operate businesses whose main customer base is in the immediate or close locality. It is noteworthy that geographical mobility weakens the intergenerational transmission of self-employment status (Niittykangas and Tervo, 2005) – which was identified earlier in this chapter as one of the chief determinants of participation in entrepreneurship. A major drawback of empirical studies linking relative start-up rates to regional conditions is that they typically explain no more than a fraction of actual regional variations

Evidence about the determinants of entrepreneurship

149

in entrepreneurship. This signals the existence of omitted explanatory variables. Some of these variables, such as ‘local culture’, social norms and institutional restrictions, might be hard to measure accurately and comprehensively. Recognising these limitations, recent research has sought other explanations for persistent differences in regional rates of entrepreneurship based on the notions of externalities and multiple regional equilibria.61 Externalities can come in several forms: 1. Information spillovers and social ties. Entrepreneurs create information spillovers by signalling information about opportunities and resource requirements to other latent entrepreneurs through social networks (Stuart and Sorenson, 2003a; Rocha and Sternberg, 2005). This information can help to: resolve ambiguities inherent in new venture creation, so reducing uncertainty; advertise role models which increase awareness of perceived net benefits from entrepreneurship; and perpetuate favourable social norms which legitimise entrepreneurship, affecting non-pecuniary aspects of occupational choice (Giannetti and Simonov, 2004). 2. Knowledge spillovers. These tend to be spatially concentrated, partly because the costs of transmitting tacit knowledge increase with distance. Hence entrepreneurs exploit knowledge spillovers locally. This can explain the well-known tendency of spin-offs to locate close to their geographic roots, which in turn might explain industry clustering, without needing to appeal to alternative explanations such as agglomeration economies (Klepper, 2006). 3. Intergenerational transmission. As noted earlier in this chapter, an important influence on entrepreneurship is intergenerational transfers of information, knowledge and attitudes from self-employed parents to their offspring. Also noted earlier is the fact that entrepreneurs tend to be less geographically mobile than entrepreneurs. Hence geographical family immobility can perpetuate regional rates of entrepreneurship. 4. Agglomeration benefits. A geographic concentration of ventures reduces average transaction costs for each entrepreneur while facilitating efficient and rapid transfers of technical knowledge and skills. It also encourages the development of specialised services (e.g. venture capital and corporate legal expertise) which in turn encourages new firms to locate nearby. Some of the entities which service new ventures are themselves new ventures, leading to a regional ‘multiplier’ effect (Johnson and Parker, 1994, 1996; Nyström, 2007). These reasons can explain why different regions can have persistently different rates of entrepreneurship. They also help explain pronounced agglomerations of entrepreneurial activity (‘clusters’) such as Silicon Valley and Boston’s Route 128 in the USA, and Baden-Württemburg and Emilia-Romagna in Europe. These clusters no doubt also benefit from some or all of the following favourable factors: a benign regulatory regime; the proximity of advanced research universities and institutes that are well connected to industry; the availability of a flexible and mobile workforce; mechanisms for maintaining global linkages; access to venture capital; and formal as well

150

Selection

as informal associations (networks) which leverage social capital and foster collective learning for the whole cluster.62 Conversely other regions can promote deep-rooted attitudes inimical to entrepreneurship, especially in areas dominated by large organisations (Jackson and Rodney, 1994; van Stel and Storey, 2004). However, the relative importance of these factors individually is still not precisely known. As a result, appropriate recommendations for public policy to create successful new clusters remain imperfectly developed. Another possibility is that regional differences in entrepreneurship are the manifestation of multiple equilibria (see chapter 2). In this regard, Parker’s (2005b) model, which links high average levels of regional entrepreneurship with high levels of education, at least appears to be consistent with the evidence (see the final column of Table 4.4 and Georgellis and Wall, 2000). At a more disaggregated level, some research compares urban with regional locations. Urban areas enjoy lower communication and transaction costs with respect to customers and suppliers and are associated with larger, denser markets with higher average incomes – enabling entrepreneurs to reap scale economies and exploit opportunities which might otherwise not be cost-effective. Factors of production, knowledge spillovers and specialised services are also likely to be more abundant in urban areas. On the other hand, inputs such as land and labour can be more expensive in urban areas; and there are often fewer paid employment opportunities in rural areas, increasing the relative attractiveness of new firm creation there. Perhaps reflecting these offsetting effects, estimates from binary choice models reveal a mixed picture about the effects of urban location on entrepreneurship (see row 12 of Table 4.1). Nevertheless, urban ventures appear to generate the highest average entrepreneurial earnings and new firm growth rates.63 There also tend to be more transitions in and out of self-employment in rural than in urban locations, possibly because the scarcity of paid employment opportunities in rural areas makes self-employment an unstable form of ‘employment of last resort’ (Tervo, 2008). Of course, not all urban areas are the same. Lee et al. (2006) argue that urban areas with greater social diversity have lower entry barriers and so attract creative individuals. Duranton and Puga (2001) and Florida (2003) argue that diverse mixes of creative people are more likely to generate new and novel combinations, which are associated with the growth of cities and high-value entrepreneurial innovations. However, the available evidence comprises statistical associations rather than causal effects. Glaeser (2007) argues that local entrepreneurship can explain why some cities grow faster than others (rather than vice versa), with high self-employment rates in cities in 1970 predicting sustained growth in incomes and populations of those cities over the following thirty years. Furthermore, numerous small local firms correlate with subsequent employment growth in a city. Glaeser (2007) concludes that more entrepreneurial cities are more successful cities. In summary, research on the geography of entrepreneurship in general and new venture creation in particular continues to develop. Part of the attractiveness of this topic is the way it connects entrepreneurship to several other prominent issues, including innovation, human capital, spatial structure, growth-enhancing spillovers and public

Evidence about the determinants of entrepreneurship

151

policy. But much more research remains to be done in terms of both theory development and the empirical testing of theories. It seems that many hard-to-observe region-specific and individual-specific factors affect regional levels of entrepreneurship. An important but challenging task for empirical researchers is to operationalise precise measures of the various factors at work and to distinguish sharply between their causal effects. 4.7

Nascent entrepreneurship

Do the findings presented so far in this chapter also apply to nascent entrepreneurs (NEs)? As noted in chapter 1, if different factors affect entry compared with survival, then the bulk of studies measuring entrepreneurship in terms of established business owners or self-employees might give a misleading or partial impression of the factors affecting entry into entrepreneurship. This section provides a brief overview of the determinants of nascent entrepreneurship. To anticipate one of the main findings, it turns out that most explanatory variables have similar effects on nascent entrepreneurship as on established entrepreneurship (which have been discussed in earlier sections of this chapter). There are several exceptions, however, which raise some interesting questions about early entry and survival processes for some population groups. This section is divided into two parts. In the first one, characteristics of NEs are discussed and the findings are compared with those from earlier in the chapter and elsewhere in the book. The second part asks what happens to NEs as they develop their new ventures. 4.7.1

Characteristics of nascent entrepreneurs For the most part, human capital variables affect individuals’ propensities to engage in nascent entrepreneurship in a similar way to their engagement in ‘established’ entrepreneurship as discussed earlier in the chapter. This includes education, which has a significant positive effect on propensities to become an NE in some countries (e.g. Sweden and the USA) but more ambiguous effects in others (e.g. Germany).64 Also consistent with evidence from earlier in the chapter, prior business and work experience have generally positive effects on the likelihood of becoming an NE, especially the former.65 Wagner (2006a) shows that diversity of experience is associated with NE status, in line with Lazear’s ‘jack-of-all-trades’ hypothesis. In all developed countries, men are much more likely to become NEs than women are, consistent with other definitions of entrepreneurs (see chapter 6). Also consistent with evidence about ‘established’ entrepreneurs, urban density and population growth rates are positively associated with nascent entrepreneurship.66 And reflecting the mixed effects of unemployment on entrepreneurship, some studies have detected positive effects (e.g. Rotefoss and Kolvereid, 2005), while others have reported negative (Bergmann and Sternberg, 2007) and insignificant (Delmar and Davidsson, 2000) effects. Differences start to creep in when one considers family background variables. In contrast with row 6 of Table 4.1, for example, marital status has more mixed effects on propensities to become an NE.67 And unlike row 9 of Table 4.1, having self-employed

152

Selection

parents or knowing other entrepreneurs have positive associations with NE status in European and cross-national studies, but not in US studies.68 Other differences between the determinants of nascent and established entrepreneurship relate to the effects of age, wealth and ethnicity. As observed earlier, annual entry rates into self-employment are roughly independent of age (e.g. Evans and Leighton, 1989a). In contrast, the evidence suggests that NE propensities decline with age.69 It is unclear why this is so, although annual household survey data might miss many short-lived and possibly unsuccessful founding efforts made by young NEs, leading to an overly ‘flat’ age-entry profile. Regarding wealth, chapter 9 outlines evidence of an apparently strong positive relationship between this variable and established entrepreneurship (see row 15 of Table 4.1). Interestingly, this relationship tends not to be observed for NEs.70 Taken at face value, these findings of wealth not mattering for NEs might reflect modest initial capital needs at start-up, a well-functioning credit market, and/or bootstrapping by nascent entrepreneurs (Kim et al., 2006). The latter includes reliance on internal funds, borrowing from ‘friends, family and fools’, and low-cost acquisition of material resources (e.g. home-based starts economising on office costs: Harrison et al., 2004). Also, samples of established entrepreneurs are prone to over-sample the ‘successful few’, whom one would expect to be wealthier. This could be an important source of survival bias, casting some doubt on assertions that limited wealth and borrowing constraints hamper start-up efforts. As we will see in chapter 5, self-employment rates of white Americans are some three times greater than those of black Americans. So it is very striking that PSED and GEM data both show black Americans to be nearly twice as likely to engage in nascent entrepreneurship as whites.71 In fact, Parker and Belghitar (2006) show that black Americans are significantly less likely to convert start-up efforts into actual start-ups. Instead they remain NEs for longer than white Americans do. Similar patterns in the more recent PSED II data set show that this is a robust feature of contemporaryAmerican entrepreneurship. It suggests that barriers to black entrepreneurship in America lie not in a lack of interest in starting a business, but in problems of seeing the process through to the successful launch of new ventures. Precisely what these problems are and how they can be overcome are important questions for future research to address. To conclude, broadly similar factors seem to affect selection into NE status as into ‘established’ entrepreneurship. This suggests that, apart from one or two exceptions noted above, hindsight and survival biases pertaining to established entrepreneurs are unlikely to vitiate the main body of findings reported in this chapter, thereby ‘saving the entrepreneurship research community from a need for major reinterpretation of what it thought was known’ (Davidsson, 2006, p. 12).

4.7.2

Venture development paths of nascent entrepreneurs I now ask what explains the hard performance outcomes of NEs (as opposed to ‘soft’ measures such as self-reported declarations of realising intended or expected goals).72 The rationale for ‘intentions’ studies is that intentions translate into actions, including

Evidence about the determinants of entrepreneurship

153

business entry. However, these studies rarely discriminate between ‘dreamers’ and ‘doers’; and psychological research as well as longitudinal analyses of follow-through behaviour repeatedly show that intentions are not always related to actual behaviour.73 This might be especially pertinent given that approximately 10 per cent of the PSED sample of NEs claimed to be thinking about starting a business a decade or more prior to the first PSED interview. Broadly speaking, researchers measure venture development in two distinct ways. The first approach records the passing of various activity milestones during the start-up process. The second records whether an NE has started up, abandoned their venture, or is still an NE a year or more after the initial survey. A problem with the first approach is that specific start-up processes can follow almost any sequence – including even claiming to have had a first sale before having thought seriously about starting a business (Carter et al., 1996; Reynolds, 1997). Partly for that reason, studies taking the first approach usually augment measures of process development with hard measures of performance. While the second approach circumvents this problem, different start-up organising efforts are caught at different stages, with over-sampling of efforts which take a long time to gestate (Davidsson, 2006, p. 19). This can usually be dealt with by including a control variable measuring the length of the start-up organising process to date. It is sometimes claimed that written business plans promote the development and survival of new ventures (Delmar and Shane, 2004; Liao and Gartner, 2006). One reason they might do so is that a business plan which establishes a venture as a legal entity ‘legitimises’ the venture in the eyes of outside stakeholders (Delmar and Shane, 2003, 2004). Alternatively, however, rational entrepreneurs with better projects are more likely to invest time and resources in business plans because they are more likely to pay off. Indeed, Bhide (2000) points out that for many entrepreneurs the expected costs of formal planning outweigh the expected benefits. Forty-one per cent of the Inc. 500 founders interviewed by Bhide (2000) started without any business plan at all. According to Newbert (2005) and Parker and Belghitar (2006), business plans do not help predict the venture outcomes of NEs. Based on a review of previous studies, Table 4.5 chronicles what happens to NEs approximately one year after they were first identified as such. Although there is some variation from study to study, between one-third and one-half of nascent entrepreneurs launch their venture within one year. The largest sample of those in the table is drawn from the PSED: according to these data, continuation in nascent entrepreneurship is the modal outcome in the USA (Parker and Belghitar, 2006). This implies that for nearly one-half of NEs, preparing before launching a start-up takes more than the one year proposed by Reynolds and White (1997). In terms of milestones achieved, Carter et al. (1996) claim that American NEs who either start up or abandon their nascent ventures are more similar to each other than are NEs who are still trying. According to Diochon et al. (2003), Canadian NEs abandon because they want to, not because they have to. Alsos and Kolvereid (1998) identified stark differences between outcomes of different types of Norwegian nascent

154

Selection

Table 4.5. What happens to nascent entrepreneurs Study

Still Nascent

Operating

Gave Up

Total

Country

Carter et al. (1996)a Alsos and Ljunggren (1998) van Gelderen et al. (2001) Diochon et al. (2003)b Parker and Belghitar (2006)

21 (30%) 37 (25%) 89 (27%) 51 (39%) 159 (47%)

34 (48%) 68 (46%) 155 (47%) 45 (34%) 112 (33%)

16 (22%) 43 (29%) 86 (26%) 36 (27%) 69 (20%)

71 148 330 132 340

USA Norway Netherlands Canada USA

Note: Numbers of cases appear as the first cell entries, with sample percentages in parentheses. Data based on destinations of nascent entrepreneurs approximately one year after initial interview, except for a (6–18 months). b Excludes nineteen cases who could not be contacted after the first wave.

entrepreneur, with only 4 per cent of portfolio NEs giving up after one year, compared with 41 per cent of serial and 32 per cent of novice NEs. Parker and Belghitar (2006) argue that the value of waiting plays a key role in predicting NE outcomes. They propose several characteristics which make waiting an optimal strategy. For example, blacks facing discrimination in the credit market (see chapter 5) who cannot borrow enough to launch their ventures immediately might optimally defer launching and wait instead, saving to self-finance well-resourced entry at a later date (see also Parker, 2000). Membership of majority ethnic groups in contrast might be associated with more valuable outside options, promoting quits from NE to paid employment. Other factors in contrast might promote early starts rather than waiting, e.g. if NEs have already received money from customers. Parker and Belghitar (2006) performed a multinomial logit analysis of the three outcomes listed in Table 4.5, measured one year after the initial interview, conditioning on a set of exogenous and objectively verifiable covariates observed at the initial interview. They also corrected for possible attrition bias. Parker and Belghitar (2006) found that NEs are significantly more likely to transition to start-up if they have established credit with suppliers or have already received some money from customers. This is consistent with independent evidence that NEs who pass important ‘milestones’ are more likely to complete the start-up process.74 It is also consistent with evidence that individuals who have become self-employed while preparing their venture for launch are more likely ultimately to start up (van Gelderen et al., 2001; Rotefoss and Kolvereid, 2005). Parker and Belghitar (2006) further reported that transition rates to start-up are higher for NEs who own their own home, are household heads and who are preparing solo ventures (see also Kessler and Frank, 2004, for confirmatory Austrian evidence). In addition, posthigh school education and NE involvement in public- and private-sponsored business assistance programmes significantly increase the probability of transitions to start-ups. Other researchers in contrast have reported insignificant or negative effects on startup propensities from higher levels of education (see Davidsson and Honig, 2003; and Reynolds and White, 1997, chap. 4, respectively).

Evidence about the determinants of entrepreneurship

155

Although wealth does not play a significant role, NEs who save and self-finance their ventures, especially non-whites, are significantly more likely to continue as NEs (Parker and Belghitar, 2006; van Gelderen et al., 2006). This apparently confirms the value of waiting, which enables the necessary resources for start-up to be acquired. Parker and Belghitar (2006) also found that white American NEs are significantly more likely to quit than other ethnic groups. Similar to Davidsson and Honig (2003), Parker and Belghitar (2006) observed no significant differences in NE outcomes by gender, although others have found males to be more likely to start up than females.75 Davidsson and Honig (2003) estimated that greater human and social capital increases the probability that individuals remain as NEs, while only social capital affects their financial performance. For their part, Parker and Belghitar (2006) found insignificant effects on outcomes from two human capital variables, age and experience; and of their social capital variables only parental self-employment had any effect, serving to increase the probability of continuation as an NE. However, two other studies have found positive effects of industry experience on start-up outcomes (van Gelderen et al., 2001; Rotefoss and Kolvereid, 2005). There is mixed evidence about the effects of business plans and attendance at business education classes on transitions to full start-ups.76 Finally, Koellinger (2008) studied 9,549 NEs from thirty GEM countries over 2002–04 and reported that formal educational attainment, unemployment and a high degree of self-confidence are significantly associated with entrepreneurial innovativeness at the individual level. 4.8

Dependent starts and firm characteristics

This section provides evidence about entrepreneurial ‘spawning’ corresponding to the theories outlined in chapter 2. Salient characteristics of spin-offs are discussed first, followed by aspects of spin-off performance. I will not review evidence about the determinants or performance of corporate ventures or university spin-offs, which are different categories of business venturing with different imperatives.77 A growing body of research links the characteristics and performance of new hightech start-ups to the characteristics and performance of the ‘parent’ firms from which they emerged. Evidence shows that the parent companies of new American VC-backed start-ups tend to be VC-financed themselves and are disproportionately located in hightech clusters such as Silicon Valley or Route 128 in Massachusetts (Gompers et al., 2005; Klepper and Sleeper, 2005). The availability of networks which facilitate entrepreneurship in these clusters might help to explain these findings (for a suggestive descriptive analysis, see Saxenian, 1994). Parent companies which are young or which have slowing growth rates spawn more new ventures than average, as do companies which focus in one market segment rather than several simultaneously (Gompers et al., 2005; Klepper, 2006). The evidence relating to the effects of parent firm size on spawning propensities on the other hand is mixed.78 On the whole, successful parent firms which enjoy long periods of ‘product leadership’ are characterised by higher spin-off rates, and spawn new ventures with above-average survival rates (Klepper, 2006). This suggests that

156

Selection

spin-offs transfer capabilities from the parent firm to their own organisation (Helfat and Lieberman, 2002). The agency and employee learning theories of spin-offs described in chapter 2 generate different predictions about the markets which spin-offs are likely to enter. An implication of agency theory is that spin-offs can minimise conflicts with their parents by entering (or creating) different markets from those of their parents. In contrast, an implication of learning theories is that spin-offs will utilise their knowledge by entering similar markets to the parent, but differentiated somewhat to reduce the threat of retaliation. The evidence on this issue partially supports the learning perspective. For example, high-tech and high-growth start-ups in the laser industry often develop similar products to parents, rather than entering completely new markets; but the markets targeted by these start-ups tend to be only peripherally related to the core business of their parents, limiting the competitive impact on the latter.79 In other cases (e.g. the hard-disk drive industry) spin-offs have entered new sub-markets before their parents and approximately one-half of them pursued a distinctive innovation that opened up a new sub-market (Christiansen, 1993). Overall, though, the strategy of entering similar markets or selling similar products to their parents is associated with above-average growth performance (Cooper and Bruno, 1977; Feeser and Willard, 1990). Asset complementarity also bears on the question of whether spin-offs co-operate or compete with their parent firms. Examples of co-operation include licensing, strategic alliances and outright acquisitions. Agency theories predict that co-operation is easier: when IPR protection is strong, since that enhances credibility and transparency in negotiation; when the cost of contracting between the two parties is low; and when complementarity in asset ownership between the parties is high. Using a novel data set on the commercialisation strategies of 118 US start-up projects, Gans et al. (2002) found support for these predictions, with IPR protection (measured in terms of patent ownership) and asset complementarity both significantly increasing the probability of the co-operative outcomes listed above. Having a VC relationship is associated with networks which reduce contracting costs, serving to make co-operative outcomes more likely to emerge than competitive ones. Gans et al. (2002) also claimed that innovations pioneered by new ventures in biotechnology more commonly occur in co-operation with incumbents, whereas in the electronics industry they occur more commonly in competition with incumbents.80 Spinoffs outperform other kinds of start-up in several respects, including their survival prospects, speed of product development, employment creation and growth performance.81 As noted earlier in this chapter, spin-off founders can enhance the performance prospects of their ventures by forming large teams which encapsulate broad and deep pools of knowledge.82 Klepper (2001) argues that few spinoffs emerge from unresolved agency conflicts in settings where employees hide discoveries from managers. Instead, most of them are triggered by adverse events at parent firms such as involuntary exits, acquisitions or downsizing imperatives. Employees generally try to work hard for their parent firm

Evidence about the determinants of entrepreneurship

157

and usually only resort to an independent start-up when intrapreneurship opportunities are denied to them (Garvin, 1983; Klepper, 2007). However, more data are needed to flesh out the evidence base on this and related issues about spinoffs. 4.9

Conclusion

This chapter presented an extensive body of evidence about the determinants of entrepreneurship. For convenience, a variety of practical definitions of entrepreneurship were used, including self-employment, business ownership and new venture creation. Either consciously or not, most of the studies from which the evidence has been drawn are underpinned by the theories discussed in chapter 2. This is especially true of occupational choice models based on heterogeneous endowments of entrepreneurial ability and risk-taking propensities, which have been operationalised using the binary choice models set out in chapter 3. So, for example, several of the human capital, social capital and psychological variables discussed above can be thought of as proxies for the theoretical constructs of Lucas (1978) and Kihlstrom and Laffont (1979). And empirical studies of entrepreneurial spawning are predicated on various theories of spin-offs presented in chapter 2. We now know a fair amount about the drivers of the entrepreneurial selection process – the focus of this chapter. Table 4.1 demonstrates that a broad consensus has been reached on the impact of many – although not all – of the principal explanatory variables. The most important determinants of entrepreneurship are age, labour market experience, marital status and having a parent who was an entrepreneur. Greater levels of risk and higher interest rates generally have negative effects on participation in entrepreneurship, although market risk and risk attitudes remain poorly measured. More generally, data limitations frequently force researchers to use crude proxies for the theoretical constructs outlined in chapter 2. In future work, finer-grained and more theoretically grounded data are therefore needed to consolidate and extend our knowledge about the individual and environmental determinants of entrepreneurship. Notes 1. References augment the studies cited in Parker (2004) with the following: Borjas (1986); Wong (1986); Macpherson (1988); Evans and Leighton (1989b, 1990); Butler and Herring (1991); Alba-Ramirez (1994); Audretsch and Fritsch (1994a); Brynjolfsson et al. (1994); Davidsson et al. (1994); Guesnier (1994); Robinson and Sexton (1994); Bates (1995); Eisenhauer (1995); Grant (1996); Lindh and Ohlsson (1996); Magnac and Robin (1996); Sanders and Nee (1996); Schiller and Crewson (1997); Garen (1998); Honig (1998); Briscoe et al. (2000); Clain (2000); Delmar and Davidsson (2000); Kangasharju (2000); Taylor (2001); McManus (2000); Armington and Acs (2002); Burke et al. (2002); Carree et al. (2002); Ritsila and Tervo (2002); Taniguchi (2002); Davidsson and Honig (2003); Fairlie and Meyer (2003); Fan and White (2003); Acs and Armington (2004a, 2006);Amossé and Goux (2004);Arum and Müller (2004 - all chapters); Blanchflower (2004); Cowling et al. (2004); Giannetti and Simonov (2004); Guiso et al. (2004); Henley (2004); Perry and Rosen (2004); Stabile (2004); Taylor (2004); Puri and Robinson (2005); Dobrev and Barnett (2005); Brown, Farrell and Sessions (2006); Ekelund et al. (2005); Niittykangas and

158

2. 3. 4.

5. 6.

7. 8.

9. 10. 11.

12. 13.

14.

Selection Tervo (2005); Rocha and Sternberg (2005); Beugelsdijk and Noorderhaven (2005); Kanniainen and Vesala (2005); Lunn and Steen (2005); Torrini (2005); Co et al. (2005); Georgellis et al. (2005); van Tubergen (2005); Böheim and Muehlberger (2006); Bruce and Mohsin (2006); Constant and Shachmurove (2006); Constant and Zimmerman (2006); Budig (2006); de Clercq and Arenius (2006); Cowling (2000); Brown, Farrell, Harris and Sessions (2006); Djankov et al. (2006); Fairlie (2006a); Fraser and Greene (2006); van Gelderen et al. (2006); Georgellis and Wall (2006); Grilo and Irigoyen (2006); Hammarstedt (2006); Hundley (2006); Kan and Tsai (2006); Koellinger and Minniti (2006); Mesnard and Ravillion (2006); Mora and Dávila (2006); Tamásy (2006); Tervo (2006); Verheul et al. (2006); Wellington (2006); Wu (2006); Zhang et al. (2006); Andersson and Wadensjoe (2007); Constant et al. (2007); Fairchild (2008); Fritsch and Falck, 2007); Fritsch and Mueller (2007); Glaeser (2007); Goeggel et al. (2007); Mohapatra et al. (2007); Wennekers et al. (2005); White et al. (2007); Colombier and Masclet (2008); Parker (2008a); Thurik et al. (2008). See Eisenhauer (1995), Dobrev and Barnett (2005) and Wu and Knott (2006) for examples of that approach. See Creedy and Johnson (1983), Foti and Vivarelli (1994), Parker (1996), Cowling and Mitchell (1997), Audretsch and Vivarelli (1997), Goedhuys and Sleuwaegen (2000) and Lofstrom (2002). KPMG (1999) observed that ‘lifestyle’motives for self-employment were strongest among founding entrepreneurs and weakest for those operating growing and innovating firms, who were more likely to stress rapid further growth as an objective. However, Burke et al. (2002) claim that entrepreneurs who value being their own boss perform significantly better than those who do not. VandenHeuvel and Wooden (1997), Blanchflower (2004), Ajayi-obe and Parker (2005). See Katz (1993), Blanchflower and Freeman (1994), Blanchflower (2000), Blanchflower et al. (2001), Hundley (2001b), Kawaguchi (2003), Blanchflower (2004), Benz and Frey (2004, 2008), Taylor (2004), Schjoedt and Shaver (2007) and Kawaguchi (2008). Sanfey and Teksoz (2007) report similar findings for people in transition economies. See Benz and Frey (2004, 2008), Taylor (2004) and Kawaguchi (2008). Benz and Frey (2008) also reported that employees of small firms (with smaller hierarchies) enjoy greater job satisfaction than employees of large firms. This might partly explain why small-firm employees are willing to earn less on average than workers in large firms (see chapter 10). An untested prediction of this finding might be that managers at the top of a hierarchy should have similar satisfaction levels to entrepreneurs. See Parasuraman and Simmers (2001), Ajayi-obe and Parker (2005) and Schieman et al. (2006). See Cowling (2000), Reynolds et al. (2002) and Williams (2004) for cross-country evidence. Consistent with this, the median age of MIT alumni who start new ventures has declined from 40 in the 1950s to 30 in the 1990s (Hsu, Roberts and Eesley, 2007). Nevertheless, in the population as a whole, entrepreneurs of both sexes remain older on average than employees (Aronson, 1991). See also Carroll and Mosakowski (1987), van Praag and van Ophem (1995), Quadrini (1999), Lin et al. (2000) and Davidsson and Honig (2003). For supporting evidence, see Evans and Leighton (1989b), Dolton and Makepeace (1990), Blanchflower and Oswald (1990), Taniguchi (2002), Dobrev and Barnett (2005) and Djankov et al. (2006). C.f. Evans and Leighton (1989b), Ferber and Waldfogel (1998), Hamilton (2000), Holtz-Eakin et al. (2000), Williams (2000), Bruce and Schuetze (2004) and Landier (2004). For example, while Evans and Leighton (1989b) and Hamilton (2000) estimated positive effects of self-employment experience on future wages in paid employment, Bruce and Schuetze (2004) and Landier (2004) estimated weak and strong negative effects, respectively. Hyytinen and Rouvinen (2008) also detected negative effects of self-employment experience on employee wages in Europe, which they attributed to self-selection of less able types into and out of self-employment. However,

Evidence about the determinants of entrepreneurship

15.

16. 17. 18. 19. 20.

21.

22. 23.

24. 25. 26.

27.

28.

29.

30.

159

Hyytinen and Rouvinen’s (2008) data sample covers only brief spells, which probably oversample negative selectivity effects. Burke et al. (2000), Knight and McKay (2000), Cramer et al. (2002) and Cowling et al. (2004). In countries like Austria and Germany, advanced vocational qualifications at the tertiary level can be a legal prerequisite for self-employed artisans to trade (Lohmann and Luber, 2004). See Bull and Winter (1991), Guesnier (1994), Audretsch and Fritsch (1994a), Parker (2005b) and Acs and Armington (2006). See also Birley (1985), Sanders and Nee (1996) and Florin et al. (2003). See Bond and Townsend (1996), McPherson et al. (2001), Ruef et al. (2003), Aldrich and Kim (2007) and Parker (2009a). See Abell et al. (2001), Djankov, Quan et al. (2005) and Sine et al. (2007). Honig (1996), Gomez and Santor (2001), Brüderl and Preisendörfer (1998) and Bosma et al. (2004) all reported significant positive effects of these measures of social capital on performance. Foreman-Peck et al. (2006) and Guiso et al. (2004) reported insignificant effects using proxies for social capital of trade associations or business clubs and electoral participation, respectively. See Uusitalo (2001), Cramer et al. (2002), Puri and Robinson (2005), Ekelund et al. (2005), Djankov et al. (2006) and Kan and Tsai (2006). Insignificant effects are reported by Tucker (1988) and Parker (2008a). See e.g. de Clercq and Arenius (2006), Koellinger and Minniti (2006) and Levie (2007). See Ilmakunnas and Kanniainen (2001), Mazzeo (2004), Kanniainen and Vesala (2005), Wu and Knott (2006), Wennberg et al. (2007) and Ghosal (2007). Another strategy is to use aggregate proxy variables such as the number of strikes, the inflation rate or low rates of fixed capital formation as measures of economy-wide risk (Parker, 1996; Robson, 1996; and Cowling and Mitchell, 1997). For references, see de Meza and Southey (1996), Camerer and Lovallo (1999) and Coelho et al. (2004). See Baron (1998), Forbes (2005) and Schade and Koellinger (2007). In terms of chapter 2’s terminology, let F(x) be the actual distribution function of the random variable x, and let G(x) be an individual’s subjective distribution function. Suppose higher values of x are preferred to lower values of x. Rational expectations implies F(x) = G(x), whereas overoptimism implies that G(x) first-order stochastically dominates F(x), e.g. the first central moment of the subjective probability distribution is excessively large. This is because an over-optimistic person overestimates the probability of high-x outcomes and underestimates the probability of low-x outcomes. Overconfidence in contrast implies that G(x) second-order stochastically dominates F(x), e.g. the second moment of the subjective probability distribution is excessively small, perhaps reflecting the individual’s misplaced self-confidence in the accuracy of their beliefs. See also Pinfold (2001) and Koellinger et al. (2007). Cooper et al. (1988) sample was of new business founders who had only recently started up, so survival bias is unlikely to contaminate these results. In contrast, survival bias might vitiate evidence about the accuracy of entrepreneurs’ forecasts about future growth (Ashworth et al., 1998; Cassar and Gibson, 2007), which are in any case imprecise measures of over-optimism. See e.g. Landier and Thesmar (2008) and Puri and Robinson (2005). Puri and Robinson (2005) measured over-optimism as the difference between self-reported life expectancy and statistical life expectancy based on smoking-, age-, education-, race- and gender-specific mortality tables. This might explain why some industries are characterised by persistent negative profits, although an alternative reason might be that potential entrepreneurs confuse the median payoff with the mean payoff (Capone and Capone, 1992). In the economics of entrepreneurship literature, positive effects from Evans and Leighton (1989b) and Schiller and Crewson (1997) contrast with negative effects from van Praag and van Ophem (1995) and Blanchflower and Oswald (1998).

160

Selection

31. See Begley and Boyd (1987) and the review by Rauch and Frese (2000). Lee and Tsang (2001) provide some dissenting evidence – though rates of growth, n-Ach and locus of control were all self-reported in their study. 32. See Boyd (1991), Robinson and Sexton (1994) and Schiller and Crewson (1997). 33. Gruber and Poterba (1994), Hamilton (2000), Wellington (2001, 2006) and Perry and Rosen (2004). 34. Lentz and Laband (1990) obtained a very similar figure based on NFIB data, of 52.5 per cent. 35. Although see de Wit and van Winden (1989, 1990) for contrary evidence. 36. For GEM studies see de Clercq and Arenius (2006) and Levie (2007). A study using data from rural Spain is Lafuente et al. (2007). 37. Dunn and Holtz-Eakin (2000), Hundley (2006) and Colombier and Masclet (2008). 38. See Evans and Leighton (1989b), Lohmann and Luber (2004) and Niittykangas and Tervo (2005). 39. Contrast positive effects in Borjas and Bronars (1989) and Bernhardt (1994) with negative effects in Laferrère and McEntee (1995) and Lindh and Ohlsson (1996). 40. Fairlie (1999), Hout and Rosen (2000) and Fairlie and Robb (2007b). 41. White (1982), Acs and Audretsch (1989), Bhide (2000), Arauzo-Carod and Segarra-Blasco (2005) and van Stel et al. (2007). 42. See Shane (2003, pp. 125–9; 138–143) for a review of business studies articles which support these (rather obvious) points. 43. For evidence, see Jaffe (1989), Jaffe et al. (1993), Audretsch and Feldman (1996), Audretsch and Stephan (1996) and Audretsch, Keilbach and Lehmann (2006), among others. 44. See Armington and Acs (2002) and Acs and Armington (2004a, 2006) for US evidence and Audretsch and Keilbach (2007b) for German evidence. 45. See Audretsch (1995a), Audretsch, Keilbach and Lehmann (2006) and Audretsch and Keilbach (2007b). 46. Audretsch and Acs (1994), Grant (1996) and Carrasco (1999). 47. The formerly unemployed are also one-and-a-half times as likely to exit self-employment in the first year as former employees; but the entry rate exceeds the exit rate for these workers. For other evidence on the high proportion of unemployed workers transitioning to self-employment, see Storey (1982), Hakim (1988, 1989b), Meager (1992b), Blanchflower and Meyer (1994), Reynolds (1997), Kuhn and Schuetze (2001) and Rissman (2003). 48. Evans and Leighton (1989b), Carrasco (1999), Knight and McKay (2000) and Uusitalo (2001). 49. Evans and Leighton (1989b, 1990), Alba-Ramirez (1994), Cowling and Mitchell (1997) and Moore and Mueller (2002). A more complicated inverse-U-shaped relationship appears in Ritsila and Tervo (2002). 50. Farber (1999). See also Gordus et al. (1981), Carroll and Mosakowski (1987) and Laferrère and McEntee (1995). 51. McManus (2000), Taylor (2001), Moore and Mueller (2002). 52. Johnson (1981), Johnson and Rodger (1983) and McManus (2000). 53. See Meager (1992a, 1994) for a critique based on bias caused by mis-measurement of the dependent variable and Parker (2004, note 34, p. 111) for a response. 54. Storey and Jones (1987), Robson (1991) and Foti and Vivarelli (1994). 55. Tervo and Niittykangas (1994), Niittykangas and Tervo (2005) and Tervo (2006). 56. Examples of the former include Blanchflower (2000), Schuetze (2000), Henley (2004), Niittykangas and Tervo (2005) and Tervo (2006). Examples of the latter include Acs, Audretsch and Evans (1994) and Parker and Robson (2004).

Evidence about the determinants of entrepreneurship

161

57. E.g. Hamilton (1989), Storey (1991), Audretsch and Jin (1994) and Ritsila and Tervo (2002). 58. See Parker (2005b), Acs and Armington (2006) and Fritsch and Mueller (2007). Fritsch and Mueller (2007) explain more than one-half of the variance in German regional start-up rates in terms of regional start-up rates from fiftteen years earlier. 59. For evidence based on regional firm formation rates from a variety of countries, see Spilling (1996), Kangashargu (2000), Acs and Armington (2006), Fritsch and Falck (2007), Fritsch and Mueller (2007), Nyström (2007), Naudé et al. (2008) and Tamásy and Le Heron (2008). For evidence based on regional self-employment rates, see Boyd (1990) and Schiller and Crewson (1997). 60. See Blanchflower (2000), Fairlie and Meyer (2003) and Santarelli and Vivarelli (2007) for additional evidence. An exception is Russia, where entrepreneurs are more mobile across jobs and locations than non-entrepreneurs (Djankov et al., 2006). While international tax competition can promote the geographical mobility of entrepreneurs (Honkapohja and Turunen-Red, 2007) to date there is only anecdotal evidence (e.g. of French entrepreneurs living in England) that this is a quantitatively important phenomenon. Stam (2007) discusses other factors which bear on entrepreneurs’ mobility decisions. 61. See, e.g., Bygrave and Minitti (2000), Minniti (2005), Landier (2004) and Parker (2005b). 62. See Feldman (2001), Miller (1984) and Rocha and Sternberg (2005). 63. Evans and Leighton (1989a), Reynolds and White (1997) and Schiller and Crewson (1997). 64. Non-German studies include Davidsson (2006), Kim et al. (2006), Wagner (2006b), de Clercq and Arenius (2006) and Langowitz and Minniti (2007). For Germany, compare Wagner (2004), Mueller (2006b), Wagner and Sternberg (2004) and Bergmann and Sternberg (2007). 65. Delmar and Davidsson (2000), Davidsson and Honig (2003), Kim et al. (2006) and Mueller (2006b). 66. Delmar and Davidsson (2000), Reynolds et al. (2004) and Wagner and Sternberg (2004). 67. Davidsson (2006) concludes this from a review of conflicting evidence including Reynolds (1997), Delmar and Davidsson (2000) and Reynolds et al. (2004). 68. For example, contrast Delmar and Davidsson (2000), Davidsson and Honig (2003), Arenius and Minniti (2005) and Mueller (2006b) with US PSED evidence from Kim et al. (2006). 69. See Reynolds and White (1997), Delmar and Davidsson (2000), Arenius and Minniti (2005), Rotefoss and Kolvereid (2005), Wagner (2006b) and Bergmann and Sternberg (2007). 70. See Reynolds (1997), Delmar and Davidsson (2000), van Gelderen et al. (2001), Reynolds et al. (2004), Arenius and Minniti (2005), Kim et al. (2006) and Mueller (2006b). 71. Reynolds et al. (2004), Kim et al. (2006) and Koellinger and Minniti (2006). 72. See, e.g., Carter et al. (2003) for a survey of business studies articles exploring reasons entrepreneurs cite for starting up. 73. See Katz (1990), Delmar and Davidsson (2000) and Henley (2007). 74. Carter et al. (1996), Davidsson and Honig (2003), Delmar and Shane (2004) and Kessler and Frank (2004). According to Reynolds and White (1997, chap. 4), efforts and investment in start-up organising efforts are greatest for those who ultimately launch their ventures. 75. Reynolds and White (1997, chap. 4), van Gelderen et al. (2001) and Kessler and Frank (2004). 76. Compare Davidsson and Honig (2003), Vivarelli (2004), van Gelderen et al. (2006) and Parker and Belghitar (2006). 77. The interested reader is referred to, e.g. Feldman (2003) and Rosenberg (2003). 78. See Feeser and Willard (1989), Dahlstrand (1997) and Hyytinen and Maliranta (2008). 79. Klepper (2001), Gompers et al. (2005), Klepper and Sleeper (2005).

162

Selection

80. Interestingly, most of the successful spinoffs in the US disk-drive industry were started by engineers (Christiansen, 1993), while the most successful laser spinoffs were founded by entrepreneurs with backgrounds in marketing and sales, rather than in R&D (Sleeper, 1998; see also Roberts, 1991). 81. Dahlstrand (1997), Klepper (2001) and Koster (2005). 82. Cooper and Bruno (1977), Cooper (1985) and Roberts (1991).

5

Ethnic entrepreneurship and immigration

Entrepreneurship among ethnic minorities and immigrants is an increasingly hot topic. There are several reasons for this. It is widely believed that entrepreneurship offers a route out of poverty for disadvantaged groups and opens up opportunities for economic advancement and assimilation. The value of entrepreneurship for ethnic minorities is enhanced to the extent that they face discrimination. Thus Glazer and Moynihan (1970, p. 36) argue that ‘business is in America the most effective form of social mobility for those who meet prejudice’. Also, minority entrepreneurship can promote economic development and job creation in poor neighbourhoods (Bates, 1993, 2006), although it can also be a source of ethnic tension, as has been observed among Korean-owned businesses located in black communities, for example (Yoon, 1991; Min, 1996). It might be helpful to commence with several ‘stylised facts’ about ethnic entrepreneurship. Most of the extant evidence pertains to the USA and the UK, although it should be borne in mind that the rise of ethnic minority and immigrant entrepreneurship is an international trend, driven largely by demographic changes in both developed and developing economies (Ram and Smallbone, 2003). First, the evidence generally shows that a far higher proportion of whites engage in entrepreneurship than blacks do. For example, self-employment rates of whites are between two and three times higher than those of blacks.1 According to Fairlie and Meyer (2002), this differential has persisted since at least 1910, suggesting that little has changed since Myrdal (1944) first bemoaned the dearth of black-owned businesses in America. Second, many non-black ethnic groups have higher participation rates in entrepreneurship than whites do. For example, Clark and Drinkwater (1998) reported that at the start of the 1990s, Chinese, Pakistanis, Bangladeshis and Indians in Britain had substantially higher self-employment rates (of 26.6, 22.8, 17.8 and 19.6 per cent, respectively) than whites did (12.3 per cent). There is certainly pronounced heterogeneity among ethnic groups in terms of their tendency to become entrepreneurs. For example, 1990 US Census data show that non-rural male self-employment rates varied substantially across sixty ethnic and racial groups, both before and after controlling for age, education, immigrant status and length of time spent in the USA (Fairlie and Meyer, 1996). According to the 1990 US Census, only 4.4 per cent of black males worked 163

164

Selection

for themselves, compared with 27.9 per cent of Korean-American men. Americans of European origin had self-employment rates close to the US average. Members of ethnic groups from the Middle East and neighbouring countries such as Armenia and Turkey also had high self-employment rates; but Hispanics (other than Cubans) had low self-employment rates. Fairlie and Meyer (1996) also highlighted diversity within the ‘black’ ethnic group, with black Africans and Caribbeans having slightly higher self-employment rates than other black Americans (but still below the US average). This finding, together with similar evidence from the UK, cautions against treating ethnic minorities as a single homogeneous group. Third, trends in entrepreneurship also vary across groups, partly in response to demographic shifts. Latino entrepreneurship in the US is a case in point. According to recent data from the US Census Bureau, Hispanics now represent nearly 14 per cent of the US population, slightly above the population share of blacks. Latinos have similar self-employment rates to blacks, yet the number of Latino enterprises grew by 31 per cent between 1997 and 2002, three times the US national average, albeit from a small base.2 Much of this growth has been in low-value self-employment in services, especially among foreign-born and poorly educated Latinos with modest assets – although the number of larger-scale minority businesses with gross revenues exceeding $1 million per annum has also grown three times faster than the overall group (Bates, 2006). A fourth stylised fact about ethnic entrepreneurship in the USA is that black-owned businesses perform less well on average than white-owned businesses do. This is not the case for American-Asian entrepreneurs, though, who do at least as well as whites on average (Bates, 1997). Consider for example the following findings from US studies: • Black male self-employees earn about one-half of white male self-employees; the corresponding figure for Hispanics is about two-thirds (Borjas and Bronars, 1989; Flota and Mora, 2001). • Black-owned firms are less profitable than white-owned firms. According to 1992 CBO data, only 13.9 per cent of black-owned firms recorded annual profits of $10,000 or more, compared with 30.4 per cent of white-owned firms; and nearly 40 per cent of all black-owned firms recorded negative profits (Fairlie and Robb, 2007b). • Sales of black-owned firms are a small fraction of those of white-owned firms. According to 2002 Survey of Business Owner data, annual sales for these groups averaged $74,018 and $437,870, respectively (Robb and Fairlie, 2007). • Black-owned firms have higher closure rates and hire fewer workers than whiteowned firms do. According to CBO data from 1992, black entrepreneurs hired only 0.63 employees on average, compared with 1.80 for white entrepreneurs. Despite these findings, self-employed black and Hispanic men receive higher mean and median total labour incomes than their wage and salary counterparts. So despite underperformance relative to whites, it is still possible that entrepreneurship can provide a route for relative economic advancement for ethnic minorities (Fairlie, 2004).

Ethnic entrepreneurship and immigration

165

A fifth stylised fact is that some immigrant groups can have higher rates of participation in entrepreneurship than natives, although there is mixed evidence across groups and about their relative financial performance (Schuetze and Antecol, 2006). For example, 1990 US Census micro-data show that the self-employment rate of immigrants was 23 per cent higher than that of natives, although Asian immigrants display marked heterogeneity in this respect, with Korean immigrants having the highest participation rates in entrepreneurship (Fernandez and Kim, 1998; Lunn and Steen, 2005). Overall, the US evidence relating to relative immigrant entrepreneurship rates is mixed.3 In Germany, self-employed immigrants earn 40 per cent more than immigrant employees and 7 per cent more than self-employed natives. This suggests that entrepreneurship may be an attractive option for immigrants in that country (Constant and Zimmermann, 2006). But Swedish evidence points to a major earnings disadvantage for immigrant entrepreneurs (Hjerm, 2004). This chapter explores some of the factors underlying these ‘stylised facts’. Both theoretical models and empirical evidence are discussed. Essentially, two hypotheses have been advanced to explain the observed variations in rates of entrepreneurship between ethnic groups. The first hypothesis relates to discrimination, perpetrated either by employers in the labour market, by banks in the capital market, or by consumers in the product market. The next section treats these issues in turn. The second hypothesis, explored in the section after, posits positive factors which can help make entrepreneurship attractive to members of particular minority groups. Then further evidence on the determinants of ethnic minority entrepreneurship is reviewed, followed by a discussion of theory and evidence relating to entrepreneurship among immigrants. Throughout the chapter, M and NM are used to denote ‘minority’ and ‘non-minority’ values of the variables to which they are attached. In particular, wM and wNM denote minority and non-minority wage rates in paid employment, respectively; while πM and πNM denote minority and non-minority profits in entrepreneurship, respectively. 5.1

Discrimination

5.1.1

Discrimination in the labour market If employers have an exogenous taste for discriminating against members of ethnic minorities, M , what are the implications for ethnic entrepreneurship, and entrepreneurial profits of minority members? Previous researchers have proposed two effects of discrimination by employers against employees of ethnic minority status:4

1. By preventing members of minorities from obtaining jobs in paid employment or by restricting them to relatively low-paid jobs, discrimination increases the attractiveness to them of entrepreneurship. In other words, entrepreneurship can act as an ‘escape route’ from employer discrimination, implying greater participation in entrepreneurship for these individuals. 2. Reflecting lower pay in paid employment, the ratio of minority to non-minority average entrepreneurial profits, πM /πNM , should exceed the ratio of minority to non-minority wages, wM /wNM (Moore, 1983b).

166

Selection

In fact, point 2 is inconsistent with the evidence about the relative earnings of blacks and Hispanics.5 One reason might be that crowding of M s into entrepreneurship competes down the price they can command for their output and hence their profits, πM . Alternatively, even if the distribution of ability within each ethnic group is identical, employer discrimination reduces πM relative to πNM in a free occupational choice equilibrium.6 Another problem with the employer discrimination hypothesis is that point 1 above is inconsistent with the facts about American and British blacks, who have lower participation rates in entrepreneurship than whites do. Although some ethnic minorities (such as Korean-Americans and British-Asians) have above-average entrepreneurship participation rates, this is an unsatisfactory defence of the employer discrimination hypothesis because it fails to explain why employers discriminate against some ethnic groups but not others. Indeed, Lunn and Steen (2005) argue that substantial heterogeneity of rates of entrepreneurship among Asian-Americans alone is sufficient to cast doubt on the discrimination hypothesis, since one can ask why individuals would discriminate against some Asian groups but not others.7 Finally, it is worth pointing out that when markets are inefficient (a ‘second-best’ outcome), discrimination in paid employment can sometimes actually increase social welfare. Chapter 12 explains that some individuals may engage in ‘rent-seeking’ activities, whereby resources are expropriated from productive entrepreneurs. Suppose that minorities (e.g. Jews in medieval Europe) are debarred from profitable rent-seeking activities (e.g. collecting and enforcing taxes) and are pushed into particular commercial activities instead (e.g. diamond trading). This form of discrimination increases productive entrepreneurship among the minority group, to their benefit; while the nonminority rent-seeking group benefits from having a greater source of income to tax away (Murphy et al., 1991). Of course, in a ‘first-best’ world there would be neither rent-seeking nor discrimination.

5.1.2

Discrimination in the capital market

I now turn to research claiming that lenders in the small business credit market discriminate against ethnic minority entrepreneurs. I commence with a brief overview of some theoretical considerations; go on to outline methods used in empirical research on this issue; and then review the evidence and its implications. It is sometimes argued that lenders practice statistical discrimination against some ethnic groups. Statistical discrimination describes the situation where an ethnic group has different characteristics on average from others, which are then used to adversely screen all members of that group. For example, many UK minority-owned businesses establish themselves in sectors such as retailing, transportation and catering, which have above-average failure rates (Bank of England, 1993). And African Americans possess less personal wealth on average than whites do (one-eleventh, according to Robb and Fairlie, 2007) and hence can offer banks less collateral.8 So even if banks do not discriminate on the basis of ethnicity, bank competition may generate lending rules

Ethnic entrepreneurship and immigration

167

that reward high-collateral and safe-sector start-ups with larger loans – resulting in outcomes that resemble discrimination, since blacks will be observed to receive fewer and/or smaller loans on average. Coate and Tennyson (1992) demonstrate how employer discrimination can spill over into statistical discrimination in the credit market. Consider again the simple occupational choice model outlined in chapter 2. If minorities receive a low outside wage, then more of them will enter entrepreneurship but with a lower average ability (see note 6). Lower ability reduces the probability that an entrepreneurial venture succeeds, which forces lenders operating under conditions of competition to increase the interest rate they charge them in order to recoup their losses. This reduces the profitability of ethnic ventures net of interest repayments. Effectively, lenders observe that higher default rates are positively correlated with ethnicity, and so have incentives to practise statistical discrimination.9 Nevertheless, statistical discrimination is illegal under US federal law. Even if bank profits vary systematically with the ethnicity of the borrower, banks are prohibited from conditioning loan decisions and loan terms on ethnicity (Blanchard et al., 2008). I now turn to empirical issues. Recent empirical research is based on two econometric models. One is a binary choice model (see chapter 3) where the dependent variable takes the value 1 if the individual is denied a loan, and 0 if it is granted. The second is a regression model where the dependent variable is the loan rate of interest. Among the independent variables in both models are one or more dummies for ethnicity. If the coefficient on an ethnic dummy is positive and significant in either case, then this is consistent with racial discrimination, provided the specification omits no explanatory variables which might themselves explain adverse ethnic loan outcomes. Any omitted or unobserved variables that explain actual loan application success or interest rates, but which are correlated with ethnicity, may lead to the incorrect inference that discrimination is present. It is therefore essential that rich and detailed data sets are utilised in applied research, in order to minimise the risk of omitted (or unobserved) variable bias. A data set which satisfies these criteria is the US National Survey of Small Business Finances (NSSBF), which contains numerous variables (including, in recent years, credit ratings). Several authors have exploited it to estimate the extent of discrimination in the US small business credit market. For example, Cavalluzzo and Cavalluzzo (1998) estimated loan denial rates using NSSBF data from the late 1980s. They uncovered evidence of substantially and significantly higher loan denial rates forAfrican-Americans than for white Americans, with no significant differences by gender or between whites and Hispanics. Similar results were obtained by Blanchflower et al. (2003), using richer NSSBF data from 1993 and 1998, which also included controls for credit histories, housing and nonhousing net worth. Blanchflower et al. (2003) estimated that black-owned businesses are about twice as likely to be denied credit as white-owned businesses; and that blacks face interest rates that are about one full percentage point higher than those facing whites. This is a substantially larger difference in loan denial rates than is commonly reported by studies of discrimination in the mortgage loan market. This difference

168

Selection

might be larger still if discouraged black borrowers – who anticipate discrimination – are also taken into account. A distinct though informal impression of discouraged black borrowers emerges from a descriptive study of 235 small businesses in a Chicago ethnic neighbourhood (Bond and Townsend, 1996).10 Further work by Cavalluzzo et al. (2002) and Cavalluzzo and Wolken (2005) suggests that smaller stocks of personal wealth play a modest role in explaining higher loan rejection rates among Asian-American and Hispanic men compared with whites. Cavalluzzo and Wolken (2005, 2005) observe that self-reported credit histories have greater power in explaining ethnic differences in loan denial rates, as does the market concentration of lenders proximate to the entrepreneur. Market concentration implies weaker inter-bank competition and hence greater scope for lenders to discriminate and still make profits – since one might expect discrimination which trades off profits for indulgence in a Beckerian ‘taste’ for discrimination to be unsupportable under perfect competition. Another interesting finding relates to credit scoring. It has been proposed that credit scoring might increase the objectivity of loan decisions (Berger et al., 2005) and hence reduce discrimination. The fact that signs of credit discrimination persist even after controlling for credit histories and credit-scoring practices must therefore cast doubt on credit scoring as a straightforward solution to (statistical) discrimination. In a recent study, Blanchard et al. (2008) survey the literature on ethnic credit market discrimination and conclude that black-owned firms are at least 20 percentage points more likely to be denied access to credit than are equally qualified white-owned firms (see also Bostic and Lampani, 1999; and Cavalluzzo et al., 2002). According to Blanchard et al. (2008), only a minority of studies have found similar evidence of discrimination against Hispanic- and Asian-owned firms. Using 1993 and 1998 NSSBF data, Blanchard et al. (2008) report statistically significant and substantial discrimination against black-owned and Hispanic-owned businesses in loan approval rates. But they find no ethnic differences in interest rates; and evidence of loan decision discrimination disappears for high net worth black-owned firms. What should we make of these findings, drawn from studies which explain at most one-third of the variation in loan rejection rates; yet use a comprehensive set of covariates which approximate those used by small business loan officers in their loan application procedures; and which find that ethnic groups have substantially higher loan rejection rates – all in apparent contravention of US Federal law? On the one hand, the comprehensiveness of the set of control variables used by researchers, which corresponds to the criteria actually used by banks in actual loan application procedures, makes it hard to avoid the conclusion that discrimination exists in the US market for small business credit. According to Blanchard et al. (2008), the most likely explanation of these findings is statistical discrimination. On the other hand, Cavalluzzo and Wolken (2005) acknowledge the possibility that some omitted variable bias remains, owing in part to lack of data about projected cash flows, which loan officers claim to use in practice. Also, loan officers can validate the accuracy of self-reported responses about loan applicants’ financial conditions, which the researcher typically cannot do. And if researchers use a different model of loan decisions from the bankers,

Ethnic entrepreneurship and immigration

169

any mis-specification of the former may resemble credit market discrimination without any actually being present. Another problem with the research findings discussed above is that they are based on ‘reduced form’ specifications which do not disentangle supply from demand factors (Berger and Udell, 2003, p. 317–18). Acknowledging these problems, Blanchflower et al. (2003) propose that a truly convincing test of discrimination should match loan decisions with actual default rates. Despite the need for caution, the recent evidence does seem to suggest that some ethnic minorities, in particular blacks, find it harder to borrow and to become entrepreneurs in the USA. It also bears out earlier research showing that on average blacks are granted smaller loans for start-ups, and have more limited access to venture capital and trade credit than whites do, even after controlling for characteristics such as education and financial assets.11 An implication of this research is that blacks are both less likely to be able to start businesses and more likely to be under-capitalised and therefore vulnerable to subsequent failure than whites (Robb and Fairlie, 2007). In this regard, a striking finding from Bates’ (1991) analysis of 1982 CBO data is that blacks’ failure rates would have been no different from those of whites if blacks had received the same amounts of external finance. The UK evidence seems to paint a somewhat different picture. There, the main financing difference appears not to be between whites and blacks, but between Asians and Afro-Caribbeans. According to Jones et al. (1994), Asians have a higher probability of obtaining a bank loan than Afro-Caribbeans and whites do, and leverage more funds from banks. These findings cast doubt on the proposition that UK banks are guilty of blanket discrimination, though it does beg the question about why Afro-Caribbeans report greater difficulties in raising bank loans than whites do.12 More recent work by Fraser (2007), which employs a similar econometric approach to Blanchflower et al. (2003) (using fewer covariates) detects no signs of significant racial discrimination in access to loans in the UK loan market. If for whatever reason minorities encounter difficulties with raising bank loans, then they presumably have an incentive to raise capital elsewhere. One possibility is venturecapital funds which specialise in financing minority business enterprises. But recent research shows that, despite earning similar yields to those of ‘mainstream’ venture capital companies, only a tiny number of minority entrepreneurs are served by this type of finance (Bates and Bradford, 2007). Other funding possibilities include (i) family finance, (ii) trade credit, (iii) franchising, and (iv) use of minority-owned banks and Rotating Savings and Credit Associations (Roscas). Regarding (i), Afro-Caribbeans in Britain have below-average access to family finance as well as to formal finance. In partial compensation, British Afro-Caribbeans make the greatest use of formal advice and assistance from third parties (Ram and Smallbone, 2003). There is less information about (ii), although evidence from developing countries points to more limited access among blacks (and women) to trade credit, primarily reflecting more limited social networks (Fafchamps, 2000). More promising is (iii). Franchising appears to be an effective way for blacks to avoid credit restrictions, since banks give preferential treatment to loan applicants wishing to set up franchise

170

Selection

outlets (Felstead, 1991). But there is as yet little hard evidence that members of ethnic minorities choose franchising in order to escape borrowing constraints. I will say more about franchising among blacks in the next subsection. Finally, concerning (iv), there is a small but growing literature on minority-owned banks and Roscas, which have proven especially popular among Chinese, Japanese and Korean immigrant groups in the USA. Some sociologists claim that Roscas have enabled impecunious immigrants to bootstrap their way to business success.13 For example, in Yoon’s (1991) survey of 199 Korean merchants in minority neighbourhoods in Chicago, 27.6 per cent used loans from Korean Roscas, 27.1 per cent used loans from banks and 34.7 used loans from kin. However, this coverage does not seem to extend to national data sets such as the CBO, where Rosca lending appears to be of marginal importance and is associated with smaller and more failure-prone businesses (Bates, 1997). Involvement in Roscas in the USA seems to have peaked in the early part of the twentieth century, becoming progressively less important over time as ethnic groups have gained greater access to formal credit markets (Besley, 1995). Another problem with the proposition that minority-owned banks can solve capital market discrimination problems is that many ethnic groups (including blacks) have not emulated Asian Roscas – which might have been expected to occur if they were so effective. Closer inspection reveals that Roscas’ value is actually rather dubious. Much Rosca finance is short-term and provided at high interest rates that can exceed 30 per cent per annum. This presumably makes a Rosca a lender of last resort to many borrowers. It is also pertinent that many Roscas are designed primarily to encourage savings rather than business investment by members.

5.1.3

Discrimination in the product market

Another possibility is that NM consumers dislike buying goods and services from M entrepreneurs. This can be expected to reduce the latter’s returns in entrepreneurship, and hence the number of M entrepreneurs. Historically, this type of discrimination might explain the proliferation of black personal service companies which catered specifically for black Americans in response to white consumer discrimination in the first half of the twentieth century (Myrdal, 1944). Evidence from US Census data suggests furthermore that disadvantage continued to shape black American entrepreneurship well into the 1970s (Boyd, 1991). There is evidence though that these consumer patterns started to change in the 1980s. Bates (2003) outlines the post-1970s trend among AfricanAmerican entrepreneurs away from traditional niches such as barbershops and ‘mom and pop’ food stores into new lines of business such as finance, business services and professional services. Borjas and Bronars (1989) studied a model of consumer discrimination in which NM consumers are assumed to have a taste for discrimination against M sellers, while M consumers are indifferent to the race of the seller. Their model generates the following testable predictions. First, the mean income of M sellers will be lower than that of NM sellers, directly reflecting consumer discrimination. As a result, skilled Ms have

Ethnic entrepreneurship and immigration

171

greater incentives to enter paid employment than skilled NMs. Unskilled Ms have greater incentives to enter entrepreneurship than unskilled NMs. Second, NM sellers have a higher return to ability than M sellers. These predictions contrast with those of the employer discrimination model and accord with the evidence about low relative minority entrepreneurial incomes cited at the start of this chapter.14 While the consumer discrimination hypothesis appears useful for understanding ethnic differences in entrepreneurship, it seems less suitable for explaining gender differences. As Aronson (1991) pointed out, women are commonly employed in sales jobs, which would not occur if firms knew that consumers discriminated against them. Another problem with the consumer discrimination hypothesis is that black businesses are relatively common in industries patronised by white customers (Meyer, 1990). One reason could be franchising, since franchisors often discourage attempts by franchisees to differentiate their units (Kaufmann and Lafontaine, 1994) – so reducing consumers’ ability to discriminate. Indeed, Williams (1999) found that black entrepreneurs were more likely than any other ethnic group to become franchisees. Williams also estimated that blacks earned more as franchisees than they would as independent business owners, a finding that is also consistent with Borjas and Bronars’ (1989) discrimination model. If Williams’ findings are true more generally, they suggest that franchising could be a successful way of increasing the level of entrepreneurial activity among blacks. Finally, we note that discrimination might not be the only negative factor inhibiting black entrepreneurship. It has sometimes been argued that slavery perpetuated attitudes which are inimical to entrepreneurship. However, in the early and middle parts of the previous century, southerners dominated African-American entrepreneurship, especially in banking, insurance and publishing – which suggests that there was indeed a black entrepreneurial tradition, despite slavery in the American South (Boyd, 2006). Furthermore, attitudes inculcated by slavery cannot explain the rise of predominantly black entrepreneurs in the northern mass-media entertainment industry catering to racially integrated audiences. 5.2

Positive factors

Discrimination can be regarded as a factor which ‘pushes’ members of ethnic minorities into the escape route of entrepreneurship. Another possibility is that ‘pull’ factors can make entrepreneurship positively attractive to members of minority groups. The following pull factors have been proposed: 5.2.1

Positive expected relative returns in entrepreneurship Positive rewards in entrepreneurship, rather than discrimination in paid employment, may explain high rates of entrepreneurship among some ethnic groups (Bearse, 1984). For example, American and British estimates of the binary choice model with relative incomes (see chapter 3) reveal that relative income differences help explain differences in self-employment rates across ethnic lines (Fairlie and Meyer, 1996; Clark and Drinkwater, 2000).

172

5.2.2

Selection

Ethnic enclaves ‘Enclaves’are geographical clusters of ethnic group members who form self-supporting economic communities. Enclaves can provide information networks, sources of credit, ‘protected’ or ‘niche’ markets for the output of ethnic entrepreneurs and a steady supply of workers, possibly drawn from close-knit extended families (Light and Bonacich, 1988). Ethnic minority entrepreneurs may know more about the tastes of ethnic consumers, in such ‘protected markets’as clothing, foodstuffs, religious goods and services (Aldrich et al., 1985). These advantages, and the absence of consumer discrimination by co-ethnics, presumably increase the opportunities and ease with which minority group members can operate a business. Set against this argument, however, is the possibility that the scope for expanding operations into broader markets is more difficult for enclave producers. Also, enclaves can foster intense competition among ethnic entrepreneurs, so limiting entrepreneurial opportunities (Aldrich and Waldinger, 1990) and reducing survival prospects (Bates and Bradford, 1992). In contrast, paid employment incomes may be relatively high in enclaves if ethnic employers do not discriminate against members of their own group. Opportunities for profitable entrepreneurship in enclaves may be limited if ethnic disposable incomes are low and consumer demand is modest, while segregation might generate social problems which are antithetical to successful entrepreneurship and exacerbate unfavourable negative stereotypes about ethnicity among non-ethnics (Fairchild, 2008). The available evidence from a range of countries certainly points to a concentration of ethnic minority entrepreneurs in particular industrial sectors. For example, US Census data from 1980 revealed that 27 per cent of self-employed immigrants worked in the retail sector, compared with 17 per cent of the native-born self-employed (Borjas, 1986). Whereas white self-employed people are concentrated in managerial, technical and professional occupations, their black counterparts tend to be concentrated in manual jobs (Becker, 1984; Boyd, 2006). In contrast, self-employment rates of Koreans are relatively high in all industries (Lunn and Steen, 2005). Even more pronounced ethnic industry concentration patterns are observed in the 1991 British Census, with 90 per cent of Asian self-employed people working in services, and only a tiny minority working in the construction sector (Clark and Drinkwater, 1998). Clark and Drinkwater (1998) also reported that 50 per cent of Indian, Pakistani and Bangladeshi self-employed people work in retail distribution, restaurants and taxi driving, while 80 per cent of the Chinese self-employed work in the restaurant industry. A direct empirical test of the ethnic enclave hypothesis examines whether the proportion of an area’s population belonging to one ethnic group positively affects that group’s rate of entrepreneurship in that area. Borjas (1986) pioneered this approach by including as an explanatory variable in a logit model of entrepreneurship participation the proportion of individuals’ local populations which was Hispanic. Borjas estimated that male Hispanics aged 18–64 were significantly and substantially more likely to be self-employed in areas with large Hispanic populations, whereas no such effect was

Ethnic entrepreneurship and immigration

173

detectable for whites. However, subsequent empirical investigations of this issue have generated mixed results.15 The enclave hypothesis also has problems explaining low black self-employment rates, especially since there is evidence in the USA and the UK of strong black networks, including a loyal black customer base and a practice of blacks hiring other blacks (Aronson, 1991; Jones et al., 1994). In contrast, incidentally, members of other selfemployed minority groups (especially Asians) predominantly hire workers from outside their ethnic group (Aldrich and Waldinger, 1990). Nor is the use of family labour confined to ethnic minorities: it appears to be a common practice among all ethnic groups (Jones et al., 1994). Others have argued that enclaves are not a route to entrepreneurial success, and primarily serve as a fallback for marginal ethnic entrepreneurs. Empirical work along these lines demonstrates that successful Asian-American entrepreneurs predominantly serve non-minority clients; raise finance from conventional lenders; and employ nonminority employees. More generally, the success of ethnic American entrepreneurs appears to be attributable not to ethnic resources but to heavy financial and human capital investments (Bates and Dunham, 1993; Bates, 1997). For example, of Asian immigrant entrepreneurs operating young firms in 1987, 58 per cent were college graduates (compared with 38 per cent of non-minority business owners), with an average start-up capital of $53,550 compared with just $31,939 for non-minority business owners (Bates, 1997). Earlier, Bates (1985) had observed that the most successful minority entrepreneurs were located outside of the ‘traditional’ personal service and retail sectors. The same research drew a contrast with the least successful Asian-owned firms, which relied on social support networks in enclaves; it appears that entrepreneurs with a predominantly ethnic minority clientele who are located in areas with large minority populations have significantly lower survival and profitability rates than the average.

5.2.3

Culture Building on Weber’s (1930) ‘Protestant Work Ethic’ thesis, it is possible that attitudes to entrepreneurship are determined by the culture and religion of particular ethnic groups (Rafiq, 1992). Some prominent figures in Islam and the Sikh religion were businessmen; and some Hindu castes specialise in business activities. In Britain, Clark and Drinkwater (2000) found that, all else equal, Muslims, Hindus and Sikhs had significantly higher probabilities of being self-employed than Christians from ethnic minorities. Related to this, some Asian cultures stress self-sufficiency, thrift and hard work, which might help to explain high British-Asian self-employment rates (Borooah and Hart, 1999). However, most empirical studies find little or no effect from religion on entrepreneurial choices.16 Poor command of the host country’s language might increase the likelihood of ethnic entrepreneurship, by restricting employment opportunities in the formal job market

174

Selection

without affecting trading opportunities among members of one’s own language group (Bates, 1997). Once again, the evidence on the issue derived from binary choice models is mixed.17 Yet host country language proficiency does appear to be associated with profitability in entrepreneurship.18 Another possibility is that belonging to a minority group may create a feeling of insecurity that encourages a drive for entrepreneurial success (Kilby, 1983; Elkan, 1988). 5.2.4

Role models and inculcation of positive attitudes

An absence of black entrepreneurial role models might help explain the lower rates of entrepreneurship among blacks, as well as their lower average entrepreneurial incomes. A lack of black role models within and outside the immediate family can also explain why Hout and Rosen (2000) found intergenerational links in self-employment to be strong for every American ethnic group except blacks (see also the discussion in the next section). Fewer favourable role models could translate into weaker pro-entrepreneurial attitudes among young blacks. For example, a Gallup survey of over one thousand young Americans detected greater ignorance about how markets work among black than among whites; and fewer black youths knew small-business owners than white youths did (Walstad and Kourilsky, 1998). These authors’findings reflect not only lower entrepreneurship rates among blacks, but also specifically an absence of prominent successful black entrepreneurs. It is noteworthy, for example, that black self-employment incomes are similar to those of whites at the three lower quartiles, but are significantly lower than whites’ incomes at the upper quartile (Hamilton, 2000).

5.3

Further evidence on determinants of ethnic differences in entrepreneurship

As the foregoing discussion has demonstrated, the literature has identified both negative and positive factors that impinge on ethnic entrepreneurship. Most of the negative factors are based on some kind of discrimination. As we saw, however, the role of discrimination has been questioned on both theoretical and empirical grounds. In the USA, for example, Bates (1997) concluded that the substantial human and financial capital inputs by Asian-American entrepreneurs relative to blacks help explain the former’s substantially higher rates of participation and average performance in entrepreneurship. And Fairlie (2006b) drew attention to below-average black transition rates from unemployment to self-employment – which is not what one would expect if black entry into self-employment was driven primarily by disadvantage considerations. Do different personal characteristics, or different returns to common personal characteristics, account for the observed differences in rates of entrepreneurship between ethnic groups? In an attempt to answer this question, Borjas and Bronars (1989) estimated what minority rates of entrepreneurship participation would have been if the coefficients from a self-employment probit based on a white sub-sample (i.e. imposing the same returns to characteristics) were applied to non-whites.19 Using this method, Borjas and Bronars (1989) estimated that blacks and Hispanics would have had the

Ethnic entrepreneurship and immigration

175

same self-employment rates as whites, and that Asians would have had a higher self-employment rate than whites. This implies that unobserved differences in rates of return, rather than differences in observable characteristics, account for most of the ethnic variation in self-employment rates. Similar findings have since been recorded by other researchers. For example, relatively low black self-employment rates in the USA cannot simply be explained by a concentration of blacks in low self-employment industries (Fairlie and Meyer, 2000) or by differences in family background (Hout and Rosen, 2000). Unfortunately, it is unclear whether discrimination, cultural factors, or unobserved characteristics are responsible for these different rates of return to personal characteristics by ethnicity. Further research is needed to dig deeper into this question. Entrepreneurship participation rates conflate entry decisions and survival outcomes. Fairlie (1999) demonstrated the importance of separating these influences in order to identify more precisely the causes of the relatively low African-American selfemployment rate. Using PSID data over 1968–89, Fairlie (1999) reported that black entry rates into self-employment were about one-half those of whites, while black exit rates were about twice those of whites. Using a decomposition analysis, Fairlie found that although some variables helped explain lower black entry rates, about twothirds of the gap was unexplained. Limited asset stocks explained about 15 per cent of the entry differential (see also Fairlie and Robb, 2007b), and the lower incidence of self-employed fathers among blacks was also a significant factor. Although Fairlie (1999) had less success explaining higher black rates of exit from self-employment compared with whites, other studies have highlighted the importance of low levels of initial start-up capital (Bates, 1997; Fairlie and Robb, 2007b). There is an important distinction here between recorded businesses and start-up organising efforts (‘nascent entrepreneurship’). As noted in chapter 4, AfricanAmericans have higher rates of entry into nascent entrepreneurship than white Americans do. There is therefore a striking difference between nascent entrepreneurship data and self-employment data in this respect. This difference can be resolved by recognising that African-American nascent entrepreneurs have higher abandonment rates than white nascent entrepreneurs (Parker and Belghitar, 2006). According to Fairlie (1999), education does not significantly affect either selfemployment entry or exit outcomes for black Americans; and he concluded that the scope for policy intervention to increase black self-employment rates is limited. Even if current black asset levels could somehow be quadrupled, that would reduce the ethnic gap in self-employment entry rates by only 13 per cent (Fairlie, 1999). Perhaps such a policy would have larger effects on enhancing the performance metrics outlined at the start of this chapter (Robb and Fairlie, 2007). Finally, there has been much less research on the determinants of growth of ethnicminority-owned businesses. Discussion of the growth of entrepreneurial ventures is deferred until chapter 11; but it is worth mentioning here a recent study of 350 small African-American-owned ‘gazelles’ by Boston and Boston (2007). Defining a gazelle as a company with an annual employment growth rate over five years of 20 per cent

176

Selection

or more, Boston and Boston (2007) reported that black-owned gazelles are: less likely than black-owned non-gazelles to compete on the basis of price; more likely to serve regional or national rather than local markets; and have fewer black employees on average. It is also notable that black-owned gazelles are more likely to sell to the government sector, which testifies to the continuing importance of public procurement policies towards American ethnic-minority entrepreneurs. We will say more about this issue in chapter 17.

5.4

Immigration and entrepreneurship

Are immigrants more likely to become entrepreneurs than native-born workers (referred to as ‘natives’ henceforth)? The first subsection below proposes several reasons why they might be, and summarises evidence about immigrants’ rates of participation in entrepreneurship. The second subsection reviews evidence about the determinants of immigrant entrepreneurship. The third subsection discusses the effects of immigration on entrepreneurship. These effects include short-term effects of competition on native entrepreneurs, as well as longer-term implications for immigrant entrepreneurs who return to their country of origin. At the outset, it is important to distinguish between within-country migrants from another region (‘in-migrants’) and migrants from another country altogether (‘immigrants’). The emphasis in this chapter is on immigrants rather than in-migrants, although one study has found little difference between the two groups in terms of nascent entrepreneurship in the UK, suggesting that ‘it is the effect of migrating, not the crossing of international borders, that influence the propensity to engage in new business activity’ (Levie, 2007, p. 164).

5.4.1

Immigrants’ entrepreneurial propensities It has been suggested that immigrants are likelier than natives to be entrepreneurs, for the following reasons:

1. On average, immigrants are better educated and motivated than natives. 2. Immigrants have access to ‘ethnic resources’and social capital (Light, 1984), including a tradition of trading, access to low-paid and trusted workers from the same ethnic group, and access to a ready market of niche products within an ethnic enclave. 3. Some immigrants are ‘sojourners’, who wish to immigrate temporarily in order to accumulate wealth before returning to their homeland. Entrepreneurship may be the most effective means to this end. 4. Immigrants turn to entrepreneurship because of ‘blocked mobility’ in paid employment, owing to language difficulties, discrimination or possession of non-validated foreign qualifications. 5. Immigrants are self-selected risk-takers by virtue of their willingness to leave their homeland to make their way in a foreign country.

Ethnic entrepreneurship and immigration

177

6. Among illegal immigrants, entrepreneurship in the form of self-employment may be a means of escaping detection by the authorities. 7. Immigrants enter industries and occupations that have high rates of entrepreneurship. Empirical studies, which mostly measure entrepreneurship as self-employment, have generated diverse findings about whether immigrants are more or less inclined to entrepreneurship than natives. Both Borjas (1986) and Lofstrom (2002) claimed to find higher self-employment rates among immigrants than natives in the USA, while Brock and Evans (1986) found no such pattern.All of these studies used US Census data. In their review of US studies about whether immigrants have higher self-employment rates or not, Levie and Smallbone (2007) concluded that the evidence is inconclusive. But in a recent comparative study, Schuetze and Antecol (2006) calculated that immigrant self-employment rates are 1 per cent higher than native rates in the USA, 2 per cent higher in Australia and 4 per cent higher in Canada. It has been argued that immigrant rates of entrepreneurship in a ‘host’ country reflect rates of entrepreneurship in a ‘home’ country. Light (1984) emphasised the diversity of self-employment experience among immigrants by ‘home’ country, which he attributed to different traditions of commerce. This theme was taken up by Yuengert (1995), whose analysis of 1980 US Census data indicated that immigrants from countries with relatively high self-employment rates were likelier to become self-employed in the USA. Yuengert (1995) estimated that 55 per cent of the immigrant-native self-employment rate differential was attributable to immigrants having above-average home-country self-employment rates. Although some supporting evidence comes from Hammarstedt (2001), for Sweden, other researchers have obtained contrary evidence (e.g. Fairlie and Meyer, 1996; van Tubergen, 2005). The nature of immigrant entrepreneurship appears to be changing. The labour economics immigration literature emphasises variations in ‘quality’among different cohorts of immigrants. For example, it has been suggested that shifts in US immigration policy which have increasingly prioritised family reunion criteria over human capital skill contributions have been responsible for a decline in immigrant ‘quality’ as measured in terms of earnings. Recent evidence suggests that in the last twenty years, immigrant cohorts to Western countries have become less rather than more likely to engage in entrepreneurship20 and now receive lower incomes there (Mora and Dàvila, 2006). These findings are consistent with the notion of declining quality of immigrants, at least in terms of entrepreneurship. Consequently, when studying this issue it is important to distinguish between ‘assimilation effects’, which capture the extent to which individuals within a given cohort assimilate into the host country, and ‘cohort effects’, which capture the possibility that cohorts differ in quality. To disentangle the two effects, Borjas (1986) proposed a useful decomposition, which is described in the chapter appendix. 5.4.2

The determinants of immigrant entrepreneurship Consider again the list of potential reasons why immigrants might be likelier than natives to become entrepreneurs. For point 1, the evidence shows that the relationship

178

Selection

between education and entrepreneurship varies across ethnic groups. On the one hand, immigrant Koreans to the USA not only have relatively high self-employment rates (at 12 per cent, more than double that of any other immigrant group), but they also tend to be more educated than average. And a non-US study by Vinogradov and Kolvereid (2007) found that more educated immigrants were significantly more likely to become selfemployed in Norway. But on the other hand, the least-educated Latino immigrants to the USA have the highest entrepreneurial participation rates, in contrast to the experience of Korean-American immigrants.21 Other evidence shows that immigrant business owners tend to obtain most of their finance from personal equity and mainstream lenders, rather than from social resources. Echoing findings from the ethnic entrepreneurship literature reviewed above, immigrant businesses depending on social resources tend to be ‘marginal’and more prone to failure (Bates, 1997).This is contrary to point 2 above.Yet it is also known that immigrants prefer to settle in locations where there are already numerous co-ethnics, as noted by Mora and Dàvila (2006) for Mexican immigrants to the USA. Waldinger et al. (1990) argue that aspiring immigrant entrepreneurs are likely to run ‘marginal’ businesses catering to ethnic or urban poor customers; to sell ethnic products to the general population or unstandardised manufactured goods; and to operate labour-intensive businesses. Another incentive (at least for Mexican-American immigrants) is to turn entrepreneur in order to exploit cross-border trading opportunities. US Census data from 2000 reveal that Mexican immigrants have significantly higher self-employment rates in US locations close to the US–Mexico border than their counterparts in the rest of the USA – and also than non-Hispanic whites in border cities (Mora and Dàvila, 2006). This is so even after controlling for socio-demographic factors, including local unemployment rates. But ‘crowding effects’ have reduced the incomes of such entrepreneurs, apart from those who still primarily operate in Mexico running cross-border businesses. The latter enjoy a significant earnings premium (Mora, 2006). Hence policies restricting trade and labour flows across the US–Mexico border might inadvertently retard Latino entrepreneurship. In summary, neither of the first two hypotheses in the list of potential reasons for immigrant entrepreneurship receives decisive support. Furthermore, contrary to ‘sojourner theory’ (point 3), entrepreneurship is not necessarily a better way of accumulating wealth than paid employment. While some studies claim that immigrants earn more in self-employment than in paid employment (Borjas, 1986; Lofstrom, 2002), the income experiences of immigrants can vary considerably, and sometimes entail disadvantage (Borjas and Bronars, 1989; Portes and Zhou, 1996). An important aspect to this debate appears to be the duration of residence in the host country. Both Brock and Evans (1986) and Lofstrom (2002) estimate that longer self-employment spells in the host country by immigrants eventually reverse initial earnings disadvantage relative to natives. This contrasts with immigrant employees, whose relative earnings disadvantage tends to persist throughout their lifetimes. Another problem for the sojourner theory is that many immigrants ultimately choose to remain in the host country, whatever their original intentions were. Fairlie and Meyer

Ethnic entrepreneurship and immigration

179

(1996) report that immigrants who had been in the USA for over thirty years had higher self-employment rates than immigrants who had been in the USA for less than ten years and who were presumably more likely to be sojourners.22 Immigrant selfemployment rates which increase with length of residence might reflect not only the positive income-duration relationship mentioned above, but also other factors. These might include: (i) greater knowledge of labour markets, tastes of ethnic groups, and institutions within the host country; (ii) accumulation of wealth required for entry into entrepreneurship and growth of a venture; and (iii) greater access to factors of production. At any event, findings that immigrant entrepreneurs have longer periods of residence than non-entrepreneur immigrants (Hjerm, 2004) cast serious doubt on sojourner theory. Point 4 above, which stresses restricted opportunities for immigrants in paid employment, has received somewhat less research attention. There are certainly assertions that disadvantage in the formal labour market pushed Koreans in Atlanta and Chinese in Vancouver into self-employment (Min, 1984; Wong and Ng, 2002). But there is little hard evidence that this operates on a systematic, large-scale basis. By way of indirect evidence, it is noteworthy that Koreans and Cubans are not only remarkably successful immigrant entrepreneur groups in the USA, but also comprise numerous people who were formerly professionals and entrepreneurs in their home countries. It might therefore be less surprising that these groups have done so well in the USA, though their self-perceptions can apparently entail status loss and feelings of failure (Godley, 2006). The importance of language is highlighted by evidence of a German-language income premium to immigrants to Germany of 38 per cent (Constant and Shachmurove, 2006). To the best of my knowledge, research has yet to explore points 5 and 6 above. Point 7 has not received much attention either, although US evidence suggests that self-employed immigrants and natives have similar industry distributions – apart from there being more natives in US construction and professional services, and fewer in retail relative to immigrants (Fairlie and Meyer, 2003). To summarise, the evidence provides relatively little support for the theoretical determinants of immigrant entrepreneurship listed above. This highlights serious limitations in our current understanding of this issue. Nevertheless, empirical research has uncovered several other determinants. A notable study is by van Tubergen (2005), who pooled census data from three ‘classic immigration’ countries (Australia, Canada and the USA) and labour force surveys from fourteen EU countries. The data spanned a range of years from 1980 to 2002 and contained data on 150,000 male immigrants from about 180 origin groups in seventeen destination countries. Van Tubergen (2005) found that immigrants from non-Christian countries of origin have the highest self-employment rates. Higher unemployment rates among natives increase the probability of immigrants being self-employed; and self-employment rates are higher among immigrant groups that are small, highly educated and have a longer settlement history. Constant and Shachmurove (2006) also report that high local unemployment rates are associated with immigrant self-employment – perhaps in response to fewer opportunities in the formal labour market. In most respects, though, the

180

Selection

determinants of immigrant entrepreneurship appear to be rather similar to those of native entrepreneurship (Constant and Zimmerman, 2006).

5.4.3

The effects of immigration on entrepreneurship

Two effects of immigration on entrepreneurship are considered below. One is the effect of immigration on native entrepreneurship and entrepreneurial earnings. This can be a politically charged issue in the host country. The second is the effect of immigrant entrepreneurship on the propensity for immigrants to become entrepreneurs when they return to their home country. In theory, the effects of immigration on native entrepreneurship could be either positive or negative, depending on whether immigrants have a greater impact on the supply of entrepreneurship or on the demand for the goods produced by entrepreneurs. In the first case, the effects are negative, since they compete down incomes in entrepreneurship and hence reduce the incentives for natives to become entrepreneurs. In the second case, the effects are positive, since greater demand increases incomes in entrepreneurship and hence strengthens everyone’s incentives to turn entrepreneur. To date, one of the few studies to estimate the effects of immigration on native entrepreneurship is Fairlie and Meyer (2003). Using 1980 and 1990 US Census data, and measuring entrepreneurs as the self-employed, these authors estimated that between 0.37 and 0.85 native men and between 0.09 and 0.19 native women are displaced by each self-employed immigrant.23 The effects of immigration on average native selfemployment earnings were found to be positive, however. The strongest displacement effects were found among less educated native workers, suggesting that self-employed immigrants might be pushing marginal low-income self-employed natives out of selfemployment, thereby increasing average earnings in the native group as a whole. As an illustrative example, Fairlie and Meyer (2003) cite cheap immigrant-owned restaurants driving out cheap native-owned restaurants, while leaving high-end native-owned restaurants unaffected. Turning to the second issue, there is growing evidence that migrants who return to their home country are more likely than average to become entrepreneurs shortly afterwards (whether or not they were entrepreneurs in the host country), and to use savings accumulated abroad as start-up capital.24 For example, Dustman and Kirchkamp (2002) found that one-half of a sample of Turkish emigrants returning from Germany started a micro-enterprise within four years of resettling in Turkey, using money saved while working abroad. Woodruff and Zenteno (2007) emphasise the importance of migration networks for Mexican entrepreneurs. Instrumenting migration networks by the aggregate rate of migration to the USA in an individual’s state of birth in Mexico, Woodruff and Zenteno (2007) report a strong and significant positive effect of migration networks on investment and profits in Mexican enterprises. Finally, it is interesting to speculate on the role of immigration policy with respect to entrepreneurship. Immigration selection is less reliant on skill-based screening criteria in the USA compared with Canada and Australia (Schuetze and Antecol, 2006).

Ethnic entrepreneurship and immigration

181

Yet relative earnings of immigrant entrepreneurs are higher in the USA than in these comparator countries. Arguably, then, immigration policy may have less impact on participation in entrepreneurship and entrepreneurial performance than other factors do, such as the tax and welfare system, or the size of the domestic market. 5.5 Appendix 5.5.1

The Borjas (1986) decomposition Let pˆ t,j denote the predicted probability of entrepreneurship for a representative member of cohort j at time t, and let pˆ t,j+10 be the probability of entrepreneurship at t for a representative member of a cohort who arrived in a host country ten years later (say) than members of j. Then     (5.1) pˆ t, j − pˆ t, j+10 = pˆ t, j − pˆ t−10, j + pˆ t−10, j − pˆ t, j+10

measures the cross-section difference between members of different cohorts at time t. The first term on the RHS of (5.1) measures the ‘within cohort j’change in entrepreneurship probability since year t − 10 (‘the assimilation effect’). The second term measures the change in entrepreneurship probability for immigrants with the same number of years’ experience since immigration (the ‘cross-cohort effect’). Borjas recognised that (5.1) could be biased if changing aggregate labour market conditions altered the attractiveness of entrepreneurship for everyone. To control for secular changes in broader labour market conditions, Borjas suggested decomposing changes in immigrant entrepreneurship relative to native-born entrepreneurs, the latter being denoted by subscript j
:     pˆ t, j − pˆ t, j+10 = pˆ t, j − pˆ t−10, j − pˆ t, j
− pˆ t−10, j
    + pˆ t−10, j − pˆ t, j+10 − pˆ t−10, j
− pˆ t, j
. (5.2) The first term in square brackets is a refined measure of the assimilation effect. It measures the change in entrepreneurship propensities of a given cohort net of the change experienced by a similar native-born cohort. Likewise, the second term measures the cross-cohort effect on entrepreneurship propensities net of economy-wide changes experienced by native-born workers between t − 10 and t. Using a sample of US Census data on male workers, and selecting t = 1980, Borjas reported substantial variation in the magnitude of the two terms of (5.1) and (5.2) across ethnic groups. Yet two common features held for almost all groups. First, the first term on the RHS of (5.2) indicated a strong assimilation effect, suggesting that the relative attractiveness of self-employment increases the longer the individual has been living in the USA. Second, the cross-cohort effect (the second term) usually indicated a greater propensity for more recent immigrant cohorts at that time to choose self-employment relative to earlier cohorts. This is reflected in the higher self-employment rates among more recent immigrants.

182

Selection

An advantage of Borjas’ decomposition technique is that it separates two important and distinct aspects of immigrant self-employment propensities. But it has yet to be widely adopted by researchers in the field, the occasional exception (e.g. Schuetze and Antecol, 2006) notwithstanding.

Notes 1. See, e.g., Clark and Drinkwater (1998) for evidence from British Census data; and Fairlie (1999) and Robb and Fairlie (2007) for US evidence. 2. See Fairlie (2006b), Lofstrom and Wang (2007) and Robles and Cordero-Guzman (2007). 3. Contrast Light (1984) and Waldinger (1986) with Borjas (1990) and Bates (1994b). 4. Sowell (1981), Moore (1983b) and Metcalf et al. (1996). 5. See Moore (1983b), Borjas and Bronars (1989), Fujii and Hawley (1991) and Clark and Drinkwater (1998). 6. This could occur if entrepreneurs’ profits are an increasing function of entrepreneurial ability x: πj = π(xj ) (j = {M , NM }), with ∂πj /∂x > 0 ∀j. To see how, let the distribution function of x, namely G(x), be the same for each group. Recall that wNM > wM because of employer discrimination. Denote the marginal entrepreneur in each ethnic group, i.e. who is indifferent between paid employment and entrepreneurship, by x˜ NM and x˜ M, respectively. These individuals are defined by the equalities wNM = π(˜xNM )

7.

8.

9.

10.

11. 12. 13.

and

wM = π(˜xM ).

Now wNM > wM ⇒ π(˜xNM ) > π(˜xM ), i.e. the minority marginal entrepreneur is less able, and less well remunerated, than the non-minority marginal entrepreneur. There are also more M than NM entrepreneurs, since 1 − G(˜xM ) > 1 − G(˜xNM ). In contrast, employers do appear to discriminate against individuals with previous criminal convictions, who are significantly more likely to be self-employed: see Fairlie (2002) for evidence from the NLSY (National Longitudinal Survey of Youth). A striking statistic, cited by Robb and Fairlie (2007), is that fully one-half of all black households in America have net worth (measured in terms of housing equity, savings, retirement and mutual funds accounts and other assets) of less than $6,200. As it stands, a higher proportion of M s than NM s is predicted to enter entrepreneurship in the Coate and Tennyson (1992) model – which is contradicted by the empirical evidence for blacks cited earlier. Coate and Tennyson (1992) showed that this prediction can in principle be overturned if entrepreneurial ability is partly determined by human capital investments since, facing employer discrimination, it is rational for Ms to acquire less human capital than NMs. If this reduces M entrepreneurs’ returns πM by more than it reduces wM (see chapter 13), then fewer Ms might choose entrepreneurship than NMs. Bond and Townsend (1996) report scant use of bank finance: in fact, only 11.5 per cent of respondents use it. Half of the respondents in this study used only personal resources, primarily personal savings, to start their businesses; a majority cited ‘lack of need’ as the main reason for not trying to get some kind of loan. Bond and Townsend (1996) concluded that formal sector loans are insufficiently flexible to meet the needs of business borrowers in poor neighbourhoods, with a clear lack of interest from the community, while informal sector funds are insufficient to meet all needs. See Knight and Dorsey (1976), Bates and Bradford (1992), Bates (1997) and Coleman (2005). Jones, McEvoy and Barrett (1994), Bank of England (1999) and Smallbone et al. (2003). Light and Bonacich (1988), Aldrich and Waldinger (1990) and Yoon (1991).

Ethnic entrepreneurship and immigration

183

14. Also consistent with this argument, Borjas and Bronars (1989) estimated significant positive selection into self-employment amongst whites, significant negative selection amongst Hispanics and Asians (see also Flota and Mora, 2001), but zero selection amongst blacks. 15. Some studies offer support for this hypothesis (Boyd, 1990, 1991 – for blacks; Le, 2000; Flota and Mora, 2001; Lofstrom, 2002) but others fail to find significant effects (Borjas and Bronars, 1989; Boyd, 1990 – for Asians; Yuengert, 1995; Razin and Langlois, 1996; Clark and Drinkwater, 1998, 2000, 2002). Fairchild (2008) is a recent study finding a significant negative effect, with both black and white self-employment rates being higher in less segregated areas, though blacks gain more than whites from lesser segregation in this respect. 16. See Pickles and O’Farrell (1987), O’Farrell and Pickles (1989), de Wit and Winden (1989) and de Wit (1993). Exceptions are Carroll and Mosakowski (1987) and Clark and Drinkwater (1998). Cultural factors might deter female ethnic self-employment though. Clark and Drinkwater (2000) found that in Britain, ethnic female self-employment rates were substantially below those of males, except among Chinese people. 17. Some studies find that poor English language skills increase self-employment participation in the USA and the UK (Boyd, 1990 – among Asians but not blacks; Fairlie and Meyer, 1996; Portes and Zhou, 1996; Clark and Drinkwater, 2002) but others find the opposite (Evans, 1989; Flota and Mora, 2001; Lofstrom, 2002). 18. Flota and Mora (2001) provide US evidence; Constant and Zimmermann (2006) provide German evidence. 19. To make this precise, consider the probit equation (3.7), and (suppressing the intercept purely for notational clarity) let βˆNM denote the estimated coefficients of that equation obtained using a purely non-minority data sample. Then for a set of personal characteristics denoted by Wi , the predicted probability of M self-employment given the same returns to characteristics across ethnic groups would be  (βˆ
Wi ) NM pˆ M = , (5.3) n i∈M

20. 21. 22. 23. 24.

where n is the sample size, (·) is the cumulative distribution function of the normal distribution, and where the summation takes place over all persons in the minority group, M . Different versions of (5.3) have been suggested by Clark and Drinkwater (1998), Clark et al. (1998) and Borooah and Hart (1999), but with similar qualitative findings. See Fernandez and Kim (1998) for US evidence; Hammarstedt (2006) for Swedish evidence; and Schuetze and Antecol (2006) for evidence from the USA, Canada and Australia. See Fairlie (2006b), Lofstrom and Wang (2007) and Robles and Cordero-Guzman (2007). For complementary evidence, see Lofstrom (2002), Schuetze and Antecol (2006) and Vinogradov and Kolvereid (2007). Earlier work by the same authors (Fairlie and Meyer, 1996) detected insignificant effects of immigration on numbers of black self-employed natives. See Ilahi (1999), McCormick and Wahba (2001), Dustman and Kirchkamp (2002) and Mesnard (2004) for evidence from Pakistan, Egypt, Turkey and Tunisia, respectively.

6

Female entrepreneurship

An important topic on the entrepreneurship research agenda is gender differences between men and women entrepreneurs. There is growing awareness that, for a variety of reasons, women face different opportunities and constraints in entrepreneurship from men; and that these considerations affect their participation and performance in entrepreneurship. As with ethnic minority entrepreneurship, there has been a concern that women’s prospects in entrepreneurship might be shaped by discrimination. Discrimination in the workplace might promote entrepreneurship among women in an effort to escape it, while discrimination against women by lenders might impede entrepreneurship by restricting access to finance. This chapter has the following structure. The first section describes some basic facts about female entrepreneurship, starting with cross-country evidence about its prevalence and some descriptive findings about gender differences in industry composition and personal characteristics. The second section explains how family factors bear on female entrepreneurship, including marriage, household production and child-rearing. The third section is devoted to the performance of women entrepreneurs, in absolute terms and relative to men. A gender performance gap is identified, and several potential explanations are discussed. The fourth section briefly treats the subject of women and entrepreneurial finance. The final section concludes. 6.1

Some basic facts about female entrepreneurship

Whether entrepreneurship is defined in terms of new venture creation, business ownership or self-employment, a higher proportion of men than women engage in entrepreneurship in all developed economies, despite a recent trend increase in female entrepreneurship in many of them. Among American workers aged twenty-five and over in 2003, for example, 9.2 per cent of women and 15.5 per cent of men were selfemployed. Yet over the period 1975–95, female self-employment grew by 60 per cent compared with 20 per cent for men (Budig, 2006). To put this in perspective, female selfemployment grew from about one-quarter of the US non-agricultural self-employed workforce in 1975 to about one-third by 1990 (Devine, 1994a). According to Aronson (1991) this represents a continuation of a trend increase since at least 1955, when the 184

Female entrepreneurship

185

female share of the US self-employed workforce stood at only 12 per cent. However, recent CPS data suggests that the growth in the relative female self-employment rate ceased in the 1990s (Wellington, 2006). Women entrepreneurs are also in the minority in Europe, although female selfemployment rates in the EU vary considerably, from just over 20 per cent in the UK, Ireland and Sweden to 40 per cent in Belgium and Portugal (Cowling, 2000). In Italy, women are one-half as likely as men to enter but are significantly more likely to exit self-employment. While Italian men tend to be more likely to enter self-employment following layoff or for career advancement considerations, women are more likely to enter from inactivity or unemployment (Rosti and Chelli, 2005). There is spotty evidence about trends in female entrepreneurship in European countries, though analogous to Wellington’s (2006) American study, Ajayi-obe and Parker (2005) observed stasis in the female self-employment rate in Britain in the 1990s, which has settled at around 7 per cent of the female workforce there. The gender gap in entrepreneurship is less pronounced using alternative definitions of entrepreneurship. For example, Brush (2006, chap 23.3) calculated that 10.6 million of the approximately 22 million firms in the USA are at least 50 per cent owned by women, and claims that women’s participation in entrepreneurship is growing in most countries. Data from the Centre for Women’s Business Research also show that American women formed businesses at twice the national rate over 1997–2002 (Greene et al., 2007). By 1997, female-owned firms employed 23.8 million workers, an increase of 262 per cent over 1987–92 (Coleman, 2007, p. 303). The figures mentioned so far relate to numbers of participants in entrepreneurship. Another salient consideration is work hours, in particular whether women entrepreneurs participate in a part-time or full-time capacity. Evidence shows that American and British self-employed women are more likely to be part-time workers than women employees or men in either employment category. For example, calculations made from the BHPS data set by the present author reveal that females comprise only 16 per cent of the full-time, but as much as 70 per cent of the part-time, self-employed workforce in contemporary Britain. In the USA, self-employed women work either very long or very short working weeks, leading to greater overall dispersion in weekly work hours than is observed for other groups (Devine, 1994a, 1994b; Budig, 2006). Indeed, NLSY data over 1979–98 reveal two apparently distinct groups of women entrepreneurs, with one group working less than fifteen hours per week in their business and the other working more than forty-one hours per week. Budig (2006) argues that the first group of women engage in non-professional self-employment primarily to limit their work hours and juggle family commitments (possibly because non-professional waged jobs tend not to be family-friendly); whereas the second group enters professional selfemployment to advance their careers. These arguments are based on evidence that family factors (especially children) help explain women’s entrance into non-professional self-employment, but (as in the case of males) have little impact on female entry into professional and managerial self-employment. Budig concluded that ‘women entering self-employment in professional occupations are more similar to their male peers in

186

Selection

self-employment than they are to women entering non-professional self-employment’ (2006, p. 2235). The industry composition of business ventures also has a gender dimension. Women entrepreneurs remain heavily over-represented in a few industry sectors, notably ‘sales’ and ‘other services’ (which include financial, insurance and real estate; professional services; and business services: Bates, 1995). Few self-employed women are found in the construction industry, which remains dominated by males. The concentration of self-employed women in service industries is particularly pronounced, being greater than that of either female employees or either category of males. To the extent that entrepreneurs identify opportunities to start businesses of similar types and in similar industries in which they formerly worked, one might be able to explain some of the concentration of women entrepreneurs in particular industry sectors. But as Aronson (1991) pointed out, the secular growth in services pre-dated the growth in female self-employment and hence cannot explain it. And between 1975 and 1990, growing female entrepreneurship in America was accompanied by few changes in occupational or industrial concentrations. Turning next to human capital, the evidence shows that British self-employed women possess more advanced educational qualifications on average than self-employed men. The exception is the employer group, where men have higher academic educational qualifications on average and women tend to have more vocational qualifications (Cowling and Taylor, 2001). The education qualifications of American nascent entrepreneurs are similar by gender (Kepler and Shane, 2007). On an individual level, the bulk of US studies find that advanced education and skills are associated with female entrepreneurship.1 Consistent with this observation is the pronounced concentration of female paid employees in clerical and administrative jobs which require less advanced qualifications and which yield work experience that is arguably ill-suited to switching into entrepreneurship (Boden, 1996). There are gender differences in other aspects of human capital, too. Descriptive evidence from the business studies literature suggests that male entrepreneurs are more likely than female entrepreneurs to have education and experience with technical and managerial elements (Brush, 1992). The same review showed that male entrepreneurs are more likely than female entrepreneurs to have been employed prior to start-up; they also have more prior work and business experience on average (see also Kepler and Shane, 2007). American women business owners tend to be older than women employees, although this difference narrowed during the 1980s (Devine, 1994a). Other reviews of the business studies literature on female entrepreneurship suggest that female entrepreneurs’ demographic characteristics, business practices and motivations appear to be broadly similar to those of male entrepreneurs. There are in contrast greater differences between genders in relation to choices of industry sector, as noted above; in amounts of start-up capital deployed; in management style; and in sales performance (Birley, 1989; Brush, 1992, 2006). Some of these differences, especially those relating to access to finance and financial performance, will be explored further below.

Female entrepreneurship

6.2

187

Family factors

There are several interrelated family issues which affect female entrepreneurship. They include marriage and household production, and child-rearing. Despite their inter-relatedness, it is convenient to discuss them sequentially. 6.2.1

Marriage and household production It is scarcely possible to understand female entrepreneurship without taking account of the importance of marriage (or cohabitation, which is subsumed under ‘marriage’ hereafter). Self-employed American women are significantly more likely than women employees to be married with a spouse present and to be covered by somebody else’s health insurance.2 To give some illustrative numbers, ‘a much larger proportion of selfemployed women (81 per cent) than of wage-and-salary women (64 per cent) is married, and these proportions are fairly stable on an age-adjusted basis’ (Lombard, 2001, p. 214). The figures are especially stark for self-employed women who employ others (‘job creators’), with nine out of every ten female job creators in Britain being married, predominantly to income-earning husbands. These women have fewer children on average than their sole-trader female counterparts (Cowling and Taylor, 2001). In terms of entry propensities, the presence of a self-employed husband approximately doubles the probability that an American woman switches into self-employment (Bruce, 1999). These findings should be interpreted with caution. Evidence that marriage and children are significantly and positively associated with entrepreneurship among women is sometimes interpreted to mean that women use entrepreneurship primarily to balance work and family commitments, whereas men use it to advance their careers (Carr, 1996). But positive relationships between marriage and entrepreneurship are also observed for men (q.v. Table 4.1 in chapter 4). Chapter 4 outlined several reasons why marriage is associated with entrepreneurship for both men and women. A couple of additional reasons might be germane to women in particular. One relates to the fact that husbands systematically contribute less to household production than women, even when the woman is working, and whether or not they are engaged in paid employment or entrepreneurship.3 With less time to spend on formal work, part-time entrepreneurship can offer married women the flexibility to juggle home and work commitments. Second, to the extent that women end up in low-paying non-professional service sector self-employment, marriage can provide a valuable financial cushion which possibly makes such choices viable. On a related point, explored further in section 6.3 below, men operate larger and more profitable enterprises on average than women do. In which case, marriage might enable women to work as ‘joint entrepreneurs’ in their husband’s business; and husbands might be more tolerant of a wife taking time out for household production than a typical employer. It is less clear what has the greatest impact on a woman’s choice to be an entrepreneur: the presence of a husband of any kind; the presence of a husband who is an entrepreneur; or the presence of a husband who earns a high income. Some studies find that both the presence of a self-employed husband and a husband’s income have separate positive

188

Selection

effects, with the former being stronger than the latter (Caputo and Dolinsky, 1998; Bruce, 1999). Budig (2006) observed that self-employed women with self-employed husbands tend to be non-professionals in clerical and service occupations, especially childcare, which might offer better opportunities to combine work and family responsibilities. Other researchers however have detected larger effects from incomes than from husbands’participation in entrepreneurship (MacPherson, 1988; Clain, 2000). For example, in Clain’s (2000) analysis of full-time, full-year workers drawn from 1990 US Census data, marital status was significantly associated with self-employment for men, while for women, self-employment was primarily associated with their spouse’s income. It is important not to overlook the possibility that female entrepreneurship can also promote male entrepreneurship, through higher female self-employment incomes (Lombard, 2001) or knowledge spillovers (Parker, 2008a). With the exception of Parker (2008a), the impact of female entrepreneurship on male entrepreneurship has received relatively little attention from researchers. 6.2.2

The impact of children

If externally provided childcare is costly, entrepreneurship might be attractive to women by offering a flexible way of combining work and domestically provided childcare. This is especially true if women can run their businesses from home: ‘time spent in home-based work may have a non-monetary benefit, such as being able to spend more time with one’s children’ (Edwards and Field-Hendrey, 2002, p. 172). The notion that entrepreneurship offers the flexibility to combine childcare and other domestic duties with a work schedule will be referred to hereafter as the ‘flexibility hypothesis’. A crude test of this hypothesis is to look for a positive association between children and female entrepreneurship; more sophisticated tests entail in-depth comparisons of the work and childcare choices of women entrepreneurs and non-entrepreneurs. On the face of it, the evidence would seem to support the crude version of the flexibility hypothesis. There is now compelling evidence that being married and raising children are both strongly associated with self-employment among women.4 Having children under six years old has the greatest impact on the probability that women are self-employed, especially among homeworkers.5 CPS, NLS and NLSY data from the USA all suggest that these effects are most pronounced for highly educated women (Wellington, 2006). The effects of children on male entrepreneurship appear to be weaker. While some studies record that both men and women are likelier to be entrepreneurs the more children they have, others have found little impact from family variables on male selfemployment participation.6 However, the propensity to become self-employed is higher among women with husbands who provide childcare (Caputo and Dolinksy, 1998). Before proceeding any further, a note of caution should be sounded. The number of children, which is commonly used as an independent variable in binary choice models of female entrepreneurship, is likely to be endogenous. We can ask, for example, whether women choose to become entrepreneurs in order to have children more easily,

Female entrepreneurship

189

or whether having children makes entrepreneurship more attractive to them. Reflecting two-way causality, respondents of the GSS (General Social Survey) have approximately 0.2 to 0.4 more actual and expected children if they are self-employed rather than employees (Broussard et al., 2003). Entrepreneurs might choose to have more children in part to increase the probability of a family member making a good match with running the family firm. This suggests that family structure and entrepreneurship are jointly determined. This is an intriguing possibility which has been insufficiently explored in the literature to date: further research on this issue is called for. Some more detailed evidence is also consistent with the flexibility hypothesis. According to US Census data from 1980, American self-employed mothers are more likely to look after their child (more than 20 per cent did) when in work than their employee counterparts are (less than 2 per cent did). Furthermore, 20 per cent of selfemployed women work from home, compared with just 6 per cent of self-employed men (Carr, 1996). Edwards and Field-Hendrey (2002) report that home-based workers are more likely to choose self-employment (63 per cent) than on-site workers (33 per cent). Apart from joint production of market and household goods and services, home-based self-employment enjoys several advantages over on-site employment, including the absence of employer monitoring and office rental costs. And while the desire for independence and monetary rewards are important for both men and women, British women seem to value the freedom of choosing when they work more than men do (Hakim, 1989a). Interestingly, Wellington (2006) observed no trend in the relationship between child-rearing and female self-employment over the last decades of the twentieth century, contrary to what might be expected if technological change made home-working – an important form of self-employment for women according to Edwards and Field-Hendrey (2002) – more pervasive. Lombard (2001) measured the demand for flexibility in terms of the variability of weekly work hours, relative to a reference point a year before. Lombard (2001) estimated a positive association between a woman’s demand for flexibility and the probability of being self-employed; the association was strongest for women with young children. According to Lombard, relatively low self-employment rates among black women may be associated with their apparent preference for a standard working week. Other findings lend less support to the flexibility hypothesis, however. For example, ECHP data shows that self-employed workers in Europe do not necessarily spend more time with their children than employees do. Indeed, self-employed women actually spend less time caring for children on average than women employees in several European nations (Gustafsson and Kjulin, 1994; Hildebrand and Williams, 2003). This is because self-employment is associated with long work hours as well as flexible work schedules (see chapter 12). 6.3

Performance of women entrepreneurs

This section commences by documenting evidence about the gap in earnings and profitability between men and women entrepreneurs. For the most part these constructs

190

Selection

are measured in terms of self-employment income, or less commonly business owners’ profits. I then propose several plausible explanations for the performance gap and review the available evidence. Finally, I briefly discuss alternative measures of performance, namely growth and survival.

6.3.1

The gender earnings gap It is now well established that self-employed women earn less on average than selfemployed men do – as well as less than employees of either gender. For example, using 1983 SIPP (Survey of Income and Program Participation) data, Haber et al. (1987) calculated that the ratio of US median full-time female to male self-employment incomes was only 0.30, compared to a ratio of 0.60 in paid employment. Among fulltime, full-year workers in 1990, self-employed women earned 73 per cent of the annual income of women wage-and-salary workers, whereas self-employed men earned 107 per cent of the annual income of their employee counterparts (Devine, 1994a). In the same study, self-employed men were more frequently observed to work in high-paying sectors than women, including executive, administrative, managerial and precision artisan jobs. In contrast, self-employed women were more frequently found in lowerpaying service and retail sectors. These sectors are more competitive and also exhibit lower business survival rates. Earnings in non-professional female self-employment can be very low indeed, especially for women who were not previously employed. For instance, Budig (2006) calculated an average hourly self-employed wage of only $4.08 for young American women over the period 1979–98. Women who do not work or who are in non-professional wage jobs, or who have family commitments, are the likeliest to become non-professional self-employed (Clain, 2000; Budig, 2006). Self-employed women perform somewhat better on average in some other countries, with female/male earnings ratios reaching 87 per cent in the case of Australia (OECD, 1986). But in developing countries, women receive substantially lower average returns than men, especially if they are not employers (e.g. see Honig, 1996, 1998, for Jamaican evidence). The same story is observed when profits and sales turnover are compared across gender lines. Regarding profitability, for example, the average employer-firm owned by a woman generates only 78 per cent of the profits of the comparable business owned by a man (Robb and Wolken, 2002). Women also generate lower sales turnover relative to men – even in same-industry comparisons (Loscocco and Robinson, 1991; Chaganti and Parasuraman, 1996).According to Loscocco and Robinson (1991), women generate only one-quarter of men’s average business receipts. More recent work based on data from the 1998 NSSBF compiled by the US Federal Reserve shows that men-owned businesses are twice as large on average as women-owned firms in terms of sales and assets (Coleman, 2007). Aronson (1991) offers an interesting historical perspective on female relative selfemployment incomes in the USA. In the inter-war and early post-war periods of the twentieth century, self-employed women earned more than women employees did –

Female entrepreneurship

191

although the comprehensiveness of these data is limited. What is clear, however, is that between 1955 and 1984 the relative earnings position of self-employed women declined steadily relative to their employee counterparts and also relative to self-employed men. It is less clear whether this decline reflects greater part-time participation by selfemployed women or a relative worsening of the average human capital of women choosing self-employment. While the inclusion of incorporated self-employed workers in sample data tends to close the income gap of male self-employees relative to male employees, it makes relatively little difference to the female self-employed– employee income gap, partly because relatively few women run incorporated businesses (Aronson, 1991). For the self-employed generally, it should be borne in mind that self-employment incomes usually omit important dimensions of job remuneration, including health care coverage. Self-employed women have relatively scant job-related health care coverage compared with women employees and men in either self-employment or paid employment (Devine, 1994a). 6.3.2

Explanations of the earnings gap

What explains the gender entrepreneurial earnings gap? Possible explanations include gender differences in: • • • •

Human capital Start-up capital Social capital Preferences, motivations and household production

Consider human capital first. On average, women entrepreneurs possess fewer years of experience than women employees or men in either entrepreneurship or paid employment (Aronson, 1991; Lee and Rendall, 2001). Women entrepreneurs also tend to have more diverse backgrounds than their male counterparts and are more likely to set up a business without having a track record of achievement, vocational training or experience (Watkins and Watkins, 1984). It is notable in this regard that age does not seem to enhance the earnings of self-employed women, in contrast to men (Clain, 2000). Start-up capital appears to be another important consideration. A widespread observation is that women entrepreneurs operate smaller businesses on average, utilising less capital and finance from banks and other lenders than men do.7 For example, Watson’s (2002) analysis of Australian Federal Government data on 13,551 male SME business owners and 875 female SME business owners revealed that women business owners earn similar rates of return on equity and assets to male business owners; but the female business owners invested less to start with, which explains why they ended up with lower absolute income and profits than the male business owners. This suggests that women entrepreneurs might be no less able than males in making returns, but simply invest less to start with. Possible reasons why women entrepreneurs tend to use less start-up capital than men are explored in the next section.

192

Selection

There is less evidence which relates social capital to the gender earnings gap. What we do know is suggestive only: female entrepreneurs include more women in their social networks whereas male entrepreneurs’ networks are more gender-balanced (Brush, 2006, p. 620). If homophily is a drawback to business performance, homogeneous networks might explain why relatively few women start businesses in the first place, and why they perform less well compared with men (Renzulli et al., 2001). However, this begs the question of why women entrepreneurs form homophilious social networks along gender lines (and whether gender homophily is really a handicap at all). Another open question is why male nascent entrepreneurs appear to examine more ideas and gather more new information when pursuing new start-up opportunities than female nascent entrepreneurs do (Kepler and Shane, 2007). There is much more research on motivations and preferences of women entrepreneurs. As has been noted several times in this book already, self-reported motivations are prone to all sorts of biases, so they should not be taken too literally, especially as less successful entrepreneurs might justify poor ex post performance by claiming they were not driven by a need for success in the first place. Yet it is interesting to observe that among nascent entrepreneurs, who have by definition not yet experienced success or failure, male nascent entrepreneurs and employees rate financial success and innovation as more important motives than women do (Carter et al., 2003; DeMartino and Barbato, 2003). In contrast, women rate more highly the flexibility of work schedules and family considerations – especially women with young children, in contrast to men for whom children did not matter in this respect (Boden, 1999a). Women entrepreneurs also declare stronger motivations in terms of personal interests; a desire for self-fulfilment; and job satisfaction than their male counterparts (Jurik, 1998). Buttner and Moore (1997) examined the motivations of 129 women executives and professionals who left large organisations to become entrepreneurs. Challenge, autonomy and family–work balance were the declared motivations; self-fulfilment and goal achievement ranked higher among these women than profits and business growth. Greater risk aversion among women than men might also explain lower average returns to women entrepreneurs, as they would prefer to occupy positions further down the expected profit/risk frontier than their male counterparts (Watson and Robinson, 2003; Kepler and Shane, 2007). Such women could for example be less likely than men to found businesses which are technologically intensive (Kepler and Shane, 2007). Women seeking to minimise risk might also reject growth strategies which endanger the security of the business and thereby their family’s financial position. In an exploration of this hypothesis, Watson and Robinson (2003) observed that women entrepreneurs in Australia received significantly lower average annual profits than men, but also bore substantially less business-related risk than men. Consistent with finance theory, risk-adjusted returns were similar by gender.8 In fact, the evidence about gender differences in risk attitudes is sparse and inconclusive. While Jianakoplos and Bernasek (1998) adduced survey evidence suggesting that women are more risk-averse than men, experimental evidence (albeit from a small

Female entrepreneurship

193

sample of students) from Schubert et al. (1999) detected similar levels of risk aversion among men and women. Hundley (2001a) attempted to test some of the competing explanations against each other by applying an Oaxaca decomposition to gender-specific earnings functions.9 Hundley found that the most important explanatory variables were housework, work hours and the number of young children, which together accounted for between 30 per cent and 50 per cent of the American annual self-employment gender earnings differential. This suggests that women earn less than men do because they spend less time managing and developing their businesses. The next most important explanatory variable according to Hundley (2001a) was industrial sector, which accounted for between 9 and 14 per cent of the gender self-employment earnings differential. That reflects the concentration of women in the less lucrative personal services sector, and their under-representation in the more remunerative professional services and construction industries. Differences in capital explained only 3–7 per cent of the differential, and other variables (including experience) seemed to be even less important. Researchers tend to find a positive role for human capital, especially age, schooling and work experience, when trying to explain performance differences among women entrepreneurs (i.e. not relative to men: Schiller and Crewson, 1997; and Coleman, 2007). For example, Coleman (2007) estimated that education and work experience were the main drivers of profitability in women-owned businesses, whereas financial capital was the key variable influencing the profitability of men-owned businesses. This finding is consistent with other evidence that in general human and financial capital may be substitutes for each other (Chandler and Hanks, 1998). There is, on the other hand, mixed evidence about women’s responsiveness to the prospect of financial returns in entrepreneurship. According to various estimates of earnings functions, self-employed American women would earn more than their counterparts in paid employment if they quit and made the switch.10 This suggests that self-employed women have above-average skills (Devine, 1994b). In contrast, Lombard (2001) estimates that women’s self-employment choices are in fact largely explained by their incomes in self-employment relative to what they could have expected to earn in paid employment. Another influence on women’s occupational choice of entrepreneurship might be gender discrimination. If women are discriminated against in paid employment (e.g. because they hit a ‘glass ceiling’), and so earn lower returns to their characteristics there, entrepreneurship can become more attractive. Boden (1999b) explored this possibility using longitudinally matched CPS data. In evidence suggestive of the glass ceiling idea, Boden (1999b) estimated the returns-to-skill component of the gender wage differential and found that it significantly increased the probability that white women would transition from paid employment to self-employment. Some other covariates also helped to explain transitions, including wealth, the presence of young children, education and experience (all with positive effects) and the current wage in paid employment (with a negative effect). But wealth and education also predispose women to contemplate business careers in paid employment. That could merely suggest that both occupations

194

Selection

appeal to successful women. And other studies show schooling and experience to significantly reduce the probability that women make transitions into self-employment in the USA (Schiller and Crewson, 1997), while lucrative executive jobs in paid employment seem to be preferred to self-employment by highly educated women in Germany (Constant, 2006). Another problem with the notion of a ‘glass ceiling’ in paid employment is that female self-employment participation rates do not vary systematically by job skill level (Devine, 1994b). This is inconsistent with the hypothesis that skilled women choose self-employment to avoid a glass ceiling of limited earnings in self-employment. It also casts doubt on the idea that women use their employment skills as a launching pad for entry into entrepreneurship. However, rigorous tests of the glass ceiling are notoriously hard to devise, as researchers invariably lack objective empirical manifestations of this construct. Simple occupational choice models with labour market discrimination also struggle to explain gender differences in entry rates. To see why, suppose that women receive a lower wage, w(1 − δ) with 0 < δ < 1, than men, who get w. Following Rosti and Chelli (2005), suppose also that: (i) men and women have the same distribution of general ability x, namely G(x); and (ii) entrepreneurship incomes y(x) are increasing in x; and (iii) business survival is an increasing function of x. Then using the simple occupational choice framework of chapter 2, it follows that the average ability and hence survival prospects of female entrepreneurs are less than that of men. However, this simple framework also predicts higher entry rates among women, contrary to the evidence (q.v. above).11 Finally, self-employment seems to be a less effective vehicle for upward earnings mobility for American women than men. Men and women have similar transition rates into self-employment, but men are more likely than women to move up the self-employed earnings distribution (Holtz-Eakin et al., 2000).

6.3.3

Other performance gaps: growth and survival rates

Self-employed women do not under-perform relative to self-employed men merely in terms of their incomes. The same is also true in terms of their turnover, employment creation performance, growth rates and survival prospects.12 For example, Bosma et al. (2004) estimated that male Dutch business founders outperform their female counterparts in terms of survival, profitability and employment creation. Lohmann and Luber (2004) reported that 42 per cent of self-employed German women remain self-employed after five years; for German men the corresponding figure is 63 per cent. A key source of this difference appears to be unstable work histories, which predict lower female but not male survival probabilities. On a similar point, Lee and Rendall (2001) reported that white American women experience shorter spells in self-employment on average than white men, despite working similar numbers of spells. It should be borne in mind that inferior financial performance may reflect founder goals which are less materialistic, rather than constraints due to discrimination, for example.

Female entrepreneurship

195

Swedish evidence explains low growth rates of women-owned businesses in terms of the small scale of their ventures; their tendency to trade in the service sector; and their reliance on direct sales rather than more lucrative business-to-business sales (Du Rietz and Henrekson, 2000). Yet no significant gender differences have been observed in relation to the performance outcomes of nascent entrepreneurs (Kepler and Shane, 2007), or among long-established business owners, in terms of employment growth (Fischer et al., 1993; Chaganti and Parasuraman, 1996) or business survival rates (Kalleberg and Leicht, 1991; Brüderl and Preisendörfer, 1998). So it appears that gender disadvantage occurs less at the early start-up or mature stages of business development, and more in the intermediate stages. 6.4 Women and entrepreneurial finance

Evidence from a variety of sources shows that men and women entrepreneurs use similar sources of finance, but that women tend to use smaller amounts of external finance on average than men do. For example, Carter and Rosa (1998) estimated that in Britain, women entrepreneurs use one-third less external finance than men on average. Similar evidence has subsequently been obtained for a variety of other countries.13 Kim’s (2006) analysis of US small business loan applications reveals that equally owned mixedgender small businesses occupy an intermediate position in terms of success in obtaining finance compared with women-owned (lowest) and men-owned (highest) businesses – although larger samples from the NSSBF reveal that women-owned American small businesses gain similar access to line-of-credit loans from commercial banks to menowned businesses (Haynes and Haynes, 1999). Access to start-up finance is important because low levels of initial capital may hamper the growth and survival prospects of ventures later on. Some Norwegian evidence bears out this hypothesis (Alsos et al., 2006). However, researchers should be careful to control for the innately limited scale of some opportunities, which can also explain this relationship. It has been widely observed that women entrepreneurs seek and use less formal and informal equity finance than men entrepreneurs. Women own over 35 per cent of US businesses, but receive at most only 5 per cent of venture capital funding (Brush et al., 2003). And only about 10 per cent of business angels in the USA and the UK are women (Amatucci and Sohl, 2007), although their characteristics do not seem to differ by gender (Mason, 2006; Harrison and Mason, 2007). American women entrepreneurs seek angel finance at a lower rate than men entrepreneurs, but with a similar probability of receiving investment; they are more likely to seek, but are less likely to receive, finance from female than from male angels (Becker-Blease and Sohl, 2007). As noted at the start of this chapter, research into gender aspects of entrepreneurial finance is sometimes motivated by a concern about possible gender discrimination by lenders. For example, lenders might perceive women to possess fewer of the characteristics deemed favourable to business success, and so restrict finance for that reason (Buttner and Rosen, 1988). To test this hypothesis, Fay and Williams (1993) conducted

196

Selection

an interesting experiment by mailing bank loan officers in New Zealand identical loan applications whose sole difference was the gender of the hypothetical applicant in the supporting documentation. Fay and Williams found that significantly more noncollege-educated women than men had loans that were declined or restricted (32 per cent compared with 10 per cent), but there was no difference along gender lines for college-educated loan applicants. A problem with this study, however, is that the sample sizes are very small, being only around fifty for each sex. Evidence based on larger surveys tend to reveal modest or negligible differences in the way that bank loan officers respond to gender in the context of business loan applications (Wilson et al., 2007). Other evidence casts greater doubt on the notion that women face discrimination in the credit market. It is noteworthy that nowadays relatively few men or women entrepreneurs seem to believe that gender discrimination underlies loan decisions; and women business owners report similar levels of satisfaction with their relationships with banks as men do (Buttner and Rosen, 1988; Carter and Rosa, 1998). If discrimination does not explain loan decisions, why do women start businesses with less capital on average than men? Several explanations can be proposed. First, women entrepreneurs tend to run smaller businesses in sectors with fewer market growth opportunities, so it may be optimal for both lenders and entrepreneurs to commence with limited capital (Orser et al., 2006). As noted earlier in the chapter, women entrepreneurs often start businesses in the service and retail sectors, which need less capital and have higher exit rates, making banks reluctant to lend large amounts of finance. Riding and Swift (1990) provide some tentative Canadian evidence suggesting that these factors may be sufficient to explain the bulk of apparent gender inequalities in bank lending. Similarly, once one controls for the size and age of a venture and the particular industry sector it is located in, gender differences in access to and terms of finance (including rates of loan approvals, co-signature requirements, interest rates on loans and lines of credit, and collateral requirements) tend to disappear (Coleman, 2000; Orser et al., 2006). Specifically, Coleman (2000) noted that many women run firms which need little external finance and possess little ‘inside’ (i.e. business-related) collateral, so necessitating the entrepreneur to put up more ‘outside’ (i.e. personal) collateral. This does not reflect discrimination but a simple and understandable desire on behalf of lenders to protect their capital at risk. Furthermore, to the extent that women receive lower wages, have broken job histories and reside in property held in the name of their husband, women entrepreneurs’ access to self-finance is likely to be limited. Yet women entrepreneurs might still demand less external finance than men because they want to run smaller businesses for family and lifestyle reasons. The same outcome can also arise if women are more risk-averse or less financially aware, reflecting their previous industry experience (Carter and Rosa, 1998). As yet, I am unaware of research which distinguishes cleanly between these different explanations of gender loan-size differences. Evidently this is a fruitful topic for future research.

Female entrepreneurship

6.5

197

Conclusion

Research on female entrepreneurship continues to expand and develop. Much of the groundwork describing ways that women differ from men in entrepreneurship has now been completed, and our knowledge base has improved considerably over the last twenty years or so. We now know that, relative to their male counterparts, women entrepreneurs tend to operate smaller businesses; use less finance; operate in more competitive industry sectors with lower returns and survival rates; and undertake more household duties in addition to their business activities. Despite some valuable work based on earnings decompositions, it remains unclear precisely what the relative importance of these factors is in explaining different patterns of female participation and performance in entrepreneurship, and much more research remains to be done on this issue. But there is an unmistakable trend away from blaming lower average levels of female entrepreneurial performance simply on gender discrimination. It is striking that relatively little evidence exists of discrimination against women in the credit or labour markets. As each year passes, it seems that there is less talk of discrimination and more discussion of ‘liberal’ or ‘social’ feminist rationales for lower rates of entrepreneurship among women. While the liberal feminist rationale stresses structural barriers for women, which prevent them from gaining experience necessary to access resources, the social feminist rationale emphasises different perspectives, goals and choices specific to women. This author’s view is that a particularly important difference between men and women in entrepreneurship stems from the persistent tendency of women to spend more time in household production and child-rearing activities than men. Until and unless this situation changes, women will continue to opt for forms of entrepreneurship which result in lower average returns compared with men. Given a finite number of hours in the day to run a business, undertake household production and take leisure, women are effectively forced to operate in entrepreneurship at a smaller scale and on a more part-time basis than their male counterparts. This chapter has demonstrated that a labour market perspective is invaluable for analysing female entrepreneurship. Both this and complementary approaches will no doubt be used to push the boundaries of research in the future. Several important questions remain open, although some of them are likely to be difficult to answer. For example, to what extent does a ‘glass ceiling’ in paid employment push women into entrepreneurship? More generally, how many women are being pushed into entrepreneurship as ‘necessity’ rather than ‘opportunity’ entrepreneurs? On a different tack, rather than dwelling on gender differences alone, as much of the literature does, it may be time to deepen the analysis of interactions of gender with other factors, such as skill, ethnicity and family background. Researchers continue to explore whether women entrepreneurs face binding financial constraints, despite the fact that there continues to be little empirical support for this notion. Much less is known, in contrast, about the financial literacy of women entrepreneurs. This is potentially important because in general women are less

198

Selection

financially literate than men (Lusardi and Mitchell, 2006). Running and growing a new business is a complicated affair, so public policy directed at improving financial literacy might do more to help women succeed in business than policies which try to relax minor or non-existent borrowing constraints. In this regard, improving access to credit for women entrepreneurs is one of the goals of the US Small Business Administration, through targeted loan guarantees. Other policies include gender-targeted federal procurement, training, counselling and mentoring programmes (Brush et al., 2003; and see chapter 17). In the author’s opinion, future researchers should also investigate alternative gender-sensitive policies. In view of the point about long work hours, for example, it is tempting to suggest that female entrepreneurship policies should include a child-care component. But this begs the question about whether such a policy would also be appropriate for women choosing to work in paid employment. It is unclear why policy-makers should treat the two occupations differently in this respect, especially as entrepreneurship seems to provide greater opportunities for flexible work arrangements than paid employment does. Notes 1. See Macpherson (1988), Evans and Leighton (1989a), Connelly (1992), Devine (1994a), Bates (1995), Boden (1996, 1999a), Carr (1996), Wellington (2001) and Edwards and Field-Hendrey (2002). 2. See Devine (1994a, 1994b), Lombard (2001), Taniguchi (2002) and Budig (2006). 3. This finding is based on time-use surveys and responses to household surveys: see Longstreth et al. (1987), Boden (1999a) and Bond and Sales (2001). 4. For US and UK evidence about the importance of marriage for female entrepreneurship, see Devine (1994a, 1994b), Robinson and Sexton (1994), Carr (1996), Schiller and Crewson (1997) and Cowling and Taylor (2001). Caputo and Dolinsky (1998) provide a dissenting view. On the importance of the presence of children for female entrepreneurship, see Macpherson (1988), Evans and Leighton (1989a), Connelly (1992), Boden (1996), Carr (1996), Caputo and Dolinsky (1998), Wellington (2001) and Kepler and Shane (2007). 5. See Connelly (1992), Boden (1996, 1999b), Wellington (2001, 2006) and Edwards and FieldHendrey (2002). 6. Compare Wong (1986), Connelly (1992), Boden (1999a) and Caputo and Dolinksy (1998) with Boden (1996) and Carr (1996). 7. See Aronson (1991), Carter et al. (1997) and Watson (2002). 8. Watson and Robinson (2003) defined risk-adjusted return in terms of Sharpe’s ‘reward to variability’ ratio, i.e., the quotient of mean annual profits and their standard deviation. 9. To see how this works, consider the earnings function ln yij = βj
Xij + uij , where the i subscript denotes an individual and a j subscript denotes gender: j = f indexes females and j = m indexes males. Here y represents annual earnings, X is a vector of explanatory variables and u is a random disturbance term. Consider a particular explanatory variable Xijk . Letting an overbar denote sample means and a hat denote a regression estimate, we can write 
 
X ˆ ˆ ln ym − ln yf = βˆm (6.1) mk − Xfk + βm − βf Xfk . The first term on the RHS of (6.1) is the part of the average earnings difference attributable to explanatory variable Xijk . The other term is the part of the earnings difference that is unexplained

Female entrepreneurship

10. 11.

12.

13.

199

by Xijk . Note that one can alternatively write an analogous expression using βˆf
in the first term of the RHS of (6.1). Since the choice is arbitrary, results based on this decomposition method should be quoted for both calculations. Macpherson (1988), Devine (1994b) and Clain (2000). Proof: denote the marginal male (respectively, female) entrepreneur by xˆ m (respectively, xˆ f ), which solves y(ˆxm ) = w (respectively, y(ˆxf ) = w(1 − δ)). Hence for common G(x) it follows that xˆ f < xˆ m and so the average ability and survival chances of female entrepreneurs are less than those of men: E(x|x ≥ xˆ f ) < E(x|x ≥ xˆ m ). But 1 − G(ˆxf ) > 1 − G(ˆxm ) implies that this simple framework predicts higher female entry rates. For evidence relating to developed economies, see Chaganti and Parasuraman (1996), Carter et al. (1997), Schiller and Crewson (1997), Boden and Nucci (2000), Bosma et al. (2004) and Lohmann and Luber (2004). For evidence relating to developing economies in southern Africa, see McPherson (1995, 1996). The same findings also emerge from reviews of the US and UK business studies literature: see Brush (1992) and Rosa et al. (1996). Other evidence relates to the USA (Coleman, 2000, 2007; Kim, 2006); Canada (Orser et al., 2006); the Netherlands (Verheul and Thurik, 2001); and Norway (Alsos et al., 2006).

Part II Financing

7

Debt finance for entrepreneurial ventures

Part I of this book explored various factors that bear on the willingness of individuals to try entrepreneurship. Part II recognises that sometimes individuals have limited opportunities to become entrepreneurs, because of difficulties in raising sufficient finance. Finance is needed to develop prototypes, purchase working capital and marketing services, and defray initial operational and living expenses. Most start-up finance in developed countries is in the form of personal equity (‘selffinance’), i.e. finance supplied by the entrepreneurs themselves. According to the Bank of England (2001), 60 per cent of start-ups in Britain use self-finance. The remaining funds are raised from external sources. According to the Bank, about 60 per cent of external finance is raised through debt-finance contracts (comprising bank overdrafts and term loans) followed by asset-based finance (e.g. leasing: around 20 per cent). A similar picture applies in the USA. Also important is family finance, at around 10 per cent of external finance on average, whereas venture capital (equity finance) tends to play only a very minor role for most entrepreneurs (between 1 and 3 per cent on average). This chapter focuses on the implications for entrepreneurship of raising debt finance. Chapter 8 discusses various other sources of finance, including venture capital. If lenders and entrepreneurs were both perfectly informed about every aspect of new entrepreneurial ventures, and if financial markets were flexible and competitive, then all ventures with positive net present value (NPV) would be funded. Also, it would not matter whether lenders or entrepreneurs undertook the ventures. However, this idealised scenario rarely happens in practice. Many entrepreneurs complain that they are unable to obtain enough funding, or sometimes any funding at all, for what they believe are viable ventures. Such entrepreneurs claim to be ‘credit rationed’. Credit rationing has received considerable attention in policy circles since the publication of influential reports by the Federal Reserve System (in 1958) in the USA, and the ‘Bolton’ and ‘Wilson’ reports (HMSO 1971, 1979) in the UK. These reports contended that there was a general shortage of financial capital to fund new start-ups and to expand existing small businesses. While a subsequent UK government report (HMSO, 1991) contended that problems of access to start-up finance had by then been largely resolved, the opposite conclusion was drawn by a US report on small business finance (SBA, 1996). 203

204

Financing

Another potential problem is that imperfections in the capital market can result in there being either too few or too many entrepreneurs in equilibrium – judged in terms of the efficient use of the economy’s resources (‘social efficiency’). The present chapter investigates these issues. Along the way, the roles of collateral, loan sizes and lender– borrower relationships will also be discussed. I will leave aside topics such as what lenders and borrowers think about each other; what induces entrepreneurs to apply for loans; and how entrepreneurs can write business plans to improve their chances of successfully obtaining loans. These issues are covered in any number of texts within the business studies literature. For notational brevity, lenders will be referred to simply as ‘banks’ hereafter. This chapter emphasises the importance of asymmetric information in the entrepreneurial credit market. Information is often asymmetric because although entrepreneurs may have accurate information about the quality of their risky proposed ventures and their abilities and commitment to execute them, banks cannot perfectly distinguish the quality of loan applications. Reasons for this include the lack of a track record for new ventures, and prohibitive costs of acquiring reliable information about them. Banks cannot simply ask entrepreneurs to declare truthfully their commitment and ability, because loan applicants lacking these attributes have incentives to overclaim about them. Furthermore, some entrepreneurs may be unwilling to reveal hard information about the potential of new innovative ideas which could be expropriated by others; while investors cannot disclose information about their involvement with other customers who are developing competing products. The problems are compounded because there is no equivalent institution to a credit-rating agency for entrepreneurs. Instead, banks have to rely on their own imperfect screening devices. It will be assumed below that although banks can screen entrepreneurs into groups defined by some observable characteristics, there is invariably also some residual imperfect information which forces them to pool, at least initially, heterogeneous risk types together within each group. Although established corporations also face asymmetric information problems in relation to project finance, there are several reasons to expect these problems to be more pronounced for new entrepreneurial ventures. First, most start-ups have limited funds compared with incumbents, making the absence of finance more critical to their survival prospects. Second, entrepreneurs are more likely than incumbents to develop radical innovations about which little information is available, whereas corporations tend to develop incremental innovations funded from retained profits (Bhide, 2000). In short, these considerations suggest that incentive issues come to the fore in entrepreneurial finance, necessitating the use of sophisticated financial contracting practices. Smith and Smith (2003) point out that the severity of asymmetric information is just one of several reasons why entrepreneurial finance differs from corporate finance. Another difference is that entrepreneurs’ consumption and investment decisions cannot be separated from their financing decisions, as is the case in classical corporate finance. Savings and retained profits are important sources of funds for entrepreneurs, needed in part to signal commitment to their ventures; and entrepreneurs typically

Debt finance for entrepreneurial ventures

205

possess undiversified investment portfolios, making them unusually prone to risk (see chapter 9). Other differences include end-game strategies. Unlike ‘infinitely lived’ corporations, some entrepreneurs plan at the outset, and often at the behest of investors, how they will exit within a finite time to realise value, a process called ‘harvesting’.1 Furthermore, unlike corporations, entrepreneurs often have to buy in outside management expertise; and they rarely separate ownership from control, at least initially. The next section provides some background information about entrepreneurial credit markets, as well as useful terminology used in the remainder of the chapter. The section after presents theoretical arguments for the existence of ‘credit rationing’, ‘redlining’ and ‘under-investment’. This material treats the possible causes of these phenomena; their effects on efficiency and the equilibrium number of entrepreneurs; and the scope for corrective action by policy-makers. The implication of these models is that credit market imperfections cause there to be too few active entrepreneurs for the social good. Theoretical rebuttals of the credit rationing hypothesis are presented in the third section, together with some counter-rebuttals. An important issue discussed here is the possibility that agents can write richer contracts which reveal hidden information and thereby eliminate market imperfections. The fourth section discusses some alternative models, in which over-investment occurs. The implication here is that credit market imperfections can cause there to be too many active entrepreneurs for the social good. The final section discusses other possible sources of inefficiency in entrepreneurial credit markets, and concludes. 7.1

Background and useful terminology

This section has two goals. First, it explains some of the salient aspects of entrepreneurial credit markets. Second, it introduces some useful terminology relating to asymmetric information which is deployed in the rest of the chapter. 7.1.1

Background Below, five salient aspects of entrepreneurial credit markets are explained: ‘collateral’, ‘limited liability’, ‘relationship lending’, ‘credit scoring’ and the optimality of debt as a financial contract.

Collateral is an asset belonging to a borrower that can be seized by a bank if the borrower defaults on a loan extended by the bank. Unlike large established firms which can pledge company assets (‘inside collateral’), most entrepreneurs can only pledge their own personal assets (‘outside collateral’), typically their house. Most models of debt finance assume that banks automatically capture inside assets in the case of default, so explicit analyses of collateral tend to be of the outside sort. In practice, banks typically incur costs when seizing collateral. This friction is built into several models of entrepreneurial finance discussed below. Collateral is widespread. Berger and Udell (1990) reported that nearly 70 per cent of all US commercial and industrial loans are made on a secured basis, while Cressy Collateral

206

Financing

(1993) found that 95 per cent of UK business overdrafts in excess of £20,000 were fully collateralised. Collateral serves several possible purposes. It can: • Persuade sceptical banks about the worth of ventures (Chan and Kanatos, 1985). • Encourage entrepreneurs to supply optimal amounts of effort (Boot et al., 1991) or to operate less risky ventures (Stiglitz and Weiss, 1992). • Enable entrepreneurs to make credible promises about repayment, since their own assets are at stake (Hart and Moore, 1994, 1998; Hart, 1995). • Enable banks profitably to re-negotiate debt when borrowers default (Bester, 1994). Entrepreneurs operating ventures with nugatory ‘inside collateral’ who know that banks prefer to re-negotiate than try to seize that collateral have incentives to default even when they are successful. Banks can remove this incentive if they have access to defaulters’ outside collateral, allowing the benefits of debt re-negotiation to be realised by both entrepreneurs and banks. • Serve as a screening or signalling device to help able entrepreneurs obtain more favourable borrowing terms (see below). On theoretical grounds, banks are predicted to request more collateral from riskier borrowers. Collateral compensates banks for bearing risk, which can be exacerbated when entrepreneurs are over-optimistic, choose excessively risky investment projects, or engage in opportunistic behaviour.2 On the other hand, if borrowers who post more collateral do so to signal higher ability (see below) they may be rewarded with lower interest repayments. The available evidence from the USA points to a net positive relationship between collateral requirements and project risk, although two recent British studies report negative and U-shaped relationships.3 Entrepreneurial real estate serves both as a productive factor of production and a source of collateral. In the presence of borrowing constraints, a subsidy to entrepreneurial real estate can in principle increase output, employment and welfare (Jin and Zeng, 2007). However, the effectiveness of this policy is undermined if it can be exploited by non-entrepreneurs. And when entrepreneurs are over-optimistic, collateral can actually decrease efficiency. This is because in a perfectly competitive credit market, collateral reduces the competitive interest rate banks charge, encouraging rash investment choices by over-optimistic entrepreneurs who overestimate the probability of success and so undervalue the cost of collateral (Manove and Padilla, 1999). Limited liability Entrepreneurs can obtain some protection from creditors by incorporating their businesses. Incorporation affords the entrepreneur limited liability status. In a legal sense this means that an entrepreneur whose company fails is liable only up to the value of the business assets, i.e. creditors cannot claim their personal wealth. Many credit market models of entrepreneurial finance include the limited liability condition in the form of a lower bound on entrepreneurs’ wealth in the downside state. As a general point, this gives debt contracts less incentivising power, and as we will see below leads directly to the possibility of some kinds of credit rationing.

Debt finance for entrepreneurial ventures

207

In practice, however, even limited liability does not perfectly protect entrepreneurs’ assets. Lenders who anticipate that entrepreneurs will switch funds from the firm to their own personal consumption can require entrepreneurs to secure loans on their own assets at the outset (Berkowitz and White, 2004). This clearly restricts the value of limited liability in practice. So does the threat of future disagreements with independent shareholders if incorporation is motivated by the objective of taking the venture public at a later date (Boot et al., 2006). There is some evidence that entrepreneurs with moderate wealth and capital are the most likely to incorporate, as the costs of incorporating are only worth bearing at a larger scale, while the wealthiest are the least likely to bump up against the limited liability constraint and are also the least risk-averse (Horvath and Woywode, 2005). ‘Relationship lending’ is another important feature of entrepreneurial credit markets. It describes how banks and entrepreneurs can form relationships whose value is reflected in loan terms which vary systematically over time. Relationship lending has the potential to gradually reveal to the bank ‘soft’ information about the enterprise and the entrepreneur. It is sometimes claimed that entrepreneurs rely more on relationship lending to convey messages about their creditworthiness than on the signalling, monitoring, and bonding mechanisms used by larger firms (Petersen and Rajan, 1994; Berger and Udell, 1995). A central question is whether borrowers who have built a relationship with a bank receive more favourable loan terms (e.g. lower interest rates and lower collateral requirements) than new borrowers. They might do so, for example, if lower interest rates provide entrepreneurs with incentives to exert greater effort at the crucial early stages of the relationship (Boot and Thakor, 1994). Much of the available evidence does indeed suggest that repayment terms are more favourable for entrepreneurs in longer bank relationships.4 On the other hand, if banks can lock entrepreneurs into relationships, they can offer low initial interest rates to lure them in, before charging them higher interest rates later on to exploit ‘informational rents’.5 Alternatively, subsequent rounds of finance which make borrowers spread their collateral thinly over a larger aggregate loan size might pose a greater risk for banks, which will also be reflected in interest rates which rise as relationships mature (Hanley and Crook, 2005). Relationship lending

If relationship lending is associated with personal relationships, then credit scoring is associated with the converse. Credit scoring is an automated underwriting technique used by banks to assess small business loan applications. It yields a single quantitative measure, or score, which represents an estimate of the applicant’s creditworthiness with respect to a requested loan. Applications receiving scores above a given threshold receive funding; those with scores below it do not. Credit scoring models usually weight several characteristics strongly, including personal credit history (which is strongly predictive of a venture’s loan repayment prospects) and other personal information such as the borrower’s monthly income, outstanding debt, financial assets, collateral, employment tenure, home-ownership status and cash-flow Credit scoring

208

Financing

prospects. Large banks tend to make greater use of credit scoring methods than small banks do (Berger and Udell, 2003; Berger et al., 2005). As a cheaper and arguably more accurate screening device than traditional personbased lending methods, it is hypothesised that credit scoring enables banks to: reduce the interest premium on loans; eliminate any credit rationing arising from the cost of processing loans (Thakor and Calloway, 1983); and expand the availability of more risky high-yield credit. Some evidence supports these arguments. For instance, Frame, Srinivasan and Woosley (2001) found that credit scoring is associated with an 8.4 per cent increase in the portfolio share of small-business loans among large US banking organisations. Credit scoring has also been associated with higher loan risk levels and prices (Berger et al., 2005). As well as expanding the volume of credit, credit scoring has the advantage of replacing judgement and discrimination with objective criteria in loan application decisions. As noted in chapter 5, this might be of particular importance for African-American entrepreneurs who struggle to obtain debt finance. Evidence that loan-rejection rates increase at times of high turnover of loan officers suggests that loan officers do indeed use soft information (e.g. assessments of character and information from customers and suppliers) to inform their loan decisions (Scott, 2006). Credit scoring appears to be becoming more widespread. In financially developed economies, the average distance between lenders and small firms is increasing, while their methods of communication are becoming more impersonal (Petersen and Rajan, 2002). This trend appears to be driven by greater lender productivity rather than by the closure of bank branches. It has been argued that ‘ongoing improvements in ICT are reducing information costs and making timely interventions easier, thereby slowly breaking the tyranny of distance…in small business lending’(Petersen and Rajan, 2002, p. 2535). A trend for growing geographical distance between banks and entrepreneurs carries several implications. Quite naturally, it downgrades the importance of relationship lending and individual loan officer input. On the other hand, by continually expanding the size of the available credit market, problems of insufficient credit should be reduced, and concerns about entrepreneurs’ access to credit in response to bank mergers and adverse local shocks to credit availability should ease (Petersen and Rajan, 2002). Nevertheless, it seems likely that geographical proximity will remain an important factor for other types of financial intermediation which involve close monitoring and assistance, such as venture capital (Lerner, 1995). And local financial development is still important in some developed economies. For instance, in Italy a person’s odds of starting a business increase by nearly six per cent if they move from the least financially developed region to the most financially developed one (Guiso et al., 2004). Hence local expertise in screening projects may still be valuable in some places, notwithstanding the broader global trends identified by Petersen and Rajan (2002). Optimality of debt as a financial contract Finally, the common usage by

entrepreneurs of debt as a financial contract (the type of contract exclusively studied in

Debt finance for entrepreneurial ventures

209

this chapter) is an outcome which needs explaining. Several reasons for the optimality of debt finance have been proposed. By ‘optimal’ is commonly meant Pareto optimal. That is, debt is the financial contract which maximises entrepreneurs’ returns without reducing lenders’ returns. Perhaps the best-known rationale for debt finance is the ‘costly state verification’ argument, which recognises that banks incur costs whenever they check entrepreneurs’ actual as opposed to declared performance. With costly state verification, it is efficient for banks not to monitor non-defaulting ventures. By economising on state verification costs, debt becomes the optimal financial contract, by allowing lenders to offer finance more cheaply than costlier alternatives, such as equity.6 A different rationale is that debt finance enjoys the advantage over equity finance of providing partial insurance to entrepreneurs if the latter can renegotiate terms of the financial contract in the bad state (Fender and Sinclair, 2006). In fact, debt turns out to be the optimal contract under very general conditions, including when banks even lack information about the distribution of project returns, entrepreneurs’ types and entrepreneurs’ investment decisions.7 By giving banks all of the project returns if entrepreneurs default, debt contracts can also address moral hazard problems.8 Another argument is that short-term debt contracts can be a good ‘second best’ to contingent contracts (which switch control from optimistic entrepreneurs to realistic investors in bad states) when entrepreneurs over-optimistically overestimate their prospects of success (Landier and Thesmar, 2008). Over-optimists demand debt finance because they do not want to share their high expected returns with financiers as they would have to under an equity contract. If entrepreneurs are predominantly overoptimistic (see chapter 4), then debt finance will prevail over equity for this reason. But even if they are realists, abler types prefer debt finance over equity if they are most likely to succeed and so gain all of the upside returns (net of their fixed debt repayment). Under equity finance, in contrast, entrepreneurs have to share their upside returns. Although this advantage is less compelling for less able types, who have lower probabilities of venture success, these types also have to also request debt finance in order to avoid transmitting an adverse signal about their abilities. Hence in equilibrium every entrepreneur requests debt finance. 7.1.2

Terminology I commence with definitions of some salient concepts discussed in this chapter.

Definition 1 (Type I credit rationing) . Type I credit rationing occurs when some or all loan applicants receive a smaller loan than they desire at the quoted interest rate. Definition 2 (Type II credit rationing) . Type II credit rationing occurs when some randomly selected loan applicants are denied a loan altogether despite being observationally identical to applicants who receive one, and despite being willing to borrow on precisely the same terms; and when banks have such rationing as an optimal equilibrium policy.

210

Financing

This credit rationing typology follows Keeton (1979). In the case of Type II rationing, a rationed borrower might offer to pay a higher interest rate in order to obtain funds; but the last clause of the definition indicates that this cannot break the rationing outcome. The inability of entrepreneurs to borrow at the interest rate they think is appropriate is in contrast not a valid definition of credit rationing, because it fails to take into account the real-world cost of funds and informational constraints prevailing in the market. The manifestations of Type I and II credit rationing analysed below are ‘equilibrium’ rationing outcomes, in the sense that they can be expected to persist until a favourable shock impacts the credit market. In contrast, ‘temporary’ credit rationing is caused by transient imbalances between the demand for and supply of loans. This case is of less interest because such imbalances will eventually be eliminated as markets adjust. Neither temporary credit rationing, nor equilibrium credit rationing caused by governments fixing interest rates below market clearing levels by diktat (e.g. usury laws), will be discussed in what follows.9 Usury laws have no role in modern deregulated economies. Figure 7.1 illustrates both types of credit rationing. The interest repayment Dm would clear the market for credit, but the actual interest rate is stuck at Dcr < Dm , with an excess demand for funds of LD − LS. In the case of Type I rationing, LS is the equilibrium loan size for an individual entrepreneur: LD is the desired loan size. In the case of Type II rationing, LS is the number of entrepreneurs who obtain a loan: LD is the number of loan applicants. Reasons why the interest rate may not rise above Dcr to Dm to eliminate rationing are explored below. They have their roots in two problems arising from asymmetric information: adverse selection and moral hazard. In the present context, adverse selection is the name given to the situation where the average quality of the borrower pool worsens when the price of funds faced by entrepreneurs (the debt repayment, D) increases. For instance, suppose in the terminology of chapter 2 that entrepreneurs possess different abilities x which are known

Suppl y

Quantit y

DL L

m

SL

e D mand 0 D

cr

D

m

e D tb repa

Figure 7.1 The supply of and demand for loans

yment, D

Debt finance for entrepreneurial ventures

211

to themselves but not to banks. Entrepreneurs cannot credibly signal their ability to banks, because every entrepreneur has an incentive to claim they are of high ability in order to improve their chances of obtaining funds. (That is why this problem is often referred to as one of ‘hidden types’.) Banks do, however, know the average quality of the borrower pool, x. Suppose banks increase D. By increasing the cost of funds it is less profitable to be an entrepreneur. If the ablest entrepreneurs respond by dropping out of the market and deploying their abilities outside entrepreneurship (as in, e.g. case (a) of Figure 2.3, where the expected returns schedule EN shifts downwards by D), while the less able entrepreneurs do not, the average quality of the remaining pool of entrepreneurial borrowers decreases. Then we say that adverse selection occurs. A different possibility is that borrowers do not quit entrepreneurship or the credit market when D increases, but instead devote less effort to their enterprise since the returns to their effort are lower. For example, entrepreneurs might shirk or spend more time on pleasurable but less profitable pursuits, such as solving interesting technical problems at the expense of commercialisation activities. This is known as moral hazard, or alternatively as a ‘hidden action’ problem, since banks cannot directly observe, prevent or contract on actions like effort which are at the discretion of the entrepreneur. The cases just discussed suggest that the danger of adverse selection and moral hazard might give banks pause for thought before increasing the interest rate they charge entrepreneurs. Indeed, the next section will explain how these problems can directly lead to credit rationing, as well as several other forms of inefficiency in the credit market, including redlining and under- or over-investment. I will now formally define these terms. Definition 3 (Redlining) . Redlining occurs when a bank refuses to lend to any loan applicant from a particular group because the bank cannot obtain its required return from these borrowers at any interest rate. Definition 4 (Under-investment) . Under-investment occurs when some socially efficient ventures (i.e. ventures whose expected values are greater than those obtained from employing their resources in their best alternative use) are not undertaken by entrepreneurs. Definition 5 (Over-investment) . Over-investment occurs when some socially inefficient ventures (i.e. ventures whose expected values are less than those obtained from employing their resources in the next best use) are undertaken by entrepreneurs. Redlining is a distinct concept from credit rationing because with redlining no member of a group, whether defined by wealth, location or some other characteristic, can obtain any finance. Credit rationing, in contrast, is the random denial of credit to some, but not all, potential borrowers in a group. Under-investment refers to a situation where too few entrepreneurs are financed for the social good, while over-investment refers to a situation where too many entrepreneurs are financed. Under-investment is distinct from credit rationing and can occur in the absence of credit rationing. Conversely, with over-investment it would be preferable from the standpoint of social efficiency if

212

Financing

some of the currently active borrowers were to quit entrepreneurship and deploy their resources to alternative, more productive uses. 7.2 Theories of credit rationing and redlining 7.2.1

Type I credit rationing Models of Type I credit rationing have a long pedigree, stretching back to the 1960s. Few of the early models are discussed in the literature nowadays (for a survey see Parker, 2004, chap. 5). One exception is based on the idea that larger loan sizes increase the temptation for entrepreneurs to ‘take the money and run’, which increases default rates and hence banks’ costs of monitoring and seizing collateral. Banks’ best response is to restrict loan sizes, so Type I rationing results.10 Contemporary research tends to focus more on moral hazard as a potential source of Type I credit rationing. For example, if banks, aware that high repayment rates deter entrepreneurs’ effort and hence repayment rates, are unable to reduce interest rates in times of tight credit, their only way of protecting themselves against losses caused by moral hazard may be to restrict loan sizes (Piketty, 1997; Laffont and Martimort, 2002, chap. 4.8.4). Another form of moral hazard emerges because limited liability places a fixed lower bound on entrepreneurs’ downside risk, while reserving to entrepreneurs all of the upside benefits. If venture risks and returns are both functions of the loan size, entrepreneurs optimally request larger loans than banks wish to provide. Larger loans scale up the upside returns which entrepreneurs can capture, as well as the downside risk borne disproportionately by banks. Anticipating this, banks protect themselves by capping loan sizes.11 A formal proof of this result, based on a model by Bernhardt (2000), appears in the first part of the chapter appendix. Although it is possible to propose models with asymmetric information or overoptimism which reverse Type I credit rationing (resulting in entrepreneurs taking larger loans than they would choose under perfect information,12 the available evidence appears to be consistent with Type I credit rationing. According to Parker and van Praag (2006a), Dutch entrepreneurs who were interviewed in the mid-1990s obtained on average only four-fifths of the starting capital they requested. Over one-third of respondents claimed to have faced some degree of Type I credit rationing when they started up. As this was a self-selected sample of surviving ventures, the actual degree of Type I credit rationing might be even higher than this. On the other hand, Parker and van Praag’s (2006a) survey respondents might have exaggerated the difficulties they faced when starting up their ventures. For example, over-optimistic but objectively low-quality entrepreneurs might claim to be credit-rationed even though realistic banks rationally deny them funds (Hillier, 1998). This should caution against placing too much faith in self-reported measures of ‘credit rationing’. 7.2.2 Type II credit rationing, redlining and under-investment The ‘canonical’ model of Type II credit rationing is that of Stiglitz and Weiss (1981), referred to as SW hereafter. This model is based on asymmetric information about the

Debt finance for entrepreneurial ventures

213

values of new ventures. Entrepreneurs are assumed for simplicity to be risk-neutral, to have identical stocks of wealth, and borrow a unit of capital from a competitive riskneutral bank.13 Each entrepreneur operates one investment project with a given risk profile. Entrepreneurs know the risk of their projects but banks do not. It is assumed that there is no credible way for entrepreneurs to convey their private information to banks. As noted in the introduction to the chapter, imperfect information is taken to be residual in the sense that bank screening has already been performed. Although banks do not observe individual entrepreneurs’ types, they do observe the frequency distribution of these types in the population. It is important to note at the outset that without asymmetric information all entrepreneurs with positive NPV investment projects would receive finance from competitive lenders; no entrepreneur with negative NPV projects would be funded. Each active entrepreneur would pay an interest rate which reflected their own riskiness to banks. They would not be pooled with other borrowers. There would be no credit rationing, under-investment or redlining. There are two key features of the SW asymmetric information model. One is that ventures generate the same expected return, but some entrepreneurs operate projects with greater risk than others, in the sense that their project returns are a mean preserving spread (MPS) of other project returns (see chapter 2). The second important feature of SW’s model is that banks offer debt contracts. This assumption, together with limited liability, means that entrepreneurs have bounded downside risk. That is important because although entrepreneurship becomes less attractive for all entrepreneurs as mandated interest repayments D increase, entrepreneurs operating the riskiest projects are the least affected. That is because these entrepreneurs enjoy greater upside risk than entrepreneurs operating safe projects, and are less likely to end up paying D. Hence the first entrepreneurs to drop out of the market as D increases are entrepreneurs operating the safest projects. But these are the best customers from banks’ perspective. There is therefore adverse selection in the credit market. The pool of borrowers becomes riskier as banks increase D, which might reduce banks’ expected profits from doing so. A similar outcome can arise if entrepreneurs can choose the degree of risk of their projects, for then all entrepreneurs would rationally respond to a higher D by selecting riskier projects. The borrower pool once again worsens from banks’ perspective, warranting a higher interest rate and possibly restricted levels of lending. This particular mechanism exemplifies moral hazard, which has qualitatively similar implications to adverse selection. But in this case, restricted lending can cause entrepreneurs to engage in even more moral hazard, reducing lending and hence effort further in a downward spiral. The outcome can be the continued propagation of negative shocks, leading to a credit crunch (Minetti, 2007). There are three major implications of the SW model. One is under-investment. Asymmetric information causes the competitive loan rate to reflect loan applicants’ average riskiness, which may be so high that low-risk entrepreneurs drop out of the market first. If there were perfect information these entrepreneurs would face an interest rate which

214

Financing

reflected their risk (but not that of riskier borrowers) and that would be sufficient to entice them to become entrepreneurs (de Meza and Webb, 1987, Proposition 5(A)). The under-investment problem can be solved by subsidising interest income, which increases both the equilibrium number of entrepreneurs and social efficiency. A second implication of SW is that Type II credit rationing might occur. As noted above, the worsening pool of borrowers means that increasing the interest rate to satisfy an excess demand for loans might actually reduce bank profits, rather than increasing them as might be expected. Effectively, banks’expected returns may eventually become a decreasing function of the interest rate D, as the pool of entrepreneurial ventures becomes dominated by risky types. This is illustrated in part (a) of Figure 7.2, where Dcr , termed the ‘bank optimal’ interest rate, is the rate that maximises bank expected profits and which therefore holds under competition.14 In this case, the supply of funds also becomes a decreasing function of D above Dcr , since depositors receive the banks’ returns and so supply fewer funds when banks’ average portfolio rate of return ρ decreases. This is illustrated in part (b) of Figure 7.2, which shows how credit rationing can occur if there is a ‘high’demand for funds, and if Dcr < Dm , where Dm is the ‘marketclearing’ interest rate (i.e. which matches supply with demand). Here banks deny credit to L1 − L∗ randomly chosen borrowers who are observationally indistinguishable from those who do receive loans. In effect, credit rationing occurs because banks do not respond to an excess demand for funds by increasing the interest rate. Hence the excess demand persists. A formal derivation of the credit rationing result appears in the second part of the chapter appendix, which also shows why credit rationing can be only a possible, rather than an inevitable, feature of credit markets. Intuitively, if adverse selection effects are very weak, bank expected return functions will not deviate sufficiently from a monotonic increasing function to yield an interior bank-optimal interest rate. With no interior bank-optimal interest rate, there can be no credit rationing in this model. Moral hazard can also generate an interior ‘bank-optimal’ interest rate which is lower than the market-clearing rate – and therefore also credit rationing.As noted above, moral hazard can arise in several ways. For example, entrepreneurs could deliberately select riskier projects in response to a higher interest rate, once again reducing the quality of the borrower pool and discouraging banks from increasing the interest rate to clear the market (Stiglitz and Weiss, 1981). Or entrepreneurs could decrease effort in response to higher repayments which cut their profits, so reducing the value of business assets which banks can claim from failed entrepreneurs.15 Other reasons for a bank-optimal interest rate can also be proposed. First, a higher interest rate which decreases the demand for loans can reduce banks’ ability to exploit scale economies, and might also restrict their ability to leverage their assets profitably. Second, a different form of adverse selection can give rise to an interior bank-optimal interest rate if ‘honest’ entrepreneurs who resist the temptation to default opportunistically are the first to quit the loans market when the interest rate increases (Clemenz, 1986). As we will see in the next section, an alternative source of credit rationing which does not depend on an interior bank-optimal interest rate arises if banks can exploit a willingness of borrowers to expose themselves to the possibility of Type II credit rationing as a tool of efficient financial contracting (Besanko and Thakor, 1987a).

Debt finance for entrepreneurial ventures

215

Banks’ expected portfolio rate of ret

urn, –␳

(a)

0 D

D

cr

(b )

High demand

N umb er of loans, L

L 1

L

L*

S upply

2

Low demand

0 D 2

D

cr

D

m

D

Figure 7.2 Stiglitz and Weiss’ credit rationing model. (a) Banks’ expected returns (b) The market for funds

The third implication of SW’s model is the possibility of redlining. To see this, suppose for expositional clarity that banks can distinguish between three distinct groups of entrepreneurs. The groups are indexed by θ , where θ can be good (g), bad (b) and ‘ok’ (o). Following the logic of Figure 7.2, there is for each borrower group an interior bank optimal interest rate, denoted by Dθ , generating expected returns to banks of ρθ (Dθ ). Banks’ expected return functions for each group are illustrated in Figure 7.3. Let ρ ∗ denote the deposit rate, which must be unique in a competitive deposit market. As Figure 7.3 shows, group g will be fully served, and so will some members of the marginal group o; but a group b generating returns of ρb (Db ) < ρ ∗ cannot make a sufficient return to enable banks to compensate depositors at any interest rate. Hence no member of group b will obtain funds, even though these ventures may have aboveaverage expected social returns. This group is redlined. The only solution to redlining

216

Financing ␳␪

␳g (D) ␳* ␳o(D)

␳b (D)

0 D g

D

o

D b

D

Figure 7.3 Redlining

is to somehow induce an outward shift in the supply of funds schedule, since that reduces ρ ∗ . One possibility is universal government lending (Ordover and Weiss, 1981), although that policy is not necessarily welfare-enhancing. Obviously, periods of tight credit or ‘credit crunch’ (as occurred in 2008, for example) are associated with more extensive redlining. SW’s is not the only model of redlining. A different model, by Aghion and Bolton (1997), generates an opposite prediction to SW with respect to the effects of the cost of capital. In SW’s model, an increase in ρ exacerbates redlining, as we have just seen; but in Aghion and Bolton’s (1997) model, where individuals can choose to be borrowers (entrepreneurs) or lenders, borrowers respond to an increase in ρ and redlining by switching out of entrepreneurship and into lending, so reducing the size of the redlined group. Another notable feature of the Aghion–Bolton model is that the number of individuals caught in the redlining trap is predicted to progressively decline as economies accumulate wealth. This suggests that credit constraints are likely to be more prevalent in less-developed than in highly developed economies, and that economies eventually ‘grow out of’ wealth-based redlining.16 Further details about the Aghion–Bolton model can be found in chapter 9. 7.3

Rebuttals of the credit rationing hypothesis and counter-rebuttals

7.3.1 Rebuttals Although it has been hugely influential, SW’s model and other asymmetric information models of credit rationing, redlining and under-investment have been criticised on several grounds. The primary objection to these models concerns their assumption that

Debt finance for entrepreneurial ventures

217

heterogeneous entrepreneurs are pooled together with one financial contract. Instead, it is argued, entrepreneurs and banks have incentives to devise richer financial contracts which separate hidden borrower types. That destroys pooling outcomes on which the credit rationing result (and other types of inefficiency) depends. I will explain this idea first, before moving on to other criticisms, some of which are specific to the SW model. So far, banks have been assumed to specify financial contracts whose only term is the interest rate. As we have seen, the interest rate does not work efficiently because its indirect effects on the quality of loans conflict with its direct price (i.e. quantity) effects. In practice, however, other loan terms (such as collateral) might be available to banks. Entrepreneurs who are better risks from banks’ point of view have incentives to request loan contracts which condition on collateral in order to signal their ability to the banks and thereby separate themselves from more risky types with whom they are pooled (in a ‘pooling equilibrium’). That way, the less risky types can obtain a lower interest rate, reflecting their more favourable borrowing profile (in a ‘separating equilibrium’). Likewise, banks have incentives to offer a menu of different contracts to appeal to different borrower types, in order to steal a competitive advantage over their rivals. That is, banks have incentives to screen heterogeneous entrepreneurs and reveal their hidden types. For expositional ease, much of the discussion below will work with just two hidden entrepreneurial types, a good type g and a bad type b. The precise aspects that make them ‘good’ or ‘bad’ are unimportant here; it suffices merely to think of g as being less prone to default than b. The population shares of the two types will be assumed fixed and known; as before, it is the identities of each individual that cannot be observed by banks. Consider Figure 7.4 for the case of borrowers b and g and two contract terms: the gross interest rate, D, and another ‘bad’ (from borrowers’ perspective), denoted BAD. It might help fix ideas to think of collateral posted by entrepreneurs at the bank’s request as an example of BAD. Indifference curves Jb and Jg in (D, BAD) space show how the different borrowers are prepared to trade off one bad for another. Indifference curves closer to the origin are associated with higher borrower utility. The key point is that the different borrowers have differently sloped indifference curves. Good borrower types g have flatter indifference curves than bad types b because they are more likely to succeed and have to repay the bank D – and hence are more willing to endure more of BAD (e.g. risking more collateral) in return for a lower D. Conversely, bs are less willing to incur the cost of more BAD in return for a lower D because they know they are more likely to default and thereby avoid repaying D but incur the BAD (e.g. lose their collateral). A formal proof of these propositions in the context of collateral, from Bester (1985a), is given in the third part of the chapter appendix.17 Crucially, if Jb and Jg cross only once (the so-called ‘single crossing property’, which certainly holds in Bester’s (1985a) model – see the appendix), then contracts b = (Db , 0) and g = (Dg , BADg ) separate types and are consistent with banks’ ‘isoprofit’ lines (i.e. lines of constant break-even profit) πb and πg . That is, g would prefer a contract (D, BAD), just to the right of g along πg , in preference to b ; but b would

218

Financing

D b

⌫b

De b t repayment, D

␲b

D

␲g g

⌫g

J

0 BAD

b

J g

g

Another b ad, BAD

Figure 7.4 The use of two-term contracts to separate hidden types

prefer b to that contract. Hence if banks offer these two contracts – and under competition they must offer these contracts18 – then each borrower type will self-select into the contract that maximises their utility, so revealing their type. This equilibrium is incentive compatible, because each type does best under the contract that reveals their type. Masquerading as a different type will result in a lower payoff and so will not be chosen. Able entrepreneurs can signal or be screened using instruments other than collateral. Another BAD they might be willing to bear is foregoing the advantage of limited liability (Chamley, 1983). To see this, suppose that entrepreneurs are risk-averse and can choose between limited and unlimited liability debt contracts. Under symmetric information, all risk-averse entrepreneurs would prefer limited liability because it places a lower bound on their risky returns. But under asymmetric information, high-ability entrepreneurs might be prepared to endure the BAD of unlimited liability in order to signal their higher ability to banks. Entrepreneurs who successfully identify themselves to banks as good risks obtain a lower interest rate (risk premium), giving rise to a partially separating equilibrium like the one in Figure 7.4.19 Indeed, more able entrepreneurs might be willing to offer banks greater personal guarantees, signing away their rights to personal assets beyond the collateral they have pledged. This signal is credible because only able types would believe that they are unlikely to lose their personal assets this way, and ‘put their money where their mouths are’ by risking the consequences. A related possibility open to some ‘international entrepreneurs’ is to cross-list their companies in

Debt finance for entrepreneurial ventures

219

foreign countries where investor protection is stronger (and hence their own protection is weaker) in order to credibly signal their types (Mougin, 2007). Other BAD financial contract terms which convey strong signals about hidden entrepreneur types include short-maturity loans; issuance of convertible preferred stock; and agreements to offer banks forfeiture and anti-dilution provisions (Sahlman, 1990; Diamond, 1993). Able entrepreneurs should also be more willing than less able entrepreneurs to waste resources by requesting inefficiently large (or small) loan sizes in order to signal their types.20 Likewise, they should also be more willing to submit to high interest rates early on in a lending relationship, followed by lower rates later on, because they anticipate a greater probability of their business surviving to the later date when rates are low (Webb, 1991; Boot and Thakor, 1994). However, it should be acknowledged that multi-period lending separation devices can be difficult to implement in practice for reasons relating to imperfect commitment. Furthermore, moral hazard can undermine the viability of long-term credit contracts. For instance, if entrepreneurs learn about their abilities over time, the resulting reduction in uncertainty means they might divert effort into riskier entrepreneurial projects which can undermine the existence of a credit market altogether (Parker, 2007c). Another factor undermining long-term credit contracts is a dynamic inconsistency problem in which entrepreneurs need a stream of finance over several periods. If the borrowing relationship has a clear end, borrowers have an incentive to default in the final period. Anticipating this, banks will not lend in the final period, giving borrowers the incentive to default in the penultimate period. By backward induction, default incentives are peeled back until the mechanism unravels altogether – unless there is sufficient uncertainty about the end date, or if there is well-established progression from one loan tranche to the next. An important consideration is that few entrepreneurs run their businesses to satisfy outside shareholders. Being both owners and managers, entrepreneurs do not separate ownership and control. While this might give entrepreneurs a competitive advantage over corporate competitors whose separation of ownership and control generates agency problems, it turns out that separation of ownership and control can serve a valuable signalling role in credit markets. Banks favour projects where the entrepreneur employs a manager who is paid simply according to whether the project succeeds or fails. These ‘low-powered incentives’remove the incentives for entrepreneurs to take hidden actions which harm lenders. Not separating ownership and control could send an adverse signal such that in equilibrium only ventures in which ownership is separated from control receive finance (Acemoglu, 1998). Somewhat less plausibly, education has also been proposed as an entrepreneurial signalling mechanism in the credit market. According to this argument, gifted entrepreneurs may optimally acquire less education than ordinary individuals if this conveys a signal of strength about their innate entrepreneurial abilities. If innate entrepreneurial ability matters more for business success than formal education, and financiers reward more able entrepreneurs by offering them favourable credit contracts, ordinary people will not find it in their interests to emulate the gifted people, because

220

Financing

with their abilities they benefit more by taking more schooling and working in paid employment instead (Orzach and Tauman, 2005). Unfortunately, the evidence refutes this theory. Quite apart from the fact that entrepreneurs and non-entrepreneurs choose similar levels of schooling on average, innovating entrepreneurs who take less time to complete their university degree (arguably a direct measure of ability) are more rather than less successful in raising finance (Backes-Gellner and Werner, 2007). Furthermore, better-educated entrepreneurs receive more rather than less credit on average and can access a greater diversity of start-up funds (Parker and van Praag, 2006a; Astebro and Bernhardt, 2003). The foregoing discussion has shown that in principle a richer contract than one based on interest rates alone can separate hidden entrepreneur types. In principle too, type separation removes all information asymmetries and hence the possibility that asymmetric information-induced credit rationing can occur. This analysis readily generalises to more than two hidden entrepreneur types. For example, with three types, three different (D, BAD) contracts can be offered that accomplish separation. However, with two or more dimensions of unobserved borrower heterogeneity, more than two contract terms are needed to separate types. For example, with two dimensions of heterogeneity (e.g. different entrepreneurial abilities and different venture risks), contracts might need to specify an interest rate, collateral and a sub-optimal loan size, say. But the principle remains the same. In the limit, one can imagine banks enriching the menu of contracts with as many contract terms as it takes to reveal all of the hidden information across every dimension. This completes my discussion of signalling and screening criticisms of credit rationing models. The next subsection will explain how these criticisms can be criticised in their turn. The remainder of the present subsection turns instead to other drawbacks of SW and related models. A particular problem with the SW model is that it assumes, rather than derives, debt to be the optimal form of finance. In fact, it can be proved that equity is actually the optimal form of finance in the SW model, where entrepreneurs’ investment projects differ in terms of their risk profiles (Cho, 1986; de Meza and Webb, 1987). But if ventures were to be financed by equity, SW’s competitive equilibrium would no longer exhibit any credit rationing or inefficiency; and banks would randomise rather than fix interest rates as SW assumed (de Meza, 2002). It is therefore necessary to appeal to some factor outside the model which favours debt over equity finance if SW’s results are to remain relevant in a strict sense. That could entail, for example, high costs of writing equity contracts. But this logical inconsistency remains an unsatisfactory feature of SW’s model. A more powerful objection is that, if risk eventually gets resolved in the future, entrepreneurs wanting to operate risky projects would do best by exercising their valuable option to wait before investing. Hence they would become the first to drop out of the market when interest rates increase. Contrary to the SW model, this implies favourable selection in the credit market, ruling out credit rationing (Lensink and Sterken, 2001, 2002). Moreover, Type II credit rationing implies that borrower entrepreneurs’marginal

Debt finance for entrepreneurial ventures

221

costs of funds become infinite. Hence rationed entrepreneurs have an overwhelming incentive to avoid rationing by scaling down the size of their project, decreasing consumption or delaying the project to accumulate wealth for self-finance (Parker, 2000). De Meza and Webb (2006) argue that this makes the conditions for credit rationing very stringent, although credit rationing can still emerge if the values of indivisible projects decay sufficiently quickly with delay. Another criticism of credit rationing models is that other plausible alternative models exist which generate market-clearing outcomes instead (e.g. de Meza and Webb, 1987). There are also grounds for questioning how widespread credit rationing can be in practice when sources of finance other than debt contracts (e.g. trade credit, equity finance and leasing) are available. Evidence about credit rationing is reviewed in chapter 9. 7.3.2

Counter-rebuttals

Proponents of credit rationing have offered several objections to the criticism that sophisticated financial contracting can eliminate the pooling equilibria on which credit rationing depend. First, screening and signalling which lead to separating equilibria are unfeasible if extra contract terms are unavailable. For example, specifying BAD as collateral will be ineffective if borrowers have insufficient collateralisable wealth. Then pooling and credit rationing can emerge again. This may be an important point because it is known, for example, that lack of collateral is one of the major reasons why banks in developed countries refer borrowers to public sector loan guarantee schemes (see chapter 16). Second, BAD contract terms might not be monotonically related to preferences, violating the single-crossing property underlying Figure 7.4. SW pointed out that banks may choose not to request collateral if that reduces requested loan sizes and thereby increases the risk of failure of the ventures in their portfolio owing to under-capitalisation. Furthermore, decreasing absolute risk aversion implies that most collateral will be offered by wealthy borrowers who are most willing to operate the riskiest investment projects – thereby muddying the signal. In both cases, banks can do best by not requiring collateral to deal with the credit rationing problem (Stiglitz and Weiss, 1981, 1992). Other muddy signals emerge if entrepreneurs do not behave fully rationally. For example, if entrepreneurs are over-optimistic, as the evidence in chapter 4 suggests, less able entrepreneurs who do not recognise this may self-select into the ‘wrong’ (i.e. able) group, destroying the value of self-selection contracting tools. Third, if the number of high-risk types is not too numerous, the benefits of screening entrepreneurs might not compensate for the deadweight costs entailed by the seizure of collateral. Banks might then do better by offering pooling contracts without collateral rather than devising separating contracts based on collateral (Mattesini, 1990). More generally, in complex informational environments, signalling equilibria can be infeasible, and scarcely worthwhile for small ventures (Admati and Pfleiderer, 1994; Ravid and Spiegel, 1997). If that is the case, some asymmetric information remains, restoring pooling of at least some entrepreneur types with the associated possibility of credit rationing.

222

Financing

It should also be noted that whether separating or pooling equilibria emerge also depends on the assumed nature of the game between entrepreneurs and banks, especially the timing of the moves of the game, and the definition of competitive equilibrium (Hellwig, 1987). If there is perfect competition among banks, and if the outcomes of screening or monitoring are public information (i.e. can be exploited by other banks), then the free-rider problem means that each bank has weaker incentives to perform costly screening and thereby resolve asymmetric problems. Then pooling contracts can emerge again.21 When there is imperfect competition among banks, banks can also maximise surplus by means of pooling, rather than separating, contracts (Besanko and Thakor, 1987a). On the face of it, the available evidence appears to lend greater support to the existence of pooling rather than separating equilibria. Actual lending rates tend to be rather insensitive to differences in both ex ante and ex post (e.g. actual performance) measures of ability, with bank small-business lending margins over the base rate typically varying by no more than a few percentage points.22 There is also limited direct evidence of entrepreneurial selection, whether adverse or favourable. Experimental evidence indicates that participants in risky investment games exhibit over-optimism which overwhelms any selection effects. In one such case, as the experimenter increased the opportunity cost of participating in the game (D), there was no systematic tendency for the most or least able to drop out of the game first (Coelho, 2004). On average, those who dropped out were the least optimistic, but no less or more able, than those who continued playing. This result highlights the earlier point about how over-optimism can muddy signals of ability. On the other hand, evidence from developing countries points to the existence of bank-optimal interest rates for some individual loans, where increases in nominal interest rates above 60 per cent are associated with greater default problems and lower portfolio quality (Cull et al., 2007). Even if separation does occur in credit markets, it is theoretically possible for credit rationing itself to serve as a BAD that separates types.23 To see how, suppose entrepreneurs have no collateral, and that BAD is the probability that banks randomly ration borrowers. Then gs are willing to take the risk of being rationed in return for a lower interest rate (since they are more likely to succeed and hence repay their loans) – whereas bs care less about a lower interest rate and more about receiving a loan. In effect, the risk of being credit rationed is the price that gs must pay to signal their types. However, an unrealistic feature of this model is that good risks are rationed, while bad risks are fully funded. An arguably more plausible alternative model, proposed by Stiglitz and Weiss (1983), has banks threatening to ration credit in later periods unless borrowers succeed early on. This induces good behaviour by borrowers and low default rates. But to be credible, banks must be seen to carry out their threat, so in equilibrium some credit must be rationed. In summary, the complication of financial contracts by opponents of credit rationing can be countered by the introduction of new dimensions of unobservable borrower heterogeneity by supporters of credit rationing. It would be desirable to take data to

Debt finance for entrepreneurial ventures

223

the various models to narrow the field of theoretical possibilities by discarding some explanations in favour of others. However, little progress on this front has been made to date. One reason for this is no doubt the tricky practical problem that hidden borrower and venture types are unobservable. To address this problem, the researcher would need to obtain data on hidden characteristics that are somehow unobserved by banks. What the above discussion does reveal is that credit rationing models are sensitive to changes in their assumptions. Some researchers apparently believe that this casts doubt on the relevance of the phenomenon itself (e.g. Hillier and Ibrahimo, 1993). On the other hand, as Clemenz (1986) has argued, it is very unlikely that necessary conditions for credit rationing can ever be found in models that are sufficiently general to remain interesting. Furthermore, the number of possible mechanisms by which credit rationing can arise could instead be regarded as a strength rather than a weakness, because it expands the set of circumstances under which such rationing may occur. Further theoretical refinements of existing models are unlikely to change any of these points. Instead, a more pressing need is for empirical research to address directly the question of whether credit rationing exists, and if so, to what extent. This issue is taken up in chapter 9. 7.4

Over-investment

Implicit in the models discussed so far is a presumption that financing problems are generally associated with too little entrepreneurship. However, other plausible models can be constructed in which the opposite outcome – too much entrepreneurship – arises. De Meza and Webb (1987) (hereafter DW) were the first to propose a model with this property, discussed in the first subsection below. The one after discusses over-optimism as an alternative source of over-investment. 7.4.1

The de Meza and Webb (1987) over-investment model DW utilise several similar assumptions to SW, including the pooling of heterogeneous types with a single credit contract under conditions of asymmetric information. The main difference between the DW and SW models is the assumed structure of project returns. Using the terminology of chapter 2, DW assume that entrepreneurs differ in terms of the probability their venture succeeds (first-order stochastic dominance) rather than in terms of their risk profile (second-order stochastic dominance). Starting with assets B, on which they could earn a gross rate of return ρ outside entrepreneurship, an entrepreneur with ability x whose investment project succeeds with probability p(x) compares the safe return of ρB with expected returns in entrepreneurship, namely p(x)(R − D). Banks would prefer to lend to higher-x borrowers because the probability they repay D from their gross return R is higher. An entrepreneur who fails gets zero returns, and the bank loses its capital in the venture. At the outset, DW point out that all entrepreneurs must invest all of their personal wealth B in their project before applying for loans, in order to avoid transmitting an adverse signal about their type. This is because the most able entrepreneurs are the

224

Financing

least likely to fail, and so can supply finance to themselves on better terms than they can obtain in the market. Less able entrepreneurs must emulate them in this respect to avoid transmitting an adverse signal about their true abilities which would separate them into less favourable contracts. Full self-finance, a result which has been known since at least Leland and Pyle (1977), does however reduce the welfare of risk-averse borrowers who have to take larger stakes in their own firm than they would wish to without asymmetric information. Some evidence is consistent with the hypothesis of full self-finance (Chandler and Hanks, 1998), though some other research suggests that entrepreneurs tie up only about one-half of their wealth in their business (Gentry and Hubbard, 2001, 2004). Over-optimistic entrepreneurs also offer more self-finance than realists, as one would expect of people who rate their chances of success very highly (Landier and Thesmar, 2008). In terms of the occupational choice model of chapter 2, the marginal entrepreneur x˜ is defined by p(˜x)(R − D) = ρB. Hence if G(x) is the distribution function of ability, a total of 1 − G(˜x) high-ability individuals choose entrepreneurship (recall chapter 2). A key implication of this selection is that the marginal entrepreneur x˜ is of low ability relative to other entrepreneurs (see, e.g., Figure 2.3(b)). As D increases, the least able entrepreneurs leave the market first, improving the quality of the borrower pool, and increasing the average probability of success among the remaining active entrepreneurs, p. Hence p is an increasing function of D. This implies that there is favourable rather than adverse selection. Because banks’ returns are Dp, there can be no credit rationing: banks always have an incentive to increase the interest rate D to remove any excess demand for funds, unlike the case in SW where an interior bank-optimal interest rate did not give them that incentive. Likewise, banks would be forced by competitive pressures to reduce D to the market-clearing rate Dm if there was an excess supply of funds. The impossibility of credit rationing is DW’s first key result. Their second key result is the inevitability of over-investment. This follows because the marginal entrepreneur x˜ generates an expected loss of [p − p(˜x)]D to the bank, i.e. generates expected returns that are less than the opportunity cost of the funds used. In other words, resources would be more efficiently deployed if the least able entrepreneurs cancelled their ventures and became safe investors instead. The reason this does not happen is because under imperfect information the least able are cross-subsidised by the more able. The former consequently find it privately worthwhile to undertake ventures that are socially inefficient. This inefficiency is especially damaging in financial markets characterised by limited liquidity (Vesala, 2007). With over-investment, a tax on interest income can restore the economy to full efficiency. This policy, which probably has minimal adverse work effort disincentive effects, increases the interest rate which must be paid to compensate depositors, and induces the least able entrepreneurs to exit.24 Alternatively, the government can tax entrepreneurs’incomes while promoting policies which favour non-entrepreneurs, such as increasing the minimum wage (Parker, 2003c). In general equilibrium, by removing the least able entrepreneurs from the pool of borrowers and producers this policy can

Debt finance for entrepreneurial ventures

225

allow banks to reduce interest rates to remaining entrepreneurs. At the same time, consumers are willing to pay more in markets where the average quality of goods is higher. Ghatak et al. (2007) refer to a tax on interest income as a ‘trickle up’ policy, since its benefits can work through to benefit even the entrepreneurs who pay the tax via higher interest repayments. In contrast, subsidising credit would attract the least able entrepreneurial types into the market, which would force banks to increase interest rates for every entrepreneur in the pool.

7.4.2 Over-optimism Two influential models of over-optimism and over-investment are de Meza and Southey (1996) and Manove and Padilla (1999). Where these models differ is that de Meza and Southey (1996) assume that banks can distinguish between realistic and over-optimistic borrowers, while Manove and Padilla (1999) assume that banks cannot distinguish between them and pool them together. In the de Meza and Southey (1996) model, over-optimists are the ones to suffer. They remain in entrepreneurship for too long, earning lower incomes and bearing greater risk than they would in paid employment, and committing excessive personal equity to their projects – until they run out of funds and are forced to exit. As a result, over-optimistic entrepreneurs who are denied loans may be better off ex post than those who obtain them. It is noteworthy that over-optimism can actually create or worsen redlining by causing incompetent individuals to apply for start-up funds. Hence subsidies and loan guarantee schemes designed to attract redlined borrowers can decrease rather than enhance social welfare (Coelho et al., 2004). Coelho et al. (2004) shy away from DW’s recommendation of actively taxing entrepreneurship, since they acknowledge the possibility of private and social benefits to entrepreneurship; but they regard proentrepreneurship policies as intervening in ‘the wrong direction’. As they put it, ‘A few good apples may attract all the publicity but this should not divert attention from most of the barrel being bad’ (2004, p. 413). In Manove and Padilla’s (1999) model, on the other hand, competitive banks face asymmetric information about which applicants are over-optimists and which are realists. Over-optimistic entrepreneurs of low ability pass up the chance to apply for smaller investments that would make them a profit, requesting unprofitable larger loan sizes because they incorrectly think they possess high ability. This form of over-investment can be rectified by increasing the interest rate, for example by DW’s interest tax policy. Manove and Padilla (1999) conclude that ‘conservative’ bank lending policies may therefore be justified even in the face of vocal criticism from the small-business lobby. On the other hand, banks do not always get it right, either. As recurring Latin American debt write-offs and the recent crisis in the Western banking system illustrate, banks can also be prone to exuberant over-optimism. Bank over-optimism can be especially damaging to the soundness and stability of the credit market. Suitable capital requirements and prudential regulation are required to counter this problem (Manove and Padilla, 1999).

226

7.5

Financing

Conclusion

Debt finance for new start-ups has an important bearing on entrepreneurship and public policies towards it. The availability and price of loans, and other contract terms such as collateral and loan sizes, are of primary importance to entrepreneurs who borrow from banks. The theoretical literature reviewed in this chapter showed that when entrepreneurs possess more information about their proposed ventures than banks do, it is possible for efficient contracting between banks and entrepreneurs to break down. Either too much or too little finance, involving too many or too few entrepreneurial ventures, can occur from the standpoint of social efficiency. The view of this author is that de Meza and Webb’s (1987) model of entrepreneurial finance is a particularly important contribution to the debate about entrepreneurial finance. That model challenges the widespread assumption that financing problems necessarily result in too few entrepreneurs. De Meza and Webb show that it is quite possible for there to be too many entrepreneurs in equilibrium, and that a suitable government policy for encouraging entrepreneurship could involve deterring the least able people from borrowing in the credit market. Therefore policy-makers should not automatically equate credit market imperfections with insufficient entrepreneurship, and so should not immediately reach for instruments designed to draw marginal individuals into entrepreneurship (see also Parker, 2007a). My discussion has touched on appropriate policy responses where they exist. These responses are diverse, reflecting the diversity of the models. One should certainly not take any particular policy recommendations too seriously. The models are partial equilibrium in nature, and do not take into account broader effects that need to be analysed in a general equilibrium setting (Hillier and Ibrahimo, 1993). The models also tend to be rather fragile, in the sense that altering some of their assumptions can easily reverse the predicted forms of market failure and the policy implications (Parker, 2002). I will elaborate on this theme of model fragility in the following paragraphs, before highlighting the desirability of future empirical research in this area. The easiest way to demonstrate the fragility of theoretical models of inefficient investment is to combine elements of the DW and SW models together in generalised models. Several generalised models allow entrepreneurs to operate ventures with heterogeneous risks as well as heterogeneous probabilities of success. Perhaps unsurprisingly, generalised models predict that Type II credit rationing, under-investment by some good types and over-investment by some bad types can all occur individually or simultaneously, while the aggregate level of entrepreneurial investment can be greater than or less than what transpires in the ‘first best’ equilibrium under symmetric information (Hillier and Ibrahimo, 1992; de Meza and Webb, 2000). This outcome is intellectually unsatisfying because it replaces concrete (albeit specialised) results with a nebulous demonstration that ‘anything can happen’. A similar problem occurs in a different DW-type model where expected returns in both occupations are increasing in unobserved entrepreneurial ability x, not just entrepreneurship as is assumed in DW and SW (Parker, 2003c). In all of these models, an appropriate policy instrument is income

Debt finance for entrepreneurial ventures

227

taxation or subsidisation of entrepreneurs. More will be said about income taxes and entrepreneurship in chapter 17. As a further example of ambiguity in generalised models, Bernanke and Gertler (1990) added a prior stage to DW’s model in which entrepreneurs must engage in costly search activity to identify a suitable venture. The information gathered from the search is known only by the entrepreneur who conducted the search, and good ventures are pooled with bad ventures at the second stage resulting in over-investment, as in DW. However, pooling diminishes the returns to finding good ventures, which diminishes entrepreneurs’ incentives to search, and so promotes under-investment. Whether the net effect turns out to be under-investment or over-investment is ambiguous. Some efforts have been made to cut through the ambiguity present in generalised models. It was noted earlier in the chapter that when it is costly to verify ex post project returns, debt is the optimal financial contract. But if state verification costs are trivial, equity can coexist with debt. It has recently been demonstrated that if entrepreneurs operate projects with heterogeneous risk profiles and expected returns, and can choose freely between DF and EF, over-investment arises under quite general conditions (Boadway and Keen, 2006). While this solution seems to cut through the Gordian knot of ‘anything can happen’, it is probably of little practical relevance. For reasons outlined in greater detail in the next chapter, most small entrepreneurial ventures are of no interest to equity financiers. Consequently, entrepreneurs can only realistically access one type of financial intermediary: banks. That returns us to the ambiguities of the generalised models based on debt finance alone. In any case, DW’s over-investment result is not robust to alternative generalised models which tie entrepreneurs’ success probabilities to the quality of workers hired by entrepreneurs. More entrepreneurs bid up the wage and attract highquality workers into their firms. But this positive externality is not internalised by the entrepreneurs themselves, implying under-investment in entrepreneurship. Depending on whether this effect or the DW effect is strongest, there can be too many or too few entrepreneurs (Boadway and Sato, 1999). The ensuing policy analysis is complicated. Furthermore, different forms of competition between different types of financial intermediary can contradict Boadway and Keen’s (2006) apparently decisive theoretical result. Gersbach and Uhlig (2007) analyse an economy in which two types of bank compete with each other to supply credit to entrepreneur borrowers. One type is a ‘standard’ bank, as analysed so far, which is good at ex post monitoring but not at ex ante screening. The other is an ‘investment’ bank or credit rating agency, which is good at ex ante screening but not at ex post monitoring. Investment banks which are good at screening cream off the best borrowers at the outset, leaving the worst borrower types to the standard banks. The adverse selection problem this creates forces some standard banks either to leave the market (so reducing the supply of credit) or to increase debt repayments if the market will bear it. But the latter induces moral hazard. In either case, interbank competition causes social inefficiency. The policy recommendation in this model is to reduce competition between the different types of bank. Unfortunately, this is the opposite of what Basel II regulations are deigned to achieve, namely improving

228

Financing

standard banks’ screening effectiveness which can be expected to intensify competition with investment banks. To conclude, a better basis for policy recommendations would exist if one could sort through the various models and identify the ones with the greatest empirical relevance (Parker, 2002). Future empirical investigation along these lines would be valuable, but likely to encounter major practical difficulties. Perhaps unsurprisingly, therefore, the literature to date has not progressed very far in this direction. Attempting to reject models on the basis of their indirect predictions is also unlikely to be informative, for the simple reason that models can often be modified to better fit the stylised facts when they come into conflict with the original model.25 Data limitations which hamper empirical research also bedevil efforts to measure the extent of Type I and Type II credit rationing. Evidence on these issues, commonly bundled together by applied researchers and referred to simply as ‘credit constraints’, is reviewed in chapter 9. 7.6 Appendices 7.6.1

Bernhardt’s (2000) model of Type I credit rationing

There is a single-period planning horizon, at the start of which an investment project becomes available. Entrepreneurs have the skills to expedite the project but lack the requisite capital, k, which they borrow from a bank. At the end of the period the project pays off pf (k), where p > 0 is a stochastic price with distribution function G(p), whose support is the positive half-line; and where f (·) is a strictly concave production function. Entrepreneurs and lenders are risk-neutral and symmetrically uninformed about realisations of p ex ante. Lenders supply k via standard debt contracts and lend at the competitive gross interest rate r per unit of capital. If an entrepreneur defaults, the lender takes over the project and extracts all the revenues. Entrepreneurs are protected by limited liability (their minimum return is zero) and maximise expected profits, given by max E {max [0, pf (k) − rk]}. k

(7.1)

When choosing k, the entrepreneur being protected by limited liability is concerned only with positive profit realisations, so has the first-order condition
[pfk (k ∗ ) − r] dG(p) = 0, (7.2) p≥p∗

where p∗ is the price at which the entrepreneur just begins to break even: i.e. p∗ f (k ∗ ) − rk ∗ ≡ 0; and where k ∗ denotes the privately optimal capital choice. Here a subscript denotes a derivative. Bernhardt showed that, when there is a positive probability of default, k ∗ is not the same as the efficient level of investment, k e . The first-order condition for k e is


e e pfk (k ) dG(p) ≡ pfk (k ) dG(p) + pfk (k e ) dG(p) = 1. (7.3) p

p≥p∗

p ρ ≡ ρ(D) and differentiate (7.7) to obtain characterised by (D, θ). Write ρ˜ := ρ(θ, ∞ ˜ dρ d θ˜ g(θ) ˜ [1 − F(D, θ )] dG(θ ) . (7.8) + θ (ρ˜ − ρ) =− ˜ dD dD 1 − G(θ˜ ) 1 − G(θ) By inspection, D has two effects on banks’ expected portfolio rate of return. The second term of (7.8) is positive, capturing the positive effect of a higher interest rate on bank expected returns. But the first term is negative (by (7.6)), capturing the adverse selection effect. That is, a higher interest repayment increases the riskiness of banks’ portfolios, leading to lower expected bank returns. If the first term outweighs the second, banks’ expected returns ρ may eventually become a decreasing function of the interest rate D, as illustrated in Part (a) of Figure 7.2. However, if the first term is small in magnitude, ρ(D) increases monotonically, and credit rationing cannot arise. 7.6.3

Bester’s (1985a) screening model Bester (1985a) assumed that all entrepreneurs possess some wealth B, but that it is privately costly for them to convert it into collateral. Suppose that collateral conversion costs are proportional to the amount of collateral posted C, by a factor k > 0. Project returns Rθ for entrepreneur of type θ are stochastic, with distribution function Fθ (R) for type θ ∈ {g, b}, where Fg (R) second-order stochastically dominates (SOSDs) Fb (R), y i.e. 0 [Fb (R) − Fg (R)] dR ≥ 0 ∀y ≥ 0. Entrepreneurs borrow loan size L at interest rate r and post collateral C ≤ B. They pay the collateral conversion cost kC whether they succeed or fail, but lose C to banks as well if they fail. For simplicity there is no limited liability. An entrepreneur θ’s expected profit given credit contract  = (D, C) is therefore

πθ () = E{max[R˜ θ − (1 + r)L − kC, −(1 + k)C]},

Debt finance for entrepreneurial ventures

231

where D = (1 + r)L. On a loan  to entrepreneur θ the bank receives the expected rate of return ρθ () = E{min[(1 + r)L , R˜ θ + C] − L}/L. Entrepreneurs’ preferences depend systematically upon their type. To see how, use πθ () above to obtain the marginal rate of substitution between r and C as σθ () = −

Fθ [(1 + r)L − C] + k . {1 − Fθ [(1 + r)L − C]}L

By SOSD, Fb (R) > Fg (R) which implies σg () > σb (). The ‘single crossing’ property of indifference curves illustrated in Figure 7.4 is satisfied: b has a steeper indifference curve in (r, C) space as in Figure 7.4. Hence separating contracts (g , b ) rule out pooling and therefore credit rationing. Finally, the iso-profit lines in Figure 7.4 can be obtained by using ρθ () above to obtain banks’ indifference curves for loans to type θ entrepreneurs: µθ () = −

Fθ [(1 + r)L − C] . {1 − Fθ [(1 + r)L − C]}L

Note that with k > 0 this is less steep than entrepreneur θ ’s indifference curve at , as illustrated. Notes 1. This process will not be studied explicitly in this book: see Tajnikar et al. (2006) for an overview. 2. See de Meza and Southey (1996), Boot et al. (1991), Bester (1994) and Coco (1999). 3. The US studies include Leeth and Scott (1989), Berger and Udell (1990, 1992, 1995) and Coleman (2000). The UK studies are by Hanley and Crook (2005) and Burke and Hanley (2006). However, none of these studies properly tests the assumption that bank screening based on observable characteristics has already taken place. Instead, the above evidence relates to a pool of ventures with heterogeneous observable characteristics, in which the observably riskiest have to post the most collateral. Therefore this does not rule out the possibility that collateral varies within observable groups in a manner consistent with a signalling mechanism. To convincingly reject the signalling story, evidence would be needed of a positive correlation between risk and collateral among only ventures which appeared to be observably identical to banks. 4. See Petersen and Rajan (1994), Berger and Udell (1995) and Harhoff and Körting (1998). 5. See Greenbaum et al. (1989), Sharpe (1990) and von Thadden (2004) for a correction. 6. See Townsend (1979), Diamond (1984) and Gale and Hellwig (1985). 7. See Admati and Pfleiderer (1994), Ravid and Spiegel (1997) and Povel and Raith (2004). 8. See Laffont and Martimort (2002, chap. 4.8.4) for an explicit discussion of this point. An ongoing research agenda studies optimal contracts in a different class of moral hazard financing problem where entrepreneurs can divert operating cash flows before reporting them to banks. See Biais et al. (2007) for a recent discussion of, and contribution to, this literature. 9. I will not discuss in this chapter macroeconomic implications of credit rationing, as in e.g. Jaffee and Stiglitz (1990, sec. 5), Hillier and Ibrahimo (1993, sec. 6), Acemoglu (2001) and Fender (2005). For example, Acemoglu (2001) links credit market constraints to high levels of unemployment in Europe. 10. See Barro (1976), Keeton (1979), Koskela (1983), Gale and Hellwig (1985) and Gray and Wu (1995).

232

Financing

11. Clemenz (1986, sec. 5.3), de Meza and Webb (1992) and Bernhardt (2000). 12. E.g. over-optimistic entrepreneurs request excessive loan sizes because they overestimate their future returns (Manove and Padilla, 1999). Or able entrepreneurs trying to separate themselves from less able types request excessively large loan sizes and are imitated by less able types (Boadway et al., 1998). 13. Despite external finance helping entrepreneurs to partially share project risks (Heaton and Lucas, 2004), risk-sharing and risk-aversion do not play a central role in the discussion that follows. Asymmetric information rather than risk aversion is the necessary condition for credit rationing, under-investment and redlining outcomes (Fender and Sinclair, 2006). It might also be thought that banks could refuse to finance entrepreneurs because they are risk-averse. But banks are unlikely to behave in a risk-averse manner, owing to their ability to spread risks over numerous borrowers. In short, risk aversion is not a convincing rationale for credit market inefficiencies. 14. If banks’ expected return function has more than one mode, then SW showed that credit rationing is still possible if the highest mode occurs for an interest rate ≤ Dm , where Dm is the market clearing rate (see below). The other possibility is of two interest rates, where credit at the lower interest rate is rationed, but where all borrowers can obtain funds at the higher interest rate. 15. See Watson (1984), Williamson (1986, 1987) and Ghatak et al. (2001, sec. 6.1). 16. And by increasing the average wage, it has also been argued that economic development can also promote the efficient transfer of family businesses – inherited by untalented scions of successful entrepreneurs – to the talented poor who can afford to buy them (Caselli and Gennaioli, 2005). Another possible source of redlining occurs when creditors are refused ‘absolute priority’ over debtor entrepreneurs’ assets. Anticipating violations of the absolute priority rule (APR), banks expect lower returns from lending to entrepreneurs. They respond by refusing credit altogether to entrepreneurs with little personal equity or large loan size requests (Longhofer, 1997). A culture of APR violations increases the equilibrium cost of finance; a clear policy implication is to enforce APR rules more stringently. 17. For further analysis of collateral in the context of entrepreneurial finance, see Chan and Kanatos (1985), Clemenz (1986), Besanko and Thakor (1987a) and Bester (1987). Coco (2000) surveys the literature. 18. The reason is that, in a competitive equilibrium, a pooling contract, p say, must make zero expected profits. But this involves gs cross-subsidising bs. Hence gs will prefer g to p , and any bank not offering g will lose gs to rivals that do. With the departure of gs, the p contract becomes loss-making, and is driven out of the market by competition. 19. In principle, a continuum of contracts specifying varying degrees of limited liability could be devised to perfectly separate a continuum of types and reveal the entire distribution of entrepreneurial abilities to banks. 20. See Bester (1985b), Besanko and Thakor (1987b), Milde and Riley (1988), Innes (1991, 1992) and Schmidt-Mohr (1997). 21. An open theoretical question is whether banks should share information about their borrowers with competitor banks. On the one hand, information-sharing can improve entrepreneurs’prospects and so mitigate moral hazard practised by ‘trapped’ entrepreneurs (Padilla and Pagano, 1997). This can be expected to enhance banks’ profitability. On the other hand, banks which share information lose informational rents. This relaxes competition among banks for borrowers at early stages of venture financing, which redistributes surplus from talented entrepreneurs to banks, and leads to possible adverse selection in the credit market (Gehrig and Stenbacka, 2007). 22. For UK evidence see Bank of England (1993), Keasy and Watson (1995), Cressy and Toivanen (1997) and Cowling (1998). US evidence reveals similar margins (Berger and Udell, 1992), with smaller firms paying somewhat higher interest rates than larger firms (Hubbard et al., 2002). 23. See Besanko and Thakor (1987a), Smith and Stutzer (1989) and Gale (1990a).

Debt finance for entrepreneurial ventures

233

24. Subsequent contributions have shown that the over-investment result and the interest tax policy are robust to the introduction of costly screening (de Meza and Webb, 1988); variable venture sizes (de Meza and Webb, 1989); risk aversion (de Meza and Webb, 1990); and the addition of ex ante moral hazard à la SW (de Meza and Webb, 1999). However, it should be noted that crosssubsidisation can sometimes increase social welfare, for example if less able types respond to more favourable pooled contract terms by supplying greater effort and so reducing the probability of default, thereby decreasing the pooled interest rate for all borrowers (Vercamman, 2002). 25. For example, de Meza and Webb’s (1987) model predicts a negative relationship between personal wealth and entrepreneurship, in conflict with most evidence on the issue (see chapter 9). But a subsequent paper published in 1999 by these two authors generalised the model by incorporating moral hazard, with the result that a positive relationship between wealth and entrepreneurship emerges. 26. Wette (1983) showed how adverse selection can also occur if banks vary collateral C while keeping D fixed.

8

Venture capital and other sources of finance

Chapter 7 analysed the debt finance of entrepreneurial ventures. While debt is the dominant form of external finance for most start-ups, many entrepreneurs also tap alternative sources of external funding. These include equity finance; informal sources of capital such as friends and relations; micro-finance and credit co-operatives; and trade credit. Sometimes, debt finance is unavailable, especially for entrepreneurs who lack collateral and propose investment projects comprising few tangible assets, high levels of risk and long development periods in which debt claims cannot be serviced. Banks typically do not finance entrepreneurial opportunities of these kinds; equity providers and other types of financiers do. Equity finance comprises both formal and informal venture capital. An extensive literature now documents the role of equity finance in funding high-growth and hightech entrepreneurial ventures. Formal venture capital has been used to finance such wellknown companies as Apple, Google, Amazon, Federal Express and eBay, among others. Equity finance, especially formal venture capital, tends to be ‘narrow but deep’, in the sense that few entrepreneurs use it but those who do can access large sums of funding from it. More widespread, but usually offering much more modest amounts of finance, are informal sources such as friends and family members, as well as micro-finance schemes. Mason (2006) cites survey responses from eighteen of the GEM countries which estimate that 3.4 per cent of adults are informal investors (either ‘business angels’ or friends and family). On average, they provide $196 million per year to new and growing companies, equivalent to 1.1 per cent of the GDP of these countries. Some 50 per cent of this informal investment goes to relatives; 29 per cent to friends and neighbours; 11 per cent to work colleagues; and 8 per cent to strangers. Informal finance is an important source of funding for entrepreneurs, amounting to between 60 and 90 per cent of total venture capital, including capital from institutional sources. The aim of this chapter is to outline the rich array of economic issues underlying these various sources of entrepreneurial finance. The first section briefly outlines the organisational structure of formal venture capital, as well as the size of the venture capital market. The primary aim here is to discuss the economics of this type of entrepreneurial finance, rather than to present an exhaustive description of the structure, 234

Venture capital and other sources of finance

235

investment process and functioning of the venture capital industry (on this, see Gompers and Lerner, 1999). The second and third sections explain respectively the advantages and limitations of formal venture capital finance for entrepreneurs. The fourth section briefly describes some features of business angels – the providers of informal equity finance. A variety of other informal sources of finance are discussed in the penultimate section; the final section concludes. 8.1 Venture capital and entrepreneurs 8.1.1

Organisational structure

Venture capital has developed as an important intermediary in financial markets, providing capital to firms that might otherwise have difficulty attracting financing. These firms are typically small and young, plagued by high levels of uncertainty and large differences between what entrepreneurs and investors know. Moreover, these firms typically possess few tangible assets and operate in markets that change very rapidly. Venture capital organisations finance these high-risk, potentially high-reward projects, purchasing equity or equity-linked stakes while the firms are still privately held. (Gompers and Lerner, 2001, p. 145)

This section deals with formal venture capital, which the authors of the above extract define as ‘independently managed, dedicated pools of capital that focus on equity or equity-linked investments in privately held, high growth companies’ (Gompers and Lerner, 2003, p. 267). Providers of venture capital are referred to below as ‘venture capitalists’ (VCs). They include private independent venture funds, corporate subsidiaries and special investment schemes. Most professional venture capital funds are organised as venture capital companies which attract funds from outside investors, who are limited partners; the VCs themselves are general partners. VCs frequently manage several entrepreneurial ventures at any one time, and are often actively involved in them to enhance their prospects of success. The discussion that follows focuses mainly on private independent venture funds. Special (public-sector) investment schemes are discussed in chapter 16; another source of equity finance is corporate venture capital, which is discussed below. Gompers and Lerner (1999, 2001) describe the origins of the venture capital industry, and characterise the ‘venture cycle’ of the industry. The venture cycle starts when VCs raise a ‘closed-end’ (i.e. fixed term and fixed size) venture fund. VCs then screen, invest in, monitor and add value to the projects they select. The cycle continues as VCs exit successful deals and return capital to their investors; it is renewed when VCs raise further funds. In return for providing infusions of capital, assistance, advice and expertise (see below), VCs demand a share of entrepreneurs’profits and an exit route that will generate a hefty capital gain on their investment. VC profit shares are negotiated on a projectby-project basis. They tend to vary between 20 and 49.9 per cent (Bovaird, 1990). Inderst and Müller (2004) point out that bargaining power between entrepreneurs and VCs, as reflected in equity shares for example, varies with technological and economic conditions. During times when technology shocks are favourable, entrepreneurs have

236

Financing

more valuable outside options at the same time as the supply of available funds expands, so bargaining power swings towards entrepreneurs. When valuable entrepreneurial opportunities are less common, however, bargaining power shifts back to the VCs. In fact, entrepreneurs and VCs negotiate not only over equity shares but also over control rights and other contract provisions that protect the VC against agency risks. These include staging of future rounds of financing; staged release of shares to the entrepreneur (‘vesting’); severance payments; and VC oversight including board representation (Sahlman, 1990). VCs often insist on the right to take a seat on the board of directors, and retain control rights, including the ability to appoint managers and remove members of the entrepreneurial team (Bruton et al., 1997). In practice, this often turns out to be an important provision. For example, Hannan et al. (1996) report that on average 10 per cent of CEOs and managers in young high-tech VC-backed firms are replaced in the first twenty months, rising steadily to a replacement rate of 80 per cent after eighty months. Interestingly, opportunistic behaviour by entrepreneurs is not a common reason why VCs dismiss entrepreneurs (Bruton et al., 2000). An important feature of VC contracts is that they often specify options to transfer control rights contingent on performance. These control rights usually change as the venture develops, and favour the VC more the worse is the venture’s performance (Kaplan and Stromberg, 2003). In the presence of entrepreneurial over-optimism, clearsighted VCs have incentives to propose contingent contracts which switch control from over-optimistic entrepreneurs to realistic investors in bad states – which by definition optimists think are unlikely ever to occur (Landier and Thesmar, 2008). Like the price of equity, the allocation of control rights also depends on the balance of bargaining power between VCs and entrepreneurs. Baker and Gompers (2003) measure VC bargaining power in terms of reputation, and furnish evidence that VC reputation is negatively associated with the probability that the original founding entrepreneur remains as CEO. Lerner et al. (2003) show that VCs hold more control rights when market conditions are bad, reflecting their superior bargaining power. According to Lerner et al. (2003), this outcome is not generally optimal because control rights should ideally be allocated to the agent with the greatest ability to influence marginal project returns (which is usually the entrepreneur). The likely consequence is that VC-backed ventures will perform worse when conditions for raising capital are bad. It might seem strange that entrepreneurs would agree to submit to stringent VC terms, such as losing control of their own company. Indeed, one might expect such terms to restrict the supply of entrepreneurs willing to sell equity. In fact Dessi (2005) shows that stringent control rights are economically optimal because they elicit efficient VC monitoring and VC project continuation decisions, while preventing collusion between VCs and entrepreneurs at the expense of other investors. The precise terms of control rights in any given deal do, however, depend on a range of factors in practice (Hellman, 1998b; Kirilenko, 2001). VC equity investments typically last for between three and seven years. VCs usually aim to sell their stake at the end of this period, either through an Initial Public Offering

Venture capital and other sources of finance

237

(IPO) or, more commonly, a trade sale. In either case, the VC’s return on capital is primarily the capital gain raised from selling (a share of) the business, rather than a regular and ongoing stream of dividends. Hence the development of a viable venture capital market depends on the existence of exit routes which offer investors high returns, including favourable CGT regimes (Black and Gilson, 1998). More will be said about this in chapter 16. There is some disagreement among experts about VCs’ portfolio rates of return and risk profiles. Kerins et al. (2004, p. 387) cite evidence from Venture Economics that over the twenty years to 2001, annual net-of-fee returns to VC funds was 17.7 per cent, compared with 15.6 per cent from the S&P 500 stock index. This estimate was based on a large sample of funds. However, VCs’ returns are highly cyclical over shorter periods (Gompers and Lerner, 1999, 2001). Denis (2004) furnishes an overview of the (surprisingly diverse) empirical evidence about VCs’ average rates of return. Prior to investing, VCs acquire information about proposed new ventures in order to reduce the degree of asymmetric information they face. They do so by first performing a rigorous screening process, followed by a detailed examination of the new ventures and entrepreneurs who make it through the initial screen. This detailed examination is commonly referred to as ‘due diligence’. Due diligence can take anything from six weeks to six months. Even after these exhaustive selection processes, VCs are still left with a substantial amount of asymmetric information and risk if they decide to provide funding. That is why venture capital is also commonly referred to as ‘risk capital’. These considerations have a major bearing on the structure of financial contracts in venture capital, as will be discussed below. VCs commonly leverage specialist industry knowledge and local networks, and often syndicate with other VCs.1 One might expect VCs to keep the best entrepreneurial projects for themselves, reject the bad ones, and to only syndicate intermediate quality proposals, where they can diversify risk while pooling expertise. About 90 per cent of deals in the US venture capital industry are syndicated, compared with only 30 per cent in the less mature German venture capital market.2 Syndication offers several advantages to VCs. It enhances the screening process by increasing the breadth of scrutiny by recognised experts, and prevents competition between investors after investment opportunities are disclosed (Casamatta and Haritchabalet, 2007). Corporate venture capital is prone to similar degrees of risk. However, it comprises a small part of the overall venture capital market and is of limited relevance to most entrepreneurs. Hellmann (1998a) observes that corporate venture capital accounts for less than five per cent of all venture capital financings (though this may be a lower bound on the true size of this sector: see Denis, 2004). Moreover, companies sometimes wish to purchase equity shares in other firms, including potential competitors, for strategic reasons rather than to satisfy a need to raise finance. For example, Mathews (2006) discusses how alliances between an entrepreneurial firm and an established firm can improve the efficiency of both. Alliances often require the entrepreneurial firm to transfer knowledge to the established partner. But that heightens the established firm’s incentives to subsequently exploit the knowledge themselves by entering the

238

Financing

entrepreneurial firm’s market. The risk of such behaviour can therefore deter knowledge transfers (Alvarez and Barney, 2001). In other cases, though, asymmetric information problems within a corporation can make product development in a start-up company more attractive than internal development in their own laboratories (Dix and Gandelman, 2007). In short, the conflicts of interest that arise between the objectives of the company providing funds, and the one receiving them (especially if they are in the same product line) might explain the limited size of the corporate venture capital sector. There need to be tangible strategic complementarities between companies to overcome this problem, otherwise conventional venture capital financing is preferable (Hellmann, 2002). Indeed, evidence shows that corporate venture capital is concentrated in the ICT sector, where partner firms more routinely synergise novel technologies (Dushnitsky and Lenox, 2006). An alternative solution is for an entrepreneurial firm to sell an equity stake to an established firm, since this dilutes the latter’s incentives to cannibalise the profits derived from the equity stake. But this solution will be less attractive if the sale of equity blunts the entrepreneur’s incentives to supply effort. 8.1.2 Size of the entrepreneurial venture capital market The USA has the longest-established formal venture capital market in the world. In 2001, over $40 billion of venture capital funds were invested there, compared with only $12 billion in Europe (Bottazzi and da Rin, 2005). The US figure was down from its peak of $106 billion in 2000, and dropped further to $30.5 billion in 2007. These numbers illustrate the volatility of venture capital markets. The European market has grown steadily since 1995, with EVCA (European Venture Capital Association) data recording that total investment increased from 24 billion euros in 2001 to 79 billion euros in 2007. Of particular interest is the growing importance of ‘early stage’venture capital investments (‘seed’and ‘early stage’finance), which are arguably the investments most closely identified with individual entrepreneurship. Since the early 1990s about one-third of US venture capital investment has been in early-stage projects. In Europe the corresponding figure is one-eighth. The UK has the largest venture capital market in the European Union, followed by France, Germany and the Netherlands, in that order. Most venture capital investments in Europe and the USA are devoted to buy-outs and late-stage financing deals. It is interesting to speculate why the USAhas historically led Europe in formal venture capital. Some researchers have suggested that Europe suffers from more rigid labour markets, which can deter workers from quitting to become entrepreneurs in VC-backed ventures since they may face greater difficulties rejoining the ranks of employees if they fail (Sahlman, 1990; Jeng and Wells, 2000). It might also be noteworthy that in Europe, banks and governments are prominent sources of venture capital funds, whereas in the USA, private funds predominate. If the latter are more effective commercialisers than the former, one would expect to see a more vigorous and successful US market. Yet another explanation relates to possible differences in cultural values between Europe

Venture capital and other sources of finance

239

and America. Landier (2001) suggests that in Europe, a ‘stigma of failure’ not only discourages individuals from initiating new starts, but also makes low-risk projects more attractive for those entrepreneurs who do emerge. As noted in chapter 7, the optimal financial contract for low-risk investment projects is debt. In the USA, by contrast, it is claimed that the stigma of failure is less pronounced, encouraging more risky start-ups, for which the optimal financial contract is equity. That could explain why the American venture capital market is larger than its European counterpart. If cultural values such as stigma of failure and attitudes towards success are reinforced by these outcomes, for example by providing negative or positive demonstration effects, these values could become entrenched – making distinct European and American equilibria durable as well as pronounced. Despite the size of the formal venture capital markets in the USA and Europe, relatively few entrepreneurs make use of them. It is an established fact that equity finance (EF) – whether formal or informal – accounts for only a small proportion of external finance for entrepreneurs in most countries. For example, Bates and Bradford (1992) reported that only 2.8 per cent of US small business start-ups obtain EF. Its receipt is positively associated with owners’ education, age, amount of self-finance and track record in business. Business ownership type also matters, with evidence that familyowned SMEs use less EF than non-family-owned SMEs (Wu et al., 2007). A similar picture applies in the UK, where EF accounted for 1.3 per cent of total start-up finance by the end of the 1990s, down from 3 per cent at the start of the decade – despite the strong growth performance of the companies that used it (Bank of England, 2001; and see the next section). The third section of this chapter outlines several reasons why entrepreneurs make only limited use of EF and VCs. 8.2 Advantages of venture capital finance for entrepreneurs

There are two major reasons why entrepreneurs choose venture capital as a source of finance. The first is based on the observation that in addition to funding, VCs provide assistance and advice which adds value to entrepreneurs’ investment projects. The second reason is that equity finance is the optimal financial contract for some entrepreneurs, which VCs specialise in offering. 8.2.1 Value-adding activities by VCs VC involvement often takes the form of advice and assistance which draws on VCs’own experience and contacts. Thus VCs often provide access to investment bankers, lawyers, accountants and consultants. Several recent US studies claim to have detected various additional beneficial effects from venture capital. VCs are thought to add value to entrepreneurs’ projects in three ways: by (1) monitoring, (2) professionalising ventures and (3) certification.

1. Monitoring. There is mounting evidence that VCs are active monitors. Lead VCs visit their portfolios an average of nineteen times a year, despite the costs entailed (Gorman and Sahlman, 1989; Sahlman, 1990). Time spent monitoring and recruiting

240

Financing

is estimated at about 60 per cent of VCs’available hours, adding up to some 110 hours per annum on average per venture (Gorman and Sahlman, 1989). VCs’representation on boards of directors significantly increases (by 1.75 VCs on average) during times when the need for monitoring is greatest (i.e. when CEOs are replaced), confirming their prominent oversight role (Lerner, 1995). In addition, VCs commonly play a key role in shaping the top management teams of the companies in which they invest, and allocate decision and control rights in a manner that facilitates post-investment monitoring activities (Kaplan and Stromberg, 2001, 2003) 2. Professionalisation. VCs provide a variety of support services to their companies. For example, VC-backed ventures in Silicon Valley are significantly more likely than non-VC-backed ventures to receive help for building their internal organisations, and professionalising along several dimensions: HR policies, recruitment of professional marketing and sales staff, governance structures and the adoption of stock option plans.3 As a result, VC-backed firms raise more funds at IPO and generate superior returns relative to non-VC-backed public offerings (Brav and Gompers, 1997; Bottazzi and da Rin, 2005). This can explain why it is that, although VCs finance only 1 or 2 per cent of all new businesses in the USA, they finance between one-third and one-half of all businesses that achieve IPOs (Gompers and Lerner, 2001). VC-backed enterprises are also likelier than non-VC-backed enterprises to: • Be novel innovators, bringing their products to market quicker (Hellman and Puri, 2000; Kortum and Lerner, 2000); • Start larger and grow faster (Jain and Kini, 1995; Colombo and Grilli, 2005) and • Create more jobs (Belke et al., 2005). Consider, for example, Kortum and Lerner’s (2000) study. Kortum and Lerner recognised that VCs help enterprises to become more innovative; but they acknowledged that there might also be self-selection whereby more innovative firms choose VCs for financing. Kortum and Lerner (2000) instrumented the explanatory variable of VC usage by an exogenous policy regime switch which increased the supply of venture capital. Their findings suggest that a dollar of venture capital is on average three to four times more potent in terms of innovative performance than a dollar of traditional corporate R&D. 3. Certification. The backing of a VC can signal quality and reputation, which can attract additional investors (Amit et al., 1998). Entrepreneurs frequently accept financing offers with lower valuations in order to affiliate with more prominent VCs, consistent with the certification hypothesis (Hsu, 2004). In an important study, Lerner (1999) examined 1,435 new firms over a ten-year period and found that while US Small Business Innovation Research programme awardees located in areas with substantial VC activity grew faster, larger subsidies alone did not lead to better performance – findings which are also consistent with the certification hypothesis.4 Value-adding VC benefits can be substantial, and even more important than the finance itself. Indeed, Gompers and Lerner (2003, p. 277) claim that ‘it is the

Venture capital and other sources of finance

241

non-monetary aspects of venture capital that are critical to its success’. These benefits might help explain why entrepreneurs are willing to sacrifice ownership, decision and control rights in order to access it. However, entrepreneurs might not receive these benefits in all cases. For example, Bottazzi and da Rin (2002) found that European VCs had little discernible impact on the growth rates, corporate strategies or job creation performances of their portfolio firms. That might point to a quality issue in European venture capital relative to US venture capital. Also, Bottazzi and da Rin (2005) provide some counter-evidence to Belke et al. (2005), finding that venture capital is not associated with growth in post-IPO sales or employment. A further problem, which constitutes a negative manifestation of certification, is ‘grandstanding’, whereby IPOs are launched at earlier dates than entrepreneurs would find ideal, in order to establish the reputations of young VC companies for future fund-raising. If the outcome is under-pricing of IPOs, VCs might therefore not add value on net (Gompers, 1996). This possibility has been tested by comparing the outcomes of formal VCs who might engage in grandstanding with business angels who, lacking an investor fund to satisfy, have the opposite incentives (Chahine et al., 2007). On the other hand, as noted above in the discussion of Kortum and Lerner’s (2000) important study, endogeneity and especially selection biases often bedevil academic studies that claim to measure the impact of VCs on venture performance. These problems are exacerbated by the fact that researchers usually only observe performance for self-selected ventures that receive further stages of funding or which are disposed of profitably (Cressy, 2006c). Overall, the consensus is that the benefits from venture capital can be substantial. VCs’ hands-on role is reflected in their geographical concentration. Being located near entrepreneurs and other VCs facilitates close working relationships and gives VCs better access to possible syndication partners, information about industry trends, and promising developments in university science. It also keeps entrepreneurs competitive and better informed. Somewhat less studied to date has been the role played by the quality of the entrepreneur in value-adding VC activities. In an analysis of VC-backed entrepreneurs over 1986–2000, Gompers et al. (2006) found that although VC-backed serial entrepreneurs comprised only 7–14 per cent of their sample, those serial entrepreneurs who had taken their venture public on a previous occasion enjoyed nearly twice as high a probability of successfully going public in their next venture compared with first-time VC-backed entrepreneurs. But while more experienced VCs significantly increase the probability of success of first-time or failed serial entrepreneurs, they seem to have no such impact on previously successful serial entrepreneurs. Thus it appears that more experienced VCs are better at identifying and helping ‘diamonds in the rough’, but are no better than less experienced VCs at adding value to entrepreneurs who are already proven performers. Although successful VC-backed serial entrepreneurs are rewarded with access to earlier stage financing, and more favourable control provisions than their first-time counterparts, they apparently cannot command higher prices for the equity they sell

242

Financing

(Kaplan and Stromberg, 2003; Gompers et al., 2006). At the same time, venture capital funds obtain higher returns from investing in serial entrepreneurs than in novices. These findings are consistent with an imperfectly competitive venture capital market, in contrast to many theoretical models of venture capital markets which generally assume competitive conditions. 8.2.2

Equity finance as an optimal financial contract In contrast to debt finance (DF) contracts, which stipulate in advance a given repayment due to a bank that is invariant to gross returns from a successful venture, an equity finance (EF) contract entitles a lender to a stake, or share, of a firm’s profits. The evidence shows that projects receiving EF tend to have particular characteristics that set them apart from those which do not, including a lack of tangible assets; the prospect of years without positive cash flows; and high levels of market and strategic uncertainty (Winton and Yerramilli, 2008). Banks, which mostly issue DF contracts, are typically not interested in these kinds of investments, and require collateral which many high-risk and high-tech projects do not possess (Gompers and Lerner, 1999; Freel, 2007). Indeed, projects with the characteristics listed above tend to be highly concentrated in particular industry sectors, which might explain why venture capital is also concentrated by sector. For example, VentureOne reports that more than three-quarters of venture financings over the 1990s occurred in the IT and healthcare sectors. These sectors account for a large proportion of innovative activity in the US economy. In addition to these points, the financial economics literature has identified several reasons why equity might dominate debt as an optimal financial contract. One of these reasons was explained in chapter 7, namely that when investment projects are ranked by second-order stochastic dominance, EF dominates DF under conditions of asymmetric information (de Meza and Webb, 1987). This argument was made under the assumption of competitive financial markets. But EF can dominate DF in imperfectly competitive markets as well, if entrepreneurs who sell a risky equity claim to a financial investor can thereby incentivise the investor to deny funding to other competitors seeking to enter and compete in the same market (Cestone and White, 2003). Equity can also dominate debt owing to a foreclosure option inherent in DF. Under conditions of asymmetric information, entrepreneurs are fearful of foreclosure by banks, which encourages them to behave more opportunistically and risk failure even though both parties to the DF contract would ideally prefer to continue the project (Trester, 1998). An EF contract can circumvent this problem. An unusual feature of formal venture capital finance, compared with other sources of formal and informal finance, is its use of contracts which mix elements of debt and equity together. This is especially true in the USA, where VCs make greater use of ‘convertible’ contracts than pure equity – in contrast with other developed countries such as Canada, where VCs appear to issue mainly pure equity contracts (Cumming, 2005). As its name suggests, a convertible contract allows VCs to convert the form of their compensation between equity and debt claims, depending on the performance of the venture. For example, convertible debt is a loan with an option for the VC to convert

Venture capital and other sources of finance

243

the owed sum into equity shares. To see its value, consider a project which starts out modestly but then turns out to enjoy spectacular success. With a convertible contract, a VC has the option to switch from a fixed debt repayment to an equity claim, thereby sharing in the upside return. Using real options logic, uncertainty about venture returns can explain why VCs might prefer convertible contracts to plain equity. Uncertainty also surrounds the eventual exit route for the investment project, namely an IPO or trade sale. Here again, convertibles provide the VC with flexibility to maximise returns depending on the exit route ultimately chosen (Berglöf, 1994; Hellmann, 2006). Other advantages of convertible loans is that they can reduce moral hazard, specifically the inclination of the entrepreneur to choose excessively risky projects (Green, 1984; Dewatripont et al., 2003); and they can induce efficient investment by the entrepreneur in ‘double moral hazard’ settings (Schmidt, 2003). Double moral hazard in this context refers to the situation where both entrepreneurs and VCs have incentives to under-invest in privately costly effort which enhances venture performance. Chapter 16 will provide a formal analysis of this issue, together with a discussion of some appropriate public policy responses. Other financial economists have sought to understand the use of convertible contracts in terms of VCs’ preference for releasing funding to entrepreneurs in stages, rather than all at once. So it is helpful first to understand why VCs utilise staged finance. One reason is that it enables them to gather information and monitor the progress of ventures, giving them the option to abandon unpromising projects relatively cheaply (Bergemann and Hege, 1998). As VCs acquire more information about entrepreneurs and their projects, they can make superior contingent investment decisions compared with a strategy of investing all of the required funds upfront (Li, 2008). The simple threat that a VC can walk away before the funding schedule is completed can also provide entrepreneurs with necessary incentives to conserve capital, hit targets and eschew morally hazardous behaviour – including attempting to renegotiate contracts to create possible hold-up problems for the VC (Neher, 1999; Smirnov and Wait, 2007). If these arguments are valid, one would expect the frequency of re-evaluations and number of financial stages to increase when the VC expects more conflicts with the entrepreneur. The opposite should occur as the company becomes more established or as entrepreneurs’ reputations grow. Evidence from a sample of 794 US companies receiving 2,143 rounds of investment over 1961–92 supports these hypotheses. Gompers (1995) found that as successive stages were reached, the frequency of waiting between rounds for subsequent injections of finance decreased, and the average amounts of funding given to entrepreneurs increased, as did entrepreneurs’ rates of spending (the co-called ‘burn rate’) of funds. As noted above, staged finance can explain the existence of convertible contracts. Consider entrepreneurs who respond to staged finance by engaging in ‘window dressing’ activities. These are activities which exaggerate the short-term performance of the project, in an effort to convince the VC to release the next stage of funds. VCs’ optimal response to window dressing in the presence of staged finance is a convertible

244

Financing

debt contract (Cornelli and Yosha, 2003). To see why, observe that window dressing can make low-quality projects harder to identify, so reducing the probability of the VC liquidating them. But it also renders high-quality projects easier for the VC to identify, rendering the conversion option valuable to the VC. Entrepreneurs operating promising projects would not wish to trigger the option, which serves as a suitable disincentive to engage in window dressing.5 Finally, one can ask why banks do not offer EF contracts. There appear to be several reasons why they issue DF rather than EF contracts. For a start, until 1999 the Glass–Steagall Act prohibited US banks from owning the equity of non-financial firms (Gompers and Lerner, 1999). But even in cases when banks are allowed to take equity stakes, the general view is that bank loan officers lack the specialised skills and experience to emulate VCs in terms of their value-adding capabilities.6 Banks have a comparative advantage in monitoring borrowers through long-term relationships; but change is very rapid in high-tech industries in particular and requires different skills (Gompers and Lerner, 1999). Nor can banks simply replicate the services of VCs by hiring external consultants (Casamatta, 2003). On the other hand, VCs’ expertise puts them in a better position to steal entrepreneurs’ ideas. Hence the choice between EF and DF may involve a trade-off (Ueda, 2004), and entrepreneurs are more likely to prefer EF the better their intellectual property rights are protected. Taking all of the above arguments together, it is clear that VC-backed finance can offer compelling advantages over alternative sources of finance for particular kinds of entrepreneurial venture. However, there are also good reasons why venture capital is offered to only a small minority of entrepreneurs. These reasons are discussed next. 8.3

Drawbacks of venture capital and equity finance for entrepreneurs

This section consists of two parts. The first explains why only a small minority of entrepreneurs choose to use venture capital. The second briefly considers the possibility of ‘equity rationing’ and under-investment, and suggests that EF might be scarce because its use is (inefficiently) restricted by financiers. 8.3.1

Factors inhibiting the use of equity finance

It is natural to ask why EF accounts for such a small proportion of external finance for most entrepreneurs. The following reasons can be adduced: entrepreneurial unwillingness to cede ownership and control rights; financing costs; agency costs; and information costs. • Entrepreneurial unwillingness to cede ownership and control rights. Most entrepreneurs start off with ideas whose growth potential is not even proven to themselves, let alone to outside VCs. This can make entrepreneurs reluctant to share ownership and control of a business whose future value could be substantial. Entrepreneurs might also be initially unaware of the damage they can inflict on VCs by engaging in moral hazard. But if VCs are aware of this problem, as seems

Venture capital and other sources of finance

245

likely, they will demand higher equity shares to price in the attendant risk of moral hazard. The danger is that negotiations over equity stakes can fail, with entrepreneurs attributing demands for high VC equity shares to naked greed (hence the epithet ‘vulture capital’). That makes them unwilling to sell the large stakes required by VCs (Hagen and Sannarnes, 2007). Even entrepreneurs who do anticipate growth and moral hazard issues may be reluctant to sell equity and lose complete control over their ventures. According to the ‘pecking-order’ hypothesis, entrepreneurs seek funds in an order that minimises external interference and ownership dilution (Myers and Majluf, 1984). Hence their first preference is for internal finance followed by DF, with EF as a last resort. Evidence supporting the pecking order hypothesis in the context of entrepreneurship appears in Cressy and Olofsson (1997), who reported a widespread preference among Swedish entrepreneurs to sell their business altogether rather than to cede a share of it to an outsider. This is especially pertinent as VCs and business angels often require rates of return of at least 15 per cent on projects and exit routes involving flotation or sale to another company (Kerins et al., 2004). A 15 per cent rate of return exceeds what most ventures can yield, and these exit routes can conflict with entrepreneurs’aspirations of indefinite ownership of a business which they have gone to considerable time and trouble to build themselves. • Financing costs. There are fixed costs of issuing shares and listing on secondary markets where shares can be traded. Most enterprises never grow to a size where these costs are warranted, even for ‘junior’ stock markets such as the US NASDAQ or the UK Alternative Investment Market (AIM). Also, larger deals such as management buy-outs or buy-ins generate greater and more reliable fee income for VCs than start-ups at the seed or early stages. Hence simple cost reasons restrict the viability and availability of EF for many entrepreneurs, especially those starting small-scale ventures. Furthermore, for large as well as small ventures, debt is usually tax-favoured and has prior claim to junior obligations and equity on a corporation’s assets in the event of a liquidation event. • Agency costs. The involvement of outside financiers can cause conflicts of interest with entrepreneurs which reduce the latter’s flexibility, and impose costs on all of the contracting parties. This is probably true to some extent of all financial instruments, but it appears to be especially pronounced for EF. Entrepreneurs have incentives to take perquisites which reduce the outside value of the venture since, unlike DF, EF allows entrepreneurs to share the costs with VCs. In response, VCs monitor the entrepreneur, resulting in both a perceived ‘loss of control’ by the entrepreneurs, and monitoring costs that must be recouped in the form of greater VC equity stakes. If these stakes become sufficiently large, entrepreneurs can be discouraged from seeking EF altogether. DF rather than EF is a natural solution to entrepreneurial moral hazard, since by making the entrepreneur the residual claimant DF furnishes the required high-powered incentives (see also Barzel, 1987). • Information costs. Because entrepreneurs and financiers must co-operate closely once they enter a relationship, each side typically expends a costly search effort.

246

Financing

High costs of gathering information can increase the price of funds beyond the willingness or ability of entrepreneurs to pay. In addition, full disclosure conditions may compromise confidential information and encourage competitors to appear who bid away the superior returns of the venture to be financed (Campbell, 1979). It might not be possible in these cases for entrepreneurs to erect information barriers to guarantee returns on their innovations (Himmelberg and Petersen, 1994). Furthermore, as was seen in chapter 7, if the least able entrepreneurs anticipate limited returns in good states, they will prefer EF to DF since they have a lower probability of sharing upside returns with VCs while avoiding a certain debt repayment in bad states. In contrast, abler entrepreneurs prefer DF, because they anticipate capturing greater upside returns. Attempting to sell equity can therefore convey a negative signal about an entrepreneur’s ability, so all entrepreneurs are dissuaded from requesting an EF contract.7 For all of the reasons discussed above, entrepreneurs often prefer to utilise DF, for which a cheap, well-established and relatively efficient market is available. Some of the problems with EF outlined above also apply to informal equity providers, but not all, as we will see in section 8.4 below. 8.3.2

Equity rationing, funding gaps and under-investment It is sometimes claimed that there is a ‘funding’ or ‘equity’ gap for EF. The term should be distinguished from ‘equity rationing’. ‘Equity gap’ is commonly used to refer to a mismatch between the fund sizes that interest entrepreneurs on the one hand and VCs or business angels on the other. For example, high fixed costs of screening and monitoring make VCs unwilling to supply the relatively small sums typically required by entrepreneurs. The term ‘equity rationing’ in contrast refers to the problem, analogous to credit rationing, where there is a persistent excess demand for funds which even competitive VCs that face low financing costs are unwilling to satisfy. Part of the motivation for exploring these issues derives from the simple observation that few entrepreneurs use EF in practice. Also, some studies have claimed to find indirect evidence of equity-based rationing by linking constrained access to equity capital with lower R & D expenditures and other forms of company investment (Hao and Jaffe, 1993; Himmelberg and Petersen, 1994). However, these studies were conducted using data on firms from the 1970s and the early 1980s, when the pool of venture capital was smaller than it is now. Even if these firms were genuinely equityrationed, it is not obvious that they would be today, when the supply of funds is so much greater. Hellmann and Stiglitz (2000) proposed a model of equity rationing in which debt and equity providers compete with each other to finance heterogeneous entrepreneurs who possess private information about both their project risks and their returns. Hellmann and Stiglitz’s (2000) model unifies both the Stiglitz-Weiss (SW) and de Meza and Webb (DW) models discussed in chapter 7. Hellmann and Stiglitz (2000) assumed that lenders specialise in either DF or EF, and that entrepreneurs cannot use a mixture of both. They

Venture capital and other sources of finance

247

then showed that credit and equity rationing may occur individually or simultaneously. The usual culprit of lender return functions which decrease in their own price (chapter 7) accounts for the possibility of rationing in each individual market. As in SW, lenders do not increase the price of funds to clear the market, because good types may exit the market, causing lenders’ expected profits to fall. Hellmann and Stiglitz also obtained the surprising result that competition between the two markets may itself generate the adverse selection that leads to rationing outcomes. The reason is that if many low-risk entrepreneurs switch between debt and equity markets, competition induces lenders in one or both markets to reduce the price of funds below market-clearing levels in order to attract them. Rationing is then needed to break even. However, if only EF was offered in these cases, credit rationing would disappear.8 Greenwald et al. (1984) developed a different model in which the least able entrepreneurs prefer EF and the ablest prefer DF (see above). These authors showed that the adverse signal transmitted by choosing EF can increase the cost of capital sufficiently to deter credit-rationed borrowers from availing themselves of EF altogether. This reinforces the potential importance of the SW credit rationing result, since it rebuts the argument that entrepreneurs who are rationed in the debt market can obtain funds elsewhere, e.g. in the form of EF. Another source of equity rationing is linked to the observation that most VC funds are of the ‘closed end’variety, i.e. have a fixed and predetermined scale. By having ‘shallow pockets’, VCs increase competition for funds among entrepreneurs at the refinancing stage, which improves entrepreneurs’incentives to select projects which not only have a positive NPV at the refinancing stage, but also have returns that are higher than those of competing portfolio projects (Inderst et al., 2007). However, while restricting funds is optimal for VCs, an efficiency loss ensues for society at large because shallow pockets mean that some positive NPV projects are not financed at all. Parallel to the analysis of chapter 7, under-investment is also possible when EF contracts are used. In a classic paper, Myers and Majluf (1984) analysed the problem of an established firm which issues new equity when a valuable investment opportunity appears. Managers of existing enterprises have privileged information about both the company’s assets in place and the value of the new investment opportunity. Suppose that EF is used to finance the new investment; that managers act in the interests of their existing shareholders; and that shareholders do not actively rebalance their portfolios in response to what they learn from the firm’s actions. Then Myers and Majluf (1984) showed that a new share issue can reduce the share price by so much that managers might optimally pass up the new profitable opportunity, implying under-investment. In contrast, the use of internal funds or risk-free DF removes under-investment without reducing the share price – so is preferred to EF by managers. However, these results are sensitive to assumptions about the objectives of managers of the enterprise and the behaviour of shareholders (Noe, 1988). In any case, the applicability of the Myers– Majluf result to most ‘typical’ entrepreneurs is probably limited. More recently, Bergemann and Hege (2006) analysed the financing of an entrepreneurial project under uncertainty about both the time of completion and the

248

Financing

probability of success. VCs update their beliefs about the probability of success as time goes on. Bergemann and Hege (2006) predict that funding generally stops too early relative to the efficient stopping time. This is another form of under-investment. A different strand of research recognises that the success of complex investment projects often depends on the effort of multiple external agents who provide requisite complementary resources.9 This can create problems of its own. For example, suppose that a series of investors needs to perform a costly evaluation of proposed projects. Then each investor has an incentive to free-ride on the evaluations of the others, leading to the danger of inefficiently protracted delays in assembling resources. Hellmann (2007a) showed that the entrepreneur’s optimal strategy in this problem is to ‘knock on every door’ if resource-providers like to be solicited to go first. If they do not, the entrepreneur should ‘pester’ one resource-provider continually. Overall, the literature on equity rationing appears to be more of theoretical than practical interest. Even in the 1980s, VCs regularly claimed that they had more available funds than attractive projects in which to invest, which is suggestive of an equity gap rather than equity rationing (Dixon, 1991). With the expansion in the size of the venture capital market in the last twenty years, the likelihood of equity rationing has presumably decreased even further. And industry experts regularly argue that limited opportunities rather than limited capital explain the low number of VC-backed start-ups (e.g. Bhide, 2000, p. 162). On the other hand, there is undeniably a pronounced cyclical element to the supply of VC funds, which could be asynchronous with the demand for EF by entrepreneurs. We still await compelling empirical evidence identifying systematic under-investment and equity rationing by VCs, as opposed to occasional mistakes in the evaluation of new venture funding opportunities by these financiers. 8.4

Informal equity finance: business angels

Mason (2006, p. 261) defines business angels as ‘high net worth individuals who invest their own money, along with their time and expertise, directly in unquoted companies in which they have no family connection, in the hope of financial gain’. Angel finance is a long-established source of seed and start-up capital for entrepreneurs. It differs from lending to friends and family members in several respects. In particular, the primary aim of business angels is to make money rather than to leverage personal relationships. The ‘typical’ angel is male, 45–65 years old, well educated but without a PhD. Many are successful cashed-out entrepreneurs. Measuring the size of the informal equity sector is difficult because there is no formal register of informal investors, and for understandable reasons many angels prefer to remain anonymous. This explains why estimates that have been compiled for the USA and the UK exhibit considerable sampling variation. Most estimates suggest that the US informal angel sector is at least as large as the formal equity sector, even though most individual deals are on a smaller scale. Amatucci and Sohl (2007) estimated there were about 225,000 US angels in 2004, investing about $22.5 billion in 48,000 ventures, with an average deal size of $470,000. In contrast, formal VCs invested $21.3 billion

Venture capital and other sources of finance

249

in 3,000 ventures, with an average deal size of $7.3m. This estimated number of angels is remarkably similar to early estimates compiled by Wetzel (1987) over twenty years ago, although Sohl (2003) obtained somewhat higher estimates of between 300,000 and 350,000 angels in the USA. Sohl (2003) also estimated there are some 20,000– 40,000 angels in the UK, investing between £500,000 and £1 million in 3,000–6,000 companies. According to Mason and Harrison (2000a), the UK’s informal market for start-up and early-stage venture financing is also broadly similar in total value to that of the formal market. The fact that angels’average deal sizes are so much smaller than those of VCs suggests that they serve a different part of the entrepreneurial finance market. Business angels tend to concentrate on early-stage financing in younger companies. Their role usually declines at later stages when more substantial funding is needed. Thus 45 per cent of US angel deals in 2004 focused on seed and early-stage projects, compared with just 6 per cent among VCs (Amatucci and Sohl, 2007). The complementarity between angels and VCs might explain why they frequently share information and collaborate (Mason and Harrison, 2000a, 2000b). Angels can help to close the equity gap between friends and family and other sources of informal finance on one hand, and formal VC on the other. Sohl (2003) estimates that the minimum deal sizes of the latter are about £1m in the UK, and about $5m in the USA. According to Amatucci and Sohl (2007), angels are beginning to redistribute funds away from early project stages towards post-seed stages. If this constitutes a genuine trend, it may have important implications for the availability of finance for the smallest entrepreneurial ventures. On the one hand, it might lead to less access to finance for the smallest projects, while on the other it might help to close the equity gap higher up the value-chain. One factor helping to narrow the difference between angels and VCs is the increasing frequency of syndication in angel finance (Mason, 2006). Mason speculates that this will increase efficiency for several reasons. Angels will be more visible and attractive to clients, enabling them to reap economies of scale in deal flow. Also, larger deal size may fill a vacant equity gap for more valuable projects, since angel syndicates can provide more follow-on finance than single individuals. Syndication also promises greater expertise and hence more value-added activities and professionalisation, leading to greater credibility with VCs, who as noted above are often needed for later-stage ‘follow on’ financing. Business angels share some similarities with VCs. Both types of investor reject most applications for funding, usually because the quality of investment proposals fails to meet a minimum threshold. Furthermore, Amatucci and Sohl (2007) point out that like VCs, angels are often valued by entrepreneurs largely on the basis of their value-adding activities. Partly this reflects the early-stage nature of so many of these deals, where expert guidance and advice is essential to the successful launch of new entrepreneurial ventures. Nevertheless, there are also some important differences between business angels and VCs.According to Wong (2002), angels locate geographically close to the entrepreneurs they finance. More than one-half reportedly operate within half a day’s travel of their

250

Financing

home (Sohl, 1999; Wong, 2002). Angels also use few of the control mechanisms favoured by formal VCs. For example, unlike VCs, angels are rarely involved in shaping top management teams. Angels also tend to take straight equity (and sometimes issue debt as well as equity); rarely demand board seats; and rarely use stage-financing or demand anti-dilution protection. American and British angels appear to have broadly similar characteristics, although UK investors tend to be less wealthy, investing about one-half of the sums of their US counterparts. UK business angels are also more likely to invest independently rather than in consortia, although similarly to the US eight times as many UK businesses raise finance from business angels as from institutional VC funds (Mason and Harrison, 2000a). Further details about the investment practices of business angels can be found in the recent overviews by Mason (2006) and Amatucci and Sohl (2007).

8.5 8.5.1

Other informal sources of finance Family finance

Families are the most commonly used source of business loans in the USA after banks and other financial institutions. Consider the following evidence drawn from Bates’ (1997) analysis of the US CBO database. Family finance is used by over one-quarter of non-minority-owned businesses, compared with two-thirds who access bank credit. Among some minority groups, however, family finance is used more extensively than bank finance. For example, family finance was used by 41.2 per cent of immigrant Korean and Chinese business owners, compared with 37.4 per cent who used bank finance (Bates, 1997; see also Yoon, 1991). For all groups, family loans are of a smaller average size than bank loans, although family loans remain an important source of funds by value, being worth an average of $35,446 for non-minority owners compared with an average $56,784 for bank loans. There is also a gender difference, with womenowned business owners relying more on family finance than men-owned businesses (Bates, 1997; Haynes and Haynes, 1999). UK evidence tells a similar story, with family loans comprising the largest source of funds after bank loans.10 Family finance accounts for 15–20 per cent of start-up finance by value amongst ethnic-owned businesses in the UK. According to Smallbone et al. (2003), ethnic-minority businesses are significantly more likely to draw on family finance than white-owned businesses are. Take-up rates are 45 per cent compared with 25 per cent, with the highest rates of usage being found among South Asians and the lowest rates of usage by the Afro-Caribbean and Chinese communities. For ethnic British entrepreneurs overall, family finance is a more important source of funding than bank credit. Data from other countries tell a broadly similar story. Knight (1985) reported that the following sources of funds were used by high-tech Canadian firms at the pre-start-up stage: personal savings: 60 per cent; family and friends: 13 per cent; bank loans: 12 per cent; and trade credit: 6 per cent. According to the 2003 GEM report, 31 per cent

Venture capital and other sources of finance

251

of the funds for start-ups less than three years old across the GEM countries came from family and friends. The importance of families and friends for supplying startup finance appears to be even more pronounced in developing countries, according to evidence from India (Bell, 1990; Kochar, 1997) and Côte d’Ivoire (Goedhuys and Sleuwaegen, 2000). What motivates lending within families? Family members may have private information about borrowers that is unavailable to banks (Casson, 2003); and they may be able to monitor and exert peer pressure on the borrower. For their part, family lenders may be trusted to behave sensitively if the entrepreneur encounters difficult business conditions not of their own making. Family lenders can also serve as loan guarantors to outside lenders (Jones et al., 1994). Furthermore, if the borrower stands to inherit the family lender’s estate, then a family loan effectively becomes a mortgage on his own inheritance. In contrast, banks are usually unwilling to accept the prospect of inheritance as security for a loan (Casson, 2003). Basu and Parker (2001) point out that most family loans tend to be interest-free (Light, 1972). Basu and Parker (2001) argue that family members might supply interest-free loans not only if they are altruistic towards the entrepreneur, but also if they are selfish. The selfish motive for interest-free lending arises if the loan entitles the lender to a sufficiently valuable option to ‘call in the favour’, and turn entrepreneur themselves at a later date. Using a sample of relatively affluent Asian immigrant entrepreneurs, Basu and Parker (2001) claimed to find indirect evidence of both altruistic and selfish family lending motives. Also, they estimated that greater use of family finance is positively associated with an entrepreneur’s age, the number of hours they work in their business and whether they employ their spouse in the venture. Basu and Parker (2001) estimated that family finance is a gross substitute for bank loans. However, a subsequent GEM study has failed to find any evidence that people who had acted as informal investors are any more likely to become entrepreneurs themselves later on (de Clercq and Arenius, 2006). This is consistent with an altruistic but not a selfish motive for within-family lending. Other evidence suggests that family finance is not associated with successful enterprise, being correlated with low profitability and high business failure rates.11 The reasons for this are unclear, although one can speculate that family funds are often limited, and that only marginal businesses might resort to this mode of financing. Evidence from Kenya suggests that business owners who use family finance tend to be younger and run younger firms, perhaps because they lack social capital required to obtain bank credit or loans from informal lenders such as micro-finance schemes (Akoten et al., 2006). There appears to be little appetite for government intervention in this area. Apart from the case of the Dutch government, which has offered tax exemptions for family finance for business start-ups, there are few public policy initiatives in operation. 8.5.2

Micro-finance schemes It is generally agreed that lack of access to credit is one of the principal reasons why citizens of developing countries remain poor (Hermes and Lensink, 2007). Formal credit markets are weak or non-existent in these countries, and few borrowers possess

252

Financing

meaningful collateral. As a result, valuable entrepreneurial projects go unfunded, thereby hindering economic development. Micro-finance schemes offer a possible solution to this problem. These schemes are invariably non-profit-making organisations which lend small sums to people who are unable to obtain funds from ‘conventional’ banks, usually because they lack collateral. Many such schemes currently operate around the world, mainly in developing countries with under-developed financial sectors. They include the Grameen Bank in Bangladesh, BancoSol in Bolivia and Bank Rayat in Indonesia.12 Perhaps the most famous is the Grameen scheme, founded in 1976 by Muhammad Yunus, an economics professor and winner of the 2006 Nobel Peace Prize. By the start of the present century this scheme, which makes unsecured loans for non-agricultural self-employment activities, served more than 3.2 million borrowers, 95 per cent of whom are women, through over 1,100 branches in 41,000 villages. Its assets exceed $3 billion. These schemes have now been widely copied in many developing countries around the world. Hermes and Lensink (2007, p. F1) cite evidence that the number of microfinance institutions increased from 618 in 1997 to 3,133 in 2005. The growth in the number of customers receiving credit from these institutions likewise increased over this period from 13.5 million to 113.3 million (84 per cent of whom were women). Despite some idiosyncratic variations, micro-finance schemes tend to have several features in common, including direct monitoring of borrowers; stipulation of regular repayment schedules that start immediately after loan disbursement; and the use of nonrefinancing threats to generate high repayment rates from borrowers who would not otherwise receive credit (Armendáriz de Aghion and Morduch, 2000). Hoff and Stiglitz (1990) identify several other direct and indirect methods used by micro-finance schemes to protect themselves against loan defaults. Direct methods include staged investments, with small initial loans followed by larger loans conditional upon satisfactory repayment of earlier loans; and a requirement that borrowers contribute a small percentage of their loaned funds to a savings scheme to insure against ‘covariant’ risk, whereby large-scale shocks such as bad weather hit an entire region and result in mass default. Indirect methods include delegated monitoring, among others (see below). Economists who study micro-finance have tended to focus their attention on group lending schemes with joint liability, in which individuals form into groups and are jointly liable for penalties if one member of the group defaults. The penalty might be the denial of future credit to all group members if one member defaults (as in the Grameen scheme), or group liability for loans if a single member defaults (as in the Bangladesh Rural Advancement Committee scheme). The advantage of joint liability contracts is that they give entrepreneurs incentives to exploit local information and exert pressure to discipline co-members in a manner consistent with the interests of lenders (and, by releasing funds from lenders, therefore also the entrepreneurs). The particular mechanisms involved include: 1. Peer monitoring and mitigation of moral hazard. Group members may be able to monitor each other in a manner unavailable to banks. For example, members may

Venture capital and other sources of finance

253

know or live near each other, and share information via ‘cross-reporting’, whereby borrowers submit reports on each other at village meetings (as in the Grameen scheme, where cross-reports are collected at the same time as loan repayments are made). In principle, this elicits truthful revelation by borrowers; eliminates collusion by borrowers against lenders; and avoids the need to invoke hefty punishments (Rai and Sjöström, 2003). Furthermore, because under joint liability each member’s payoff depends on whether other members’ ventures succeed, all members have an incentive to monitor other members’ behaviour, and to take remedial action against members who misuse their funds. Joint liability can therefore overcome moral hazard and monitoring problems.13 2. Cheap state verification and repayment enforcement. Group members may be in a better position than banks to learn about partners’ venture outcomes. Then joint liability can encourage them to exert peer pressure to deter partners from defaulting opportunistically in good states. Social ties make members’ behaviour easier to observe by co-members, and enable enforcement of powerful sanctions against malefactors (Besley and Coate, 1995). For example, group members might threaten others with ostracism or some other social sanction if they shirk in a manner that invites default, or if they invest in excessively risky ventures. Also, if group members have lower auditing costs than banks (as seems likely) a group lending scheme may economise on state verification costs. Only if the whole group defaults will a lender incur audit costs, so this arrangement reduces average auditing costs and enhances efficiency. Indeed, if audit costs are too high for banks to be able to offer any individual loan contract, a group lending scheme could facilitate lending where none was possible before. 3. Mitigation of adverse selection. Rather than changing borrowers’ behaviour, as above, joint liability can favourably alter the pool of borrowers. All group members are treated as being in default if any one member of the group does not repay their loan. Hidden entrepreneurial types optimally match together in pairs in a process known as ‘positive assortative matching’. This kind of matching is optimal because although all types prefer to match with the most able types, the joint benefits to the latter are greatest if they match with other able types, since then the probability of a joint default is lowest. Like collateral, joint liability can reveal types and lead to more efficient contracting even when borrowers lack ‘formal’ financial collateral.14 Many micro-finance schemes boast high repayment rates. These have reached 98 per cent (it is claimed) in the case of Grameen, compared with only 50 per cent for conventional bank loans in comparable regions. Repayment rates in Chinese schemes have attained similar levels (Armendáriz de Aghion and Morduch, 2000). While the mechanisms listed above probably do not furnish an exhaustive set of explanations for the high repayment rates of micro-finance schemes,15 some evidence confirms their usefulness, as I now explain. One part of the evidence base identifies the role of peer monitoring and peer pressure. Using data on 137 groups from a group-based lending programme in Guatemala,

254

Financing

Wydick (1999) found that peer monitoring (associated with shorter distances between group members and also with knowledge that members have of other members’ sales) and a group’s willingness to apply pressure on delinquent members were the salient factors explaining superior borrowing group performance. Paxton et al. (2000) also reported that social pressure within groups is positively related to repayment performance, while Karlan (2007) claimed that through successful monitoring, borrowers know who to punish and who not to punish after a co-member defaults. While 2.8 per cent of Peruvian group lending scheme relationships were observed to deteriorate when there was no default, this figure rose to 12.0 per cent in cases where a member did default (Karlan, 2007). Other evidence supports the notion of selective group formation. In particular, Wenner (1995) reported that repayment rates among twenty-five Costa Rican groups were highest among those which had formal rules about stating how members should behave (see also Zeller, 1998, for further evidence from Madagascan borrower groups on this point). Rules can be interpreted as a proxy for screening of members via local reputations. Sharma and Zeller (1997) claimed that groups that were formed using a self-selection (screening) process have better repayment performance records. The severity of punishments contingent upon default are significantly associated with superior repayment rates. According to Wenner (1995), more geographically isolated groups have higher repayment rates, possibly because lack of alternative sources of finance make successful group repayments imperative. Similarly, Sharma and Zeller (1997) reported superior repayment performance among groups otherwise lacking access to credit. There is also evidence that more severe default penalties and covariant risk increase repayment performance among Thai borrowing groups (Ahlin and Townsend, 2007). An important aspect of the evidence base described so far is the role played by social ties. On the whole, social ties are related to superior repayment performance, as predicted by theory (Zeller, 1998; Wydick, 1999). However, some of these studies measure social ties in terms of engagement with other people in the community. These measures are prone to selection biases if groups are formed in neighbourhoods which have both superior entrepreneurial opportunities and stronger social networks. In the spirit of a ‘natural experiment’, this problem can be avoided, and causality more reliably inferred, by studying borrowing groups with randomly assigned memberships which cut across neighbourhoods. Karlan (2007) followed this strategy using data from FINCAPeru, and found that two measures of social ties (geographical proximity and cultural similarity) were significantly associated with higher repayment rates and savings of members, owing to better monitoring and enforcement of repayments among members with stronger social ties. On the other hand, a study based on data from 140 borrowing groups in Burkina Faso by Paxton et al. (2000) found that group solidarity was more pronounced than peer pressure when groups faced repayment problems. These authors detected a general unwillingness to apply sanctions to colleagues facing repayment problems. This finding raises a serious potential problem: strong social ties might induce borrowers to

Venture capital and other sources of finance

255

refrain from imposing harsh penalties, effectively leading to collusion against the microfinance lender. Hence social ties do not necessarily improve repayment performance. In line with this possibility, Sharma and Zeller (1997) reported that groups containing lots of relatives were associated with more repayment problems in Bangladesh. And Ahlin and Townsend (2007) reported that co-operation among members, as measured by sharing of resources and joint decision-making, was associated with lower repayment performance among Thai borrowing groups. In fact, the evidence reviewed above is subject to several important limitations (Hermes and Lensink, 2007). Weak linkages of empirical constructs to theory; the use of crude and questionable empirical proxies; and possible endogeneity problems are rife in this literature. In response, there have been efforts to tease out and test some of the key implications of the theories (Ahlin and Townsend, 2007); to utilise data akin to natural experiments (Karlan, 2007); and to apply laboratory experimental methods to shed light on the relationship between social capital and the repayment performance of borrower groups (Cassar et al., 2007). Moving on from the empirical evidence, I now turn to the welfare implications of micro-finance for entrepreneurs. Micro-finance schemes promise several social welfare benefits. First, for the reasons outlined above, they can enhance efficiency and loan repayment rates compared with individual loan contracts (Ghatak, 2000; Lensink et al., 2005). The benefit of higher repayment rates can be recycled to borrowers in the form of lower interest rates and/or larger loan sizes. This can decrease further the severity of asymmetric information problems such as adverse selection (see chapter 7), while directly increasing borrower welfare. Second, ventures can be funded that would otherwise not take place. This can be especially important in poor regions, where selfsufficient entrepreneurship promotes economic development and alleviates poverty – the so-called ‘micro-finance promise’ (Morduch, 1999a). While loans can only ever be part of a solution to poverty, they can certainly assist the process (Morduch, 1999a, p. 1610). Indeed, the evidence suggests that access to micro-finance does contribute to poverty reduction (Khandker, 2005; Azevedo, 2006). Third, micro-finance schemes can carry in their train valuable social development programmes such as vocational training, civic information and information-sharing to members. These have been found to add substantial value to participants’ venture profitability rates and employment generation efforts.16 However, micro-finance schemes can also suffer from drawbacks. First, they can encourage excessive time-consuming and welfare-reducing monitoring by group members (Armendáriz de Aghion, 1999). And the joint liability clause can encourage excessively cautious investment behaviour. Second, there is no guarantee that a scheme will break even; and most micro-finance schemes are dependent on subsidies (Hulme and Mosley, 1996; Morduch, 1999b). Moreover, as Ghatak (2000) warned, joint liability contracts might drive individual loan contracts out of the market, undermining the viability of conventional lending institutions. Third, the transfer of risk from banks to borrowers presumably reduces borrower welfare. Fourth, micro-credit might increase the vulnerability of households to poor financial outcomes and thereby strain family

256

Financing

relations (see Azevedo, 2006, p. 327). Fifth, group lending schemes can have negative as well as positive effects on repayment rates, since it is possible for borrowers to collude and for whole groups to default in a ‘domino effect’, even though some members would have repaid under individual lending (Besley and Coate, 1995). In some theoretical models, it can be shown that the benefits of micro-finance schemes outweigh the costs (e.g. Stiglitz, 1990). But this is not a general property and it cannot be assumed to hold universally, notwithstanding some recent evidence of substantial pecuniary benefits from Bangladeshi micro-finance schemes.17 For example, McKernan (2002) reported that participation in micro-finance schemes increased monthly self-employment profits by 175 per cent on average. Pitt and Khandker (1998) noted substantial gender differences in Bangladesh, where micro-finance credit has a significantly larger positive effect on households in which women rather than men were the scheme participants. Pitt and Khandker (1998) suggest that this might be indicative of how access to credit unleashes women’s productive skills which, unlike men’s, are held in check by cultural and religious restrictions proscribing formal waged work. Anderson and Baland (2002) advance an alternative and more prosaic view: women have stronger preferences for using money in ways that help family members, whereas men prefer to divert any extra resources towards personal consumption goods such as alcohol and gambling. The discussion so far has centred on the use of micro-finance schemes in developing countries. It is natural to speculate whether these schemes’ success could be replicated in developed countries. Schreiner and Morduch (2001) and Schreiner and Woller (2003) are sceptical, on the grounds that developed countries usually have welldeveloped product and financial markets, reducing the demand for these programmes. Another drawback is that greater anonymity in developed countries makes group lending more difficult there. The force of these objections is reflected in the small number of US schemes; the tiny employment creation resulting from them; and their negligible impact on poverty. For example, Schreiner and Woller (2003) cite research estimating that access to a national programme would at most move one person per thousand from public assistance to micro-enterprise. They concluded wryly that ‘broad access to financial services in the United States is good for micro-enterprises but bad for microenterprise programs’ (Schreiner and Woller, 2003, p. 2006). These authors went on to propose alternative policy recommendations in this area (see also Bates and Servon, 1998). In any case, Armendánriz de Aghion and Morduch (2000) argue that micro-finance lenders in developing countries are moving away from joint lending in favour of individual lending for wealthier entrepreneurs. Thus Grameen of Bangladesh and BancoSol of Bolivia have now adopted individual-based lending as well as group-based lending practices. This increases profitability and hence promotes financial sustainability of micro-finance schemes. However, evidence from 124 micro-finance lenders in 49 countries suggests that the fractions of poor and female borrowers are lower in individual-based than in group-based lending schemes (Cull et al., 2007). Hence greater profitability seems to come at the expense of more limited outreach. This appears to be an enduring trade-off in micro-finance lending.

Venture capital and other sources of finance

257

Finally, Armendánriz de Aghion and Morduch (2000) claim that it is possible to replicate many beneficial aspects of group lending schemes in ‘conventional’ individual-lending contracts, including direct monitoring, regular repayment schedules and the use of non-refinancing threats. They identified loan sizes of $2,000 as the point at which joint contracts are effectively replaced by individual-lending contracts in Eastern Europe. 8.5.3

Other non-profit-making lending organisations and schemes Non-profit-making (NP) lenders are another alternative source of finance for cashstrapped entrepreneurs. Examples of NP lenders include mutual savings banks, credit unions and co-operative banks. The World Bank and the European Bank for Reconstruction and Development are examples of NP lenders on the international stage. Because NP lenders do not seek to maximise profits, one might wonder (akin to the case of NP entrepreneurs discussed in chapter 2) how an NP lender can survive in competitive markets against for-profit (FP) rival lenders. This is an especially acute question because FP banks should earn zero supernormal profits in the long run under competitive market conditions. One reason might be that FP banks collude monopolistically, though this does not explain why new non-co-operative FP banks do not then enter the market and compete away their rents. Another reason might be that NP lenders are more efficient in some markets than FP lenders are, because they know customers better or specialise in lending to marginal entrepreneurial groups that are redlined by FP lenders. One important rationale for NP lenders is that they can fill a credit gap left by FP lenders. However, as Canning et al., (2003) show, NP lenders might themselves ration credit. Hence this form of financial intermediation cannot be regarded as a panacea for limited market access to entrepreneurial finance. To see the basic intuition for the Canning et al. (2003) result, consider an NP lender which seeks to maximise consumer (i.e. borrower) surplus rather than profits. An NP lender therefore returns any surplus it makes to borrowers in the form of lower interest rates. But to avoid lower interest rates distorting loan volumes requested by borrowers (who will demand a greater volume of loans as a result of the lower interest rate), the NP lender has to ration loan volumes to keep them at their optimal levels. That is, credit rationing allows an NP lender to redistribute the surplus without distorting the volume of activity from the efficient level. This is yet another example of how credit rationing can be used as a tool of efficient contracting (see chapter 7). 8.5.4

Co-operative schemes I now briefly describe some co-operative (‘self-help’) schemes that can also leverage entrepreneurial finance. The following schemes are considered: credit co-operatives and rotating savings and credit associations (Roscas) and mutual guarantee schemes (MGSs). Credit co-operatives and Roscas Credit co-operatives are voluntary groupings of individuals that obtain funds from members and then allocate credit among them.

258

Financing

They resemble banks by taking deposits from their members while also extending loans. The whole co-operative is liable for the debts of a single member. While the spread of liability dilutes the incentive to perform peer monitoring relative to group lending schemes (GLSs), it does not eliminate it altogether (Banerjee et al., 1994).18 Other features of co-operatives are similar to those of group lending schemes, including the threat of (possibly non-pecuniary) sanctions to discourage opportunistic behaviour by their members. However, given their larger group sizes, co-operatives are more vulnerable to covariant risk. The problem of size also makes banks wary of dealing with co-operatives whose members may collude against them. Rotating savings and credit associations (Roscas) play a similar role to credit cooperatives. Rosca members save on a regular basis and periodically allocate a pot of funds to particular members, either by lot or by bidding. These funds can be used to purchase an indivisible good. This process continues with past winners excluded until everyone has won the pot once (see Besley et al., 1993). While Roscas exist primarily to facilitate purchases of lumpy consumption goods by their members, they can also facilitate the accumulation of capital required for business entry. For example, Dekle and Hamada (2000) observe that Roscas have long provided funds to SMEs in local sectors of the Japanese economy. Although Roscas are no substitute for a wellfunctioning formal credit market, they are probably better than nothing (Besley et al., 1993, 1994). GLSs, credit co-operatives and Roscas tend to be associated with low-income communities. While some of these schemes, especially credit co-operatives, continue to survive alongside formal lenders in developed economies (see, e.g., Scholten, 2000), these schemes tend to decline in importance as economic development occurs, or as the communities involved improve their access to formal credit markets (Besley, 1995). Mutual Guarantee Schemes Mutual Guarantee Schemes (MGSs) are privatesector versions of government-backed loan guarantee schemes (LGSs) (see chapter 16 for details). Like a credit co-operative, an MGS is a voluntary grouping of individuals. But whereas credit co-operatives issue loans directly, an MGS merely guarantees a bank that it will cover a pre-specified fraction of a bank loan made to one of its members in case they default. MGS members pay a fee or save in a fund that provides the scheme’s capital. In practice, MGSs take a variety of organisational structures, although they share several features in common. They tend to be industry-based and located in distinct geographical areas, potentially enabling the exertion of peer pressure to encourage loan repayments. MGSs are widespread in Europe, especially in Italy, where over 800 exist, with over one million member enterprises (Rossi, 1998). The values of guarantees vary from scheme to scheme, but according to Rossi (1998) the typical guarantee is for between 50 and 100 per cent of members’ loans. MGSs are also common in Germany where they are chartered as limited liability companies, with capital provided by the banking system, guilds and chambers of trade. Federal and state governments share the responsibility for guaranteeing up to 70 per cent of loans. The German MGSs claim to

Venture capital and other sources of finance

259

back loans with lower default rates than conventional bank loans (OECD, 1998). MGSs are less common in the UK and the USA. It is currently unclear whether their existence derives from underlying informational advantages or simply historical legacy effects. The value to banks of a loan guarantee is recycled in the form of lower interest rates charged to MGS members. As with group lending schemes, MGSs probably have an optimal size: small ones are best able to screen loan applications and exert peer pressure, but idiosyncratic risks are spread more widely in larger schemes.19 In practice, MGSs often enhance their effectiveness by pre-screening loan applications, as well as by providing financial advice and encouraging valuable information sharing among members. A self-selection rationale for MGSs can also be proposed based on the analysis of chapter 7. The fee paid by a member into an MGS can be thought of as a BAD, which ‘safe’ borrower types are more willing to pay than ‘risky’ types in return for a lower interest rate. Thus with a range of fees and savings requirements, MGSs can in principle separate types and facilitate efficient contracting. 8.5.5 Trade credit Another potentially valuable source of local information is trade credit (TC). TC comprises loans between firms that are used to purchase materials and goods in process. The duration of TC is the time elapsed between invoicing and payment. This is typically 30–60 days in both developed and developing countries (Fafchamps, 2000). TC is an important source of informal finance to entrepreneurs. According to Acs, Carlsson and Karlsson (1999, table 1.5), the value of TC in the USA in 1995 was $233 billion, compared with $98 billion for bank loans. TC can be an even more valuable source of finance in developing and transition economies, where contract enforcement is weak and formal credit markets are limited. In these circumstances, TC might be capable of mitigating credit rationing (see also Bopaiah, 1998).20 What is the rationale for trade credit? The demand for TC is obvious: delaying payment to suppliers is an important way that small enterprises deal with cash-flow variations, which can aid their survival. Even large firms in developed countries find TC useful. If banks are imperfectly competitive, TC can be cheaper than DF (Emery, 1987). Even when bank finance is cheaper than TC, it can be less attractive because banks hold liquidation rights, making entrepreneurs who fear early liquidation willing to pay a higher price for trade credit (Huyghebaert et al., 2006). In developing countries, TC may emerge as a source of finance of necessity, if formal credit markets are weak or absent. Reasons for the supply of TC are less obvious, although McMillan and Woodruff (1999) summarise two main sets of explanations:

• Explanations based on industrial organisation. If suppliers have market power, TC might enable them to price-discriminate covertly, hiding price cuts from other customers. At the same time, TC can serve as a warranty for product quality, since the delay in payment gives customers time to inspect suppliers’ merchandise. In less favourable economic circumstances, suppliers sometimes have little choice but to offer TC to cash-strapped customers if they want to make a sale.

260

Financing

• Explanations based on suppliers’ superior information and enforcement. A supplier already engaged in a long-term trading relationship with an entrepreneurial venture often has better information about the latter’s creditworthiness than a bank or finance company does. The day-to-day information flow inherent in exchange means that suppliers are well placed to evaluate their customers’ abilities to pay (Emery, 1984). Furthermore, they can solve incentive problems by threatening to withdraw access to future supplies, a sanction which may be more powerful than banks threatening to withhold future credit (Biais and Gollier, 1997; Petersen and Rajan, 1997). Another consideration is that suppliers are well placed to cheaply repossess and resell their goods in the event that an entrepreneur defaults (Petersen and Rajan, 1997). For all of these reasons, one can expect trade credit to be a useful substitute for formal finance. Some evidence from Vietnam supports the second group of explanations. A supplier’s trust in a customer firm’s willingness to repay is likely to be greater the harder it is for the latter to find an alternative supplier (McMillan and Woodruff, 1999). The threat of being cut off from future trade credit acts as a powerful incentive to repay; the repeated nature of the game and acquisition of reputation bypasses missing legal contract enforcement institutions. McMillan and Woodruff (1999) found that long trading relationships and more intense information-gathering efforts, possibly using social networks, are associated with greater disbursements of TC. Also, Petersen and Rajan (1997) reported that TC in the USA was negatively related to the fraction of customers’ inventories that are finished goods rather than raw supplies, which is consistent with their ‘repossession in default’ rationale. If suppliers have little choice but to offer TC to cash-strapped customers if they want to make a sale, it is possible that firms receive more TC when they receive less DF. On the other hand, the use of TC could convey a favourable signal of creditworthiness to banks, allowing entrepreneurs to use TC to leverage credit that might not otherwise have been forthcoming (Biais and Gollier, 1997). Reflecting these conflicting forces, the evidence on the substitutability or complementarity of TC vis-à-vis DF is mixed. On the one hand, econometric estimates by Rodriguez-Rodriguez (2006) suggest that bank credit and trade credit were substitutes for Canary Islands firms in the 1990s. On the other hand, Cook (1999) claims that TC and DF are complements in the Russian context, while Petersen and Rajan (1997) report that US firms enjoying longer relationships with financial institutions (and which are presumably less credit-constrained) do not receive more TC. Using Vietnamese data, McMillan and Woodruff (1999) detect no relationship between ventures offering TC and receiving a bank loan, although recipients of bank loans receive more TC than average. It seems likely too that this mixed evidence reflects the different economic and social contexts of the very different countries in which these studies were conducted. McMillan and Woodruff (1999) mention a possible drawback of TC and relational contracting more generally. As they put it, ‘exclusion is the corollary of ongoing relationships. Continuing to deal with the customary trading partner might mean refusing

Venture capital and other sources of finance

261

to deal with new entrants, which could result in some inefficiencies’ (McMillan and Woodruff, 1999, p. 1315). Nevertheless, there is evidence that industries with greater dependence on TC as a source of finance exhibit higher growth rates in countries with relatively weak financial institutions (Fisman and Love, 2003). One possible reason, consistent with the arguments outlined above, is that TC assists entry in industries dependent on external finance (Klapper et al., 2006). 8.6

Conclusion

Debt finance is far from being the only source of entrepreneurial finance. Several other sources of finance were reviewed in this chapter, including venture capital, informal angel finance, family finance, trade credit and various co-operative and not-for-profit lending schemes. This chapter’s treatment of alternative sources of finance has not been exhaustive or complete. For example, I did not discuss explicitly the role of credit cards, leasing arrangements or franchising – despite the possibility that these mechanisms can help eliminate funding gaps caused by limited bank credit (Horvitz, 1984). Leasing can be more economical and less risky for small firms than debt finance, conferring tax advantages and being cheaper than buying capital that will not be utilised intensively (Bowlin, 1984). Likewise, franchisors have been able to finance expansion by requiring franchisees to furnish some or all of the necessary capital (Dant, 1995). Nevertheless, what emerges from the chapter’s discussion is the wide variety of different financing arrangements that are available to budding entrepreneurs. Commentators who express concern about credit rationing sometimes appear to overlook this point. Even in countries where financial markets are poorly developed, and where aspiring entrepreneurs lack even nugatory amounts of collateral, micro-finance schemes have demonstrated the scope to expand financing activities, facilitating the creation of thousands of new ventures and possibly contributing to the alleviation of poverty. Yet the existence of a rich array of financing instruments still does not necessarily mean that entrepreneurs obtain all of the funds they desire in practice. Evidence is needed to evaluate the extent to which entrepreneurs actually face binding borrowing constraints. This subject is explored in the next chapter.

Notes 1. See Cressy (2006cc, chap. 14.5.2), Zacharakis and Eckermann (2007) and Tykvová (2007). 2. See Tykvová (2007). This author goes on to discuss optimal financial contracts for syndicates of VCs. 3. See Sahlman (1990), Hochberg (2002) and Hellman and Puri (2002). 4. Shane (2003, pp. 215–17) describes other kinds of certification, available to both VC- and nonVC-backed firms. 5. Tykvová (2007, sec. 4) discusses other reasons for the use of convertible contracts based on their ability to resolve conflicts between entrepreneurs and VCs at the time of exit from the investment. 6. See e.g. Keuschnigg and Nielsen (2003, 2004a, 2005).

262

Financing

7. See Ross (1977), Greenwald et al. (1984), de Meza and Webb (1987), Amit et al. (1990) and Innes (1993). 8. Bracoud and Hillier (2000) analyse optimal contracts in a generalisation of Hellman and Stiglitz’s (2000) model, in which expected returns vary among entrepreneurs. 9. Demougin and Fabel (2007) analyse incentive contracts in problems of this sort. 10. See Curran and Blackburn (1993), Metcalf et al. (1996), Basu (1998) and Basu and Parker (2001). 11. Yoon (1991), Bates (1997), Basu (1998). 12. Huppi and Feder (1990), Morduch (1999a) and Azevedo (2006) review the structure, rationale, costs and effects of various micro-finance schemes around the world. Aspects of these schemes have been replicated in poorer rural and inner city areas of developed countries, e.g., MicroBusiness International in the USA, the Calmedow Foundation in Canada, and the ADIE Credit Project for Self-employment in France (Rahman, 1993). 13. Stiglitz (1990), Varian (1990), Banerjee et al. (1994), Armendáriz de Aghion (1999) and Chowdury (2005). 14. See Ghatak (1999, 2000), Ghatak and Guinnane (1999), Armendáriz de Aghion and Gollier (2000), Laffont and N’Guessan (2000) and Gangopadhyay et al. (2005). 15. For example, Barboza and Barreto (2006) propose another, based on peer mentoring, whereby successful group members can actively coach and teach others essential aspects of business management. It is also noteworthy that theoretical models of group lending can relax assumptions of peer information and peer monitoring and still sustain the prediction that micro-credit outperforms individual lending (Daripa, 2008). 16. See Kevane and Wydick (2001), McKernan (2002) and Smith (2002). 17. Ahlin and Jiang (2008) study the long-run effects of micro-credit on economic development. 18. Guinnane (1994) and Ghatak and Guinnane (1999, sec. 3.1) describe the origins of credit cooperatives in Germany in the nineteenth century. Key features of the German system included screening of members (not all were admitted) and project proposals (not all were financed). Banerjee et al. (1994) study the optimal design of a credit co-operative. 19. According to Hughes (1992), there is also a public good character to MGSs, because the founding firms pay the greatest cost in setting up the loan guarantee, which later members can benefit from at lower cost. However, this does not necessarily provide a case for public support, because there is nothing to prevent incumbents devising ways of inducing future members to share the initial costs. 20. Evidence for the role of TC in developing economies can be found in Cook (1999) for Russia; McMillan and Woodruff (1999) for Vietnam, and Fafchamps (2000) for Kenya and Zimbabwe.

9

Wealth and entrepreneurship

This chapter explores the role of entrepreneurs in the accumulation of wealth and aggregate savings. As the first section of this chapter explains, entrepreneurial wealth creation influences aggregate capital accumulation, savings and wealth inequality. Furthermore, personal wealth influences entrepreneurs’decisions about the types of venture to initiate – and whether any are initiated in the first place. Although economic research in the last decade has greatly advanced our understanding of entrepreneurial wealth accumulation, several issues remain unresolved. Among these issues is a major puzzle about entrepreneurs’ asset portfolio decisions, which Moskowitz and Vissing-Jørgenson (2002) call the ‘private equity premium puzzle’. The essence of this puzzle is that entrepreneurs’ private returns are not high enough to fully compensate them for the risk that they bear. In a strictly financial sense, it seems that most entrepreneurs would do better by liquidating their enterprises and investing the proceeds in a balanced portfolio of publicly traded bonds and equities. Similarly, it might be thought that entrepreneurs should reduce ownership stakes in their own ventures and diversify their wealth using the public equity markets. Evidence showing that this strategy is not widely adopted in practice is discussed below, as well as possible reasons for limited entrepreneurial asset diversification. The first three sections of this chapter treat these issues. These sections are essentially descriptive in nature. Issues with greater conceptual content are explored in the remainder of the chapter, starting with the fourth section, which asks whether individuals need personal wealth in order to enter entrepreneurship. If they do, then entrepreneurship might not be a conduit of social and economic mobility for the less well off. This question has been explored intensively in the last two decades. After setting out the theoretical analysis of this topic, the evidence is evaluated. I go on to trace the implications of personal wealth for subsequent entrepreneurial performance. It transpires that the entrepreneurship–wealth relationship is a contentious one. Disagreement continues to rage over the existence of an empirical relationship between entrepreneurship and wealth and what such a relationship actually means. In particular, Evans and Jovanovic (1989), who initiated the debate on this topic, argued that a positive wealth– entrepreneurship relationship can be taken as prima facie evidence of credit constraints which could hinder entrepreneurial entry by worthy but impecunious individuals. But 263

264

Financing

others have argued that this relationship says little or nothing about borrowing constraints. It has also been argued that even if borrowing constraints exist, entrepreneurs can bypass them by adopting suitable avoidance strategies. The penultimate section of the chapter complements the treatment of ‘Type I’ credit rationing (rationing of loan sizes) underlying the analysis in sections 4–7 of this chapter with a discussion of ‘Type II’ credit rationing (rationing of the number of loans). This material therefore provides an empirical complement to the theoretical analyses of credit rationing given in chapter 7. The conclusion of that chapter was that theory alone cannot determine whether credit rationing exists and how widespread it might be in practice. Hence empirical evidence is needed to reach a judgement about the likely practical relevance of this phenomenon. The available evidence is reviewed here. Throughout the chapter, different sides of the various debates are presented. Brief conclusions are drawn in the closing section. 9.1 The role of entrepreneurs in aggregate wealth accumulation and inequality

Recent research has identified entrepreneurship as a key mechanism driving aggregate wealth accumulation, savings and wealth inequality in modern economies. This research effort has mainly been conducted by economists primarily interested in explaining macroeconomic savings and wealth dynamics in the US economy. The facts linking entrepreneurship with aggregate wealth accumulation are striking. By way of illustration, consider the following findings, drawn from the SCF (Survey of Consumer Finances) database:1 • Entrepreneurs accounted for only 8.7 per cent of US households in 1989, yet held 39 per cent of the nation’s net worth. • About one-half of the richest 5 per cent of American families are business owners, with the income and wealth shares of these families being greater at higher percentiles of the income and wealth distribution. • 81 per cent of individuals in the top 1 per cent of the wealth distribution in 1989 declared that they were either business owners or self-employed, with more business owners than self-employees at the top of the wealth distribution, and more selfemployees than business owners at the bottom. • Although entrepreneurial families hold less wealth in stocks than other similarly wealthy households, they still account for about one-third of all stockholdings. Hence entrepreneurship may have a significant impact on asset prices, as well as on real economic activity. Data from ‘Rich Lists’ tell a similar story. For example, between 60 and 80 per cent of people on the Forbes magazine list of the wealthiest 400 Americans are ‘self-made’ business men and women, while almost all of the rest inherited fortunes which originated in one or more businesses started by one of their parents or grandparents. Only

Wealth and entrepreneurship

265

three were celebrities. Collectively, this evidence shows that entrepreneurs dominate the ranks of the rich and the super-rich.2 How do entrepreneurs get rich? Consistent with the Forbes list, one reason appears to be dynastic bequests of entrepreneurial wealth. This not only preserves concentrated wealth holdings but also helps perpetuate aggregate wealth inequality across the generations (Cagetti and de Nardi, 2006). A second important factor is that successful entrepreneurs earn and save more than their employee counterparts (Quadrini, 2000; Cagetti and de Nardi, 2006). As well as making entrepreneurs more upwardly mobile in the wealth distribution, entrepreneurial saving drives wealth accumulation at the very top of the wealth distribution.3 Most intriguing is the possibility that entrepreneurial saving drives wealth inequality in society. Simulations suggest that entrepreneurs play an essential role in aggregate wealth accumulation, and their presence makes a major difference to the implied effects on wealth inequality of income tax reform (Meh, 2005). This seems to dominate an offsetting effect whereby entrepreneurs of modest means who introduce paradigm-shifting ‘Schumpeterian’ innovations can reduce inequality by redistributing wealth to themselves from previously successful incumbents (Spencer et al., 2008). In contrast, transfers obtained by entrepreneurs who sell controlling stakes in publicly traded firms to outsiders have little effect on aggregate entrepreneurial wealth accumulation (Klasa, 2007). And while philanthropy by entrepreneurs fulfils a socially useful function (Acs and Phillips, 2002), there is as yet little robust evidence showing it to stimulate entrepreneurship directly, in the way that dynastic family transfers do. Why do entrepreneurs save more than employees? Using a calibrated general equilibrium occupational choice model, Quadrini (2000) attributed higher savings rates by entrepreneurs to four principal factors. One is that wealth is needed to overcome borrowing constraints, enter entrepreneurship and run larger business ventures. Reinvestment of profits is a cheap and ready source of finance that can fuel venture growth. Second, financial intermediation costs drive a wedge between borrowing and lending rates, increasing incentives for entrepreneurs to save. Third, entrepreneurs face greater risk, including that arising from entrepreneurs’ limited asset diversification (Covas, 2006). Hence savings can act as a cushion against risk. Fourth, savings that augment wealth can perpetuate dynastic entrepreneurship, enabling entrepreneurial families to continue accumulating wealth over many decades, and even centuries. Tax reasons might also help explain high savings rates and wealth levels of entrepreneurs. Progressive taxation of labour income encourages owners of incorporated businesses to divert profits from earnings into assets, since dividends are typically more lightly taxed. There can also be tax advantages to self-finance (Kari and Karikallio, 2007). Although there are some dynamic models of entrepreneurial wealth transfers and accumulation,4 none of them explains the size distribution of entrepreneurs’ wealth, or how it compares with the distribution of entrepreneurial incomes. According to estimates compiled by Parker (2003a), older self-employed Britons enjoy above-average wealth holdings, while wealth inequality in this group is modest.

266

Financing

9.2 The ‘private equity premium puzzle’

In an influential article, Moskowitz and Vissing-Jørgensen (2002) analysed the portfolio risk-and-return patterns of American entrepreneurs and highlighted the following ‘private equity premium puzzle’. Entrepreneurs tend to invest in undiversified and relatively risky private businesses, yet tolerate an average rate of return which is no greater than what is obtainable from a diversified (and hence less risky) publicly traded portfolio of equities. This is a puzzle because if individuals are risk-averse, they should only be willing to bear the greater risk of entrepreneurship if they are rewarded with a higher return. Entrepreneurs bear some element of undiversified risk because their investment opportunities are not market assets which are equally available to all investors. Some reasons why entrepreneurs hold high shares of their wealth in undiversified businesses are explored in the next section. For now, it is of interest to consider the costs that entrepreneurs bear as a result of the private equity premium puzzle. One way of estimating this cost is by calculating the implicit cost of capital which entrepreneurs face at the time of IPO. According to Kerins et al.’s (2004) analysis of high-tech IPOs, an entrepreneur’s opportunity cost over a one-year holding period is generally between two and four times as high as that of a well-diversified investor. To put this into perspective, for an entrepreneur with 25 per cent of her total wealth invested in her venture, the cost of capital is 40 per cent. If entrepreneurs really are risk-averse, as the evidence suggests (chapter 4), it is desirable to find reasons why entrepreneurs are willing to settle for average returns which do not appear to compensate them properly for the extra risk that they bear. Several explanations have been proposed. First, individuals do not put their human capital at risk when they become entrepreneurs. In particular, future earnings in entrepreneurship or paid employment are unaffected by the risk of the current business. Hence the risk of an individual’s total net worth, which includes the present value of human capital, is much less than that of their financial wealth alone. In a calibration exercise which factors the present value of human capital into estimates of individuals’ total net worth, Polkovnichenko (2003) shows that only small non-pecuniary benefits, amounting to as little as 1.5 per cent of returns, are sufficient to induce individuals to turn entrepreneur despite the greater undiversifiable risk they face in entrepreneurship. For less educated people, with lower present values of human capital, this argument presumably becomes less compelling. On the other hand, less educated people usually have relatively little financial wealth, so the private equity premium puzzle is probably less of an issue for them anyway. For these individuals, the main locus of comparison between entrepreneurship and wage work probably relates to relative wages, as discussed in chapters 2–4. A second potential explanation of the private equity premium puzzle is that entrepreneurship provides additional private benefits which should be capitalised and incorporated into the calculus of relative occupational benefits. An important example of such a benefit is the opportunity to learn about one’s idiosyncratic ability to

Wealth and entrepreneurship

267

V(B)

Ret urns

R(B)

0 B

Figure 9.1 The convex outer envelope of returns in entrepreneurship, V(B), and paid employment, R(B)

generate business returns (Hintermaier and Steinberger, 2005). Only people who try entrepreneurship are able to realise this benefit, which helps them to make better portfolio decisions not only instantaneously, but also over the rest of their lifetimes. It may be worth accepting a riskier return in the short term in return for this long-term benefit. Third, the value of entrepreneurship when there is an option to quit can be a convex function of wealth. If so, entrepreneurs who can choose a project from a set of projects with common mean returns but different degrees of risk might optimally choose to operate a relatively risky project (Hopenhayn and Vereschagina, 2003). To see why, consider Figure 9.1, which illustrates two concave functions of wealth, B. These are the returns in entrepreneurship as a function of wealth, V (B), and the returns in paid employment, R(B). Individuals’ best choices are given as the ‘outer envelope’ of these functions, illustrated in bold in Figure 9.1. If the two functions cross once, the outer envelope is a convex function. As noted in chapter 2, individuals should choose to operate riskier projects when the return function is convex. So entrepreneurs with wealth around the point of intersection in Figure 9.1, and faced with a choice of projects with different risk profiles, might optimally choose to operate riskier projects. They do not need a risk premium to do so, so there is no private equity premium puzzle to explain. A fourth explanation of the puzzle is based on a preference for skewed returns. A preference of this kind has previously been observed in the context of gambling, where it can explain why risk-averse consumers willingly play lotteries yielding negative expected, but highly positively skewed, returns (Golec and Tamarkin, 1998). The same logic can also be applied to entrepreneurship, as demonstrated in the first part of the appendix to this chapter. It is shown there that individuals might participate in risky entrepreneurship even if expected returns are negative (µ1 < 0, µ2 > 0), as long as they are compensated by a sufficiently large positive skewness of returns, µ3 > 0. It is

268

Financing

important to note that a preference for skewness is not the mark of an inveterate gambler or a hopeless optimist, but is instead a common trait of a sober-minded risk-averse person with decreasing or constant absolute risk aversion. At the time of writing, research continues into efforts to resolve the private equity premium puzzle. It is unclear at present which (if any) of the above explanations best explains the puzzle. Finding out which does is a valuable research topic, which promises to shed further light on the dynamic structure of entrepreneurial choices. 9.3

Entrepreneurial wealth diversification

The financial portfolios of even very wealthy entrepreneurs are often remarkably undiversified. On average, about 40 per cent of an entrepreneur’s total wealth is tied up in their firm’s equity.5 According to estimates based on the SCF (Survey of Consumer Finances), around 64 per cent of entrepreneurs own their firm in its entirety, with another 10 per cent owning exactly one-half of the equity (Bitler et al., 2005). As noted above, limited asset diversification has been proposed as a possible answer to the entrepreneurial private equity premium puzzle. But this begs the question of why entrepreneurs do not diversify their wealth. Several possible explanations for undiversified asset holdings by entrepreneurs have been proposed. One was noted by Moskowitz and Vissing-Jørgenson (2002) themselves, namely correlated returns from business ownership and the entrepreneur’s human capital, thereby reducing the range of other uses to which their human capital can be put. Asecond set of possibilities relates to work effort by entrepreneurs. In contrast to most paid employment jobs, entrepreneurship offers ample opportunities for flexible working. If entrepreneurs can vary their labour supply relatively freely (see chapter 12), then households can choose riskier financial portfolios (e.g. undiversified entrepreneurship) because flexible labour supply provides an additional hedge against financial risk (Bodie et al., 1992). A different work effort argument focuses on incentive problems arising from diluted ownership stakes. In particular, high ownership stakes might be necessary to provide entrepreneurs with incentives to supply discretionary and non-contractible effort. Without entrepreneurs holding a majority stake in a venture, outside investors would fear being expropriated since entrepreneurs would possess greater incentives to slack. Bitler et al. (2005) adduced some evidence in favour of this hypothesis. Using data from the SCF and NSSBF, Bitler et al. (2005) showed that entrepreneurial effort as measured by weekly hours of work was positively associated with entrepreneurs’ ownership shares; furthermore, effort translated into better performance. As Bitler et al. noted, however, while their findings might explain concentrated ownership shares, they do not explain the puzzle about why these people become entrepreneurs in the first place – another important part of the private equity premium puzzle. A third possible reason for limited entrepreneurial diversification is that almost by definition, small firms are dependent on a limited pool of management expertise, which is a key factor in the success or failure of a business (Hall, 1992). This argument is

Wealth and entrepreneurship

269

predicated on the notion that entrepreneurs’ business assets are fundamentally illiquid, or indivisible. However, there is a problem with this explanation, because portfolios of continuing entrepreneurs tend to become more rather than less undiversified over time. So this finding does not support the explanation of down-payment constraints when starting a new business. Finally, control issues might make it optimal for entrepreneurs not to diversify their wealth. In addition to a simple non-pecuniary preference for sole control, entrepreneurs might prefer to keep their firm private rather than take it public (and thereby obtain greater access to capital) because they require decision-making autonomy to manage their firm optimally (Boot et al., 2006).

9.4 Wealth and entrepreneurship: theories

The possibility of a relationship between wealth and entrepreneurship is without doubt one of the most heavily researched topics in the economics of entrepreneurship. This relationship is of interest for several reasons, the most prominent being the possibility that it reflects the existence of credit constraints. This section provides an overview of the principal theories underlying the concept of a wealth–entrepreneurship relationship. These theories are based on a variety of models of borrowing constraints in entrepreneurial credit markets. The first subsection discusses the pioneering article of Evans and Jovanovic (1989). This is followed by other important contributions proposed by Banerjee and Newman (1993), Aghion and Bolton (1997) and Newman (2007). The emphasis in this section is on theory; evidence is evaluated in the section which follows. 9.4.1

The Evans and Jovanovic (1989) model Consider a set of individuals who can only borrow up to a multiple γ ≥ 1 of their initial assets, B, to start a business where γ is common to all individuals. Entrepreneurs can therefore operate only capital k ∈ (0, γ B). This rule, which is sometimes also referred to as a ‘limited liability constraint’ (Paulson et al., 2006), corresponds to the case of Type I credit rationing, since banks are willing to extend loans to everyone with some assets, up to some given asset-determined limit, irrespective of the interest rate entrepreneurs are prepared to pay. Although Evans and Jovanovic (1989) (referred to as EJ hereafter) did not explain why banks impose this ‘lending multiple’ policy, wealth-based loan sizes emerge naturally from a moral hazard problem, outlined below, in which entrepreneurs can ‘take the money and run’ (Banerjee and Newman, 1993). EJ assumed that borrowing and production take place in a single period. Entrepreneurs’ incomes y depend on k via the production function y = xk α , where x is managerial ability and α ∈ (0, 1) is a parameter. Constrained borrowers enter entrepreneurship only if their earnings net of capital repayments D = (1 + r)γ B (where r > 0 is the nominal interest rate) exceed their earnings in paid employment, w.

270

Financing

This occurs if x(γ B)α − (1 + r)γ B > w.

(9.1)

If a sample of individuals with given abilities are drawn at random from the population, then the probability that they are entrepreneurs is an increasing function of assets B, as can be verified by differentiating the LHS of (9.1). This is the first of two predictions of the EJ model: 1. There is a positive relationship between the probability of entering entrepreneurship, and assets prior to entering entrepreneurship. 2. Wealthier entrepreneurs will operate larger enterprises on average than poorer ones and will receive higher incomes. Perhaps counter-intuitively, however, the model predicts that, conditional on wealth, the most constrained entrepreneurs are those with the highest abilities, x.6 It will be seen later in the chapter that actual data are inconsistent with this prediction, at least if x is measured as human capital. To deal with this problem, one could extend the EJ model, perhaps by linking ability or human capital to easier access to supplies of financial capital. This could happen if greater human capital is associated with higher levels of income and savings which banks use as the basis for determining loan sizes, or if human capital serves as a positive signal of productivity which releases funds from lenders (Parker and van Praag, 2006a). Such extensions would render the sign of the theoretical relationship between human capital and borrowing constraints ambiguous (Astebro and Bernhardt, 2003, 2005). It is straightforward to use the EJ framework to demonstrate that an empirical relationship between wealth and participation in entrepreneurship is a sufficient condition for identifying imperfect capital markets. To see this, let output in entrepreneurship be f (k), and let  and φ be the cdf and pdf of the distribution of wages, w. Using the occupational choice model outlined in chapters 2 and 3, the probability that an individual becomes an entrepreneur is z ∗ = [ f (k) − (1 + r)k]. Hence ∂z ∗ /∂k = φ[f
(k) − r]. If capital markets are perfect, f
(k) = r and so z ∗ does not depend on capital. But imperfect capital markets imply f
(k) > r (because constrained borrowers necessarily have k < k ∗ ) – so z ∗ being positively related to k is a sufficient condition to indicate imperfect capital markets. It is not, however, a necessary condition. EJ estimated a simple probit model of entry into self-employment conditioned on assets (in levels and squared) – as well as experience in paid employment, education and assorted personal characteristics. Using NLS (National Longitudinal Survey) data on 1,500 white males over 1978–81 who were wage earners in 1976, EJ reported a positive and significant probit coefficient (p-value = 0.02) on initial assets. This supports prediction 1 above. Also, log self-employment incomes were significantly and positively related to log initial assets, supporting prediction 2. EJ went on to estimate a structural model of occupational choice with borrowing constraints. They estimated γ to be 1.44, and significantly greater than 1; a subsequent estimate by Xu (1998) based on

Wealth and entrepreneurship

271

more accurate data found γˆ = 2.01. Also, EJ estimated that 94 per cent of individuals likely to start a business faced Type I credit rationing. They claimed that this prevented 1.3 per cent of Americans from trying entrepreneurship. These are large effects, which have encouraged numerous subsequent researchers to explore the empirical robustness of the relationship between entrepreneurial participation and wealth. The empirical findings from these studies are reviewed in the next section of this chapter. The EJ model also implies that the distribution of wealth will influence the amount of entrepreneurship. That is because a mean preserving spread (see chapter 2) in the wealth distribution will make entry easier in the presence of borrowing constraints by increasing the number of people in the right-hand tail with sufficient wealth to selffinance or borrow the funds required for entry (Lindh and Ohlsson, 1998). This can be seen directly from (9.1), since regressively redistributing wealth from those with very low B (who will never enter anyway) to those with a higher B, but who are just unable to enter, increases entry. Using a time-series of Swedish data, Lindh and Ohlsson (1998) went on to claim an empirical link between wealth inequality and the aggregate self-employment rate. However, subsequent theorising suggests that the relationship between wealth inequality and self-employment could be more subtle, being positive when start-up costs are high but turning negative when start-up costs are low. In an empirical application of this idea, Mesnard and Ravillion (2006) obtained estimates of a negative relationship between wealth inequality and self-employment using Tunisian micro-data, a country in which start-up costs are very low on average. 9.4.2

The Banerjee and Newman (1993) model

EJ’s assumption of a wealth-based lending rule can be motivated in several ways. Banerjee and Newman (1993) suggested one rationale, based on the idea that entrepreneurs can ‘take the money and run’. To abstract from unnecessary bankruptcy issues, suppose for simplicity and without loss of generality that project returns are always high enough for entrepreneurs to be able to repay their loans. All entrepreneurs post their entire wealth B as collateral to borrow the amount k, where k > B. In return, they obtain the payoff f (k). The key point is that once they obtain their payoff, entrepreneurs can choose to renege on their repayment (1 + r)k, taking the money f (k) and running away, albeit at the cost of lost collateral, (1+r)B. At the end of the period, the entrepreneur escapes the lender’s efforts to find her with probability 1 − p. If she is caught, with probability p, she consumes all her income and is subject only to a non-monetary punishment c which reduces her utility in a linear fashion. Reneging therefore yields a payoff of f (k) − pc, while repayment yields f (k)+(1+r)(B−k). The entrepreneur will therefore renege whenever (1 + r)B + pc < k(1 + r). Knowing this, lenders will only make loans that satisfy k ≤ B + [pc/(1 + r)].

(9.2)

All loans made in equilibrium will satisfy this constraint, and entrepreneurs who receive loans will never renege. Only wealthy entrepreneurs receive loans, however: (9.2)

272

Financing

ties loan sizes directly to initial personal wealth, as assumed (but not derived) by EJ. 9.4.3 The Aghion and Bolton (1997) model This outcome of the poor being denied access to credit also emerges from a model proposed by Aghion and Bolton (1997) (referred to as AB hereafter), in which moral hazard problems turn out to be most acute at low wealth levels. Suppose entrepreneurs need a unit of capital to set up a firm. If their wealth is less than unity, i.e. B < 1, they must borrow, owing the bank D(B) per unit borrowed. The project generates a positive payoff Rs with probability p, and the payoff Rf = 0 with probability 1 − p. The entrepreneur can increase p by exerting more effort, but this costs the entrepreneur c(p) = rp2 /2α in utility, where α ∈ (0, 1]. Hence the entrepreneur faces the problem

max{pRs − p(1 − B)D(B) − c(p)} . p

  The solution to this optimisation problem is given by p(B) = α 1 − (1 − B) D(B) Rs . It

turns out that p
(B) > 0, i.e. poorer borrowers have less incentive to supply effort, as they must share a larger fraction of the marginal returns from effort with the bank. As a result, competitive banks optimally cut off credit to (‘redline’) everyone below some critical wealth threshold: only the wealthy can become entrepreneurs.7 Paulson et al. (2006) designed an encompassing empirical framework which nests the EJ and AB models as special cases, and permits the researcher to test which of the models is most congruent with the evidence. The structural model underlying their approach, and the results obtained from estimating it, are outlined in the second part of the appendix to this chapter. 9.4.4

Newman’s (2007) moral hazard model Chapter 2 discussed a moral hazard problem by Newman (2007) in which, counterintuitively, the poorest individuals become entrepreneurs and the wealthiest work for them. As noted there, Newman interpreted this outcome as being problematic for riskattitude-based models of entrepreneurial choice. Taking Newman’s result at face value, borrowing constraints which deny loans to those without wealth will presumably result in no entrepreneurship at all. In fact, modifying the Newman model by adding bankruptcy costs changes its predictions in a more realistic way. Bankruptcy costs mean that the very poorest no longer choose to become entrepreneurs because they can no longer afford to risk failure: their marginal disutility from failure becomes too high in the bad state (Banerji and van Long, 2007). In this case, only individuals with moderate wealth become entrepreneurs. The very richest and poorest choose not to, the former because they must bear an unacceptable level of income risk to be made incentive-compatible in entrepreneurship, and the latter because they cannot afford the bankruptcy cost.

Wealth and entrepreneurship

273

9.5 Wealth and entrepreneurship: evidence

Following the pioneering work of EJ, many researchers have used cross-sectional and/or longitudinal data to estimate binary choice models of entrepreneurship which include some measure of individuals’ wealth or asset windfalls among the vector of explanatory variables. Most studies report a significant positive relationship between asset endowments and entrepreneurship,8 while others find that home ownership and house prices are linked with new firm formation and self-employment at both the individual and the regional levels.9 In contrast, a handful of studies have detected an insignificant wealth/entrepreneurship relationship.10 The most influential of these studies is Hurst and Lusardi (2004), whose contribution will be analysed in greater detail shortly. Other evidence casts doubt on EJ’s model, too. Researchers have reported a positive relationship between access to capital and human capital (Bates, 1997; Astebro and Bernhardt, 2003, 2005) which, as noted above, contradicts the EJ model. This result has proven robust to allowing human capital to be endogenous (Parker and van Praag, 2006a). A key issue in this literature is the need to recognise and deal with the potential endogeneity of wealth. Endogeneity may have generated spurious estimates of a wealth– entrepreneurship relationship in previous work. One potential source of endogeneity from assets is reverse causality: entrepreneurship makes people wealthy rather than wealth making people become entrepreneurs. In an attempt to overcome this problem, some researchers have analysed the effects of windfalls and inheritances received prior to market entry. Examples of windfalls include lottery wins, redundancy lump sums and favourable exchange rate movements which increase the value of remittances received in developing countries by relatives working overseas.11 Empirical studies generally report positive, significant and substantial effects of windfalls on self-employment status and entry probabilities, with diminishing returns from higher windfall values.12 To give a flavour of the results, Blanchflower and Oswald (1990) reported that a Briton who received £5,000 in 1981 prices was twice as likely to be self-employed in 1981 as an otherwise comparable person who had received nothing. For the USA, Holtz-Eakin et al. (1994a) estimated that a $100,000 inheritance would increase the probability of a transition from paid employment to self-employment by 3.3 percentage points. And Lindh and Ohlsson (1996) estimated that the probability of self-employment in Sweden would increase by 54 per cent if lottery winnings were received, and by 27 per cent following the receipt of an average-sized inheritance. Findings of a positive and significant role for windfalls seem to be robust to several possible objections and criticisms, including entrepreneurs being more willing to gamble on lotteries (the opposite appears to be the case according to Lindh and Ohlsson, 1996 and Uusitalo, 2001); to inheritances being anticipated or taking the form of family businesses; and to industry differences in capital requirements (Bates, 1995). Indeed, true effects from windfalls may be understated to the extent that researchers cannot always measure delayed entries into entrepreneurship following the receipt of a windfall. Entry decisions may take several years to play out. On the other hand, there is some

274

Financing

evidence that future inheritances predict business formation as well as past inheritances. This implies that inheritance is not a good instrument for capital constraints, as by definition capital-constrained families cannot borrow against future inheritances (Hurst and Lusardi, 2004). Instead, it is possible that inheritances are capturing something other than borrowing constraints, for example a wealthy and successful family background. This raises the possibility that windfall studies, as well as ones based on personal wealth variables, could be vulnerable to endogeneity biases arising from unobserved attributes which make some individuals both more interested in accumulating wealth and in becoming entrepreneurs. If so, it would look as though wealth is correlated with entrepreneurship even when it is not. Some direct evidence of this type of endogeneity appears in Hochguertel (2005a), who estimated the following structural model using a panel of Dutch household data from the 1990s: sit∗ = Xit ’βs + γss sit−1 + γsb ln Bit−1 + αsi + usit ln Bit∗

= Xit ’βb + γbs sit−1 + γbb ln Bit−1 + αbi + ubit  sit =

(9.3) (9.4)

1 if sit∗ > 0 0 otherwise

where s is participation in entrepreneurship (self-employment); B is personal net wealth; the αs are (possibly correlated) fixed effects; and the us are random disturbance terms. If γsb > 0, a positive wealth–entrepreneurship relationship exists, while a positive correlation between the αs means that people who find it attractive to become entrepreneurs also have a preference for acquiring wealth. Hochguertel (2005a) estimated both αs in (9.3) and (9.4) to be statistically significant and strongly positively correlated (corr(αs , αb ) = 0.510, with estimated standard error = 0.066). This suggests correlated unobservable preferences. ML estimates of the γ s are summarised in Table 9.1. According to these estimates, lagged wealth does not affect participation in entrepreneurship, consistent with Hurst and Lusardi (2004) and contradicting the borrowing constraint hypothesis. There is instead clear evidence of reverse causality, with participation in entrepreneurship generating greater wealth (γˆbs > 0). This is also consistent with other US evidence, which shows that following entry, entrepreneurs exhibit high growth rates in wealth and savings, while entrepreneurs who exit tend to dissave (Gentry and Hubbard, 2004). Strong statedependence in entrepreneurship, γˆss , also emerges from the estimates, in accordance with Henley’s (2004) findings based on British BHPS data. These results suggest that researchers need to correct for endogeneity by using robust estimators, such as IV (see chapter 3). Recent studies are beginning to address this problem, using a range of identification strategies. Hurst and Lusardi (2004) were one of the first research teams to instrument wealth, their instrument being house price appreciation.13 Using PSID (Panel Study of Income Dynamics) data, Hurst and Lusardi estimated the relationship between becoming a business owner and wealth to be flat up to the 95th percentile of the wealth distribution, before becoming steep and positive.14 This result suggests that there is no relationship between wealth and entrepreneurship

Wealth and entrepreneurship

275

Table 9.1. Estimates of a simultaneous equation model of entrepreneurship and wealth

Entrepreneurship at t − 1 Wealth at t − 1

Entrepreneurship at t

Wealth at t

γˆss = 2.762a (0.137) γˆsb = 0.000 (0.052)

γˆbs = 1.036a (0.295) γˆbb = 0.152a (0.009)

Note: Entries are coefficient estimates with estimated standard errors in parentheses beneath; a denotes statistical significance at the 1% level. Source: Hochguertel (2005a).

for most people, and casts doubt on the borrowing constraint hypothesis since it is hard to believe that the rich are liquidity constrained. Hurst and Lusardi’s (2004) findings have been challenged in subsequent research also based on PSID data and enlarged with CPS data to cover more metropolitan areas and a longer time period. Using these richer data, Fairlie and Krashinsky (2006) discovered that the spike at the top of the wealth distribution is associated with wealthy older job losers who face limited job opportunities in paid employment and turn to entrepreneurship instead (q.v. Farber, 1999). Strikingly, when Fairlie and Krashinsky split the sample into job losers and non-job losers, they discovered increasing rates of entry into self-employment for both groups throughout the asset distribution. This suggests that Hurst and Lusardi’s (2004) findings might be an artefact of aggregation across dissimilar groups. Furthermore, Mesnard and Ravillion (2006) used IV to estimate the effects of wealth accumulated abroad by Tunisian émigrés on their propensities to become self-employed once they returned to Tunisia. Mesnard and Ravillion’s identifying instrument was the date of emigration. The reason was a new Tunisian law which made emigration to relatively rich European countries more difficult after 1974. Emigrants who left before 1974 had opportunities to acquire more wealth than those who emigrated afterwards. Since the policy change was exogenous, it should not affect the decision to become self-employed upon eventual return to Tunisia. Mesnard and Ravillion’s IV estimates revealed a positive and significant effect from wealth, with a concave diminishing effect; non-parametric estimation yielded a similar pattern. But while they are interesting, these findings still do not establish the existence of borrowing constraints. Several reasons why they do not are explained in the next section. 9.6 Alternative interpretations of a wealth/entrepreneurship relationship

While findings of a wealth–entrepreneurship relationship are consistent with borrowing constraints, alternative explanations which do not depend on the existence of borrowing constraints can also be proposed: 1. Inherently acquisitive individuals both accumulate assets and prefer entrepreneurship to working in paid employment. As noted in the previous section, this

276

2.

3.

4.

5.

Financing

mechanism is associated with a positive relationship between entrepreneurship and wealth, whether or not borrowing constraints exist. Entrepreneurs prefer self-finance to external finance. One reason might be that entrepreneurs regard the terms of external finance to be unreasonable or unprofitable. For example, suppose bank administrative costs drive a wedge between borrowing rates, rb , and saving rates, rs , where rs < rb . If the entrepreneur’s net present value conditioned on the interest rate has the ordering NPV (rb ) < 0 < NPV (rs ) then windfalls which turn borrowers into savers can increase participation in entrepreneurship even in the absence of borrowing constraints. Another reason why self-finance is preferred to external finance might be ‘control aversion’, whereby entrepreneurs do not wish to be dependent upon the decisions of banks (Cressy, 2006b). Consequently, people wait until they have saved (or inherited) enough wealth to enter entrepreneurship without borrowing. Yet all the while banks may have been willing to lend all of the required funds to every loan applicant. Alternatively, consider overoptimistic individuals who propose ventures for financing which banks perceive to be unprofitable at the proposed scale of operation. Banks therefore rationally refuse to extend all of the credit requested by entrepreneurs. The latter then delay entry until they have sufficient wealth, giving rise to a positive wealth–entrepreneurship relationship despite the actual absence of any borrowing constraints. Leisure is a normal good, so wealthier people prefer entrepreneurship because its flexibility enables them to reduce their work hours, in contrast to paid employment where standardised work-week restrictions prevail (Harada and Kijima, 2005). This can happen irrespective of whether or not borrowing constraints exist.15 Apositive association between start-ups and wealth (or windfalls such as inheritances and lottery winnings) might simply reflect the effects of decreasing absolute risk aversion (DARA) rather than borrowing constraints. Consider again the Kihlstrom and Laffont (1979) model outlined in chapter 2. Under DARA, an increase in the wealth of the marginal risk-averse individual makes them more willing to enter risky entrepreneurship, so increasing the aggregate rate of entrepreneurship (Cressy, 2000; Newman, 2007). However, there are several objections to the DARA explanation. In theory, DARA implies that entrepreneurs running larger firms will take more risks and so will have more variable returns. But this is contrary to the evidence (Hopenhayn and Vereshchagina, 2003). And it has been observed that even after controlling for measures of risk preferences, wealth still has a significant positive effect on transitions into self-employment (Kan and Tsai, 2006). Some asymmetric information models have the property that a positive relationship between the number of entrepreneurs and personal wealth is consistent with overinvestment rather than Type I credit rationing (de Meza and Webb, 1999). And if project returns decay with delay in exploiting them, over-investment models also imply that less wealthy entrepreneurs will undertake their projects at a smaller scale and will succeed less often, and may therefore choose ex ante not to undertake them at all (de Meza and Webb, 2006). The implication is once again a positive relationship between wealth and entrepreneurship.

Wealth and entrepreneurship

277

6. Financially distressed firms face higher interest rates and restricted credit (Harhoff and Körting, 1998). An inheritance or windfall merely relaxes the entrepreneur’s budget constraint and permits them to survive in business a little longer. Conversely, when wealth is plentiful, there is less financial distress and greater participation in entrepreneurship. The resulting competition discourages entry from those with sufficient wealth but poor investment projects – leading to a higher overall business survival rate (Black et al., 1996). 7. In developing economies, wealthy entrepreneurs often enjoy market power. This gives them the most favourable loan terms, inducing them to become entrepreneurs while the less well off do best by lending their capital to the entrepreneurs (Gabszewicz and Laussel, 2006). All of these arguments can explain how wealth promotes entry and persistence in entrepreneurship, despite an absence of borrowing constraints. And even if borrowing constraints do exist, their effects may be limited in scope. There are two major reasons for this, the discussion of which will take up the remainder of this section. One is that most entrepreneurs start ventures which need little or no capital. The other is that even when capital is required but is for some reason unavailable, entrepreneurs can in principle find ‘escape mechanisms’ to circumvent the constraints. There is abundant evidence that lack of capital need not deter entry. According to Meyer (1990), 60 per cent of entrants to entrepreneurship possess no depreciable capital. Borrowing constraints evidently did not prevent these entrepreneurs from entering. Furthermore, Hurst and Lusardi (2004) reported that low-wealth households are no less likely to start up in industries requiring high capital investments. Capital requirements for most business start-ups seem to be modest in any case. Dennis (1998) cites a 1992 survey of Inc.’s 500 fastest-growing firms which found that one-quarter of these firms started up with funds of less than $5,000. Nor need any entrepreneurs denied requisite finance necessarily be passive victims of borrowing constraints. Entrepreneurs have incentives to seek ‘escape routes’ such as finding other sources of funds (Bhide, 2000), or saving (Parker, 2000; Ghatak et al., 2001). The creative use of meeting resource needs without recourse to external finance is known as ‘bootstrapping’. This practice can involve leveraging credit cards and second mortgages; delaying payment to suppliers; and minimising stock and accounts receivable. According to Bhide (2000, p. 15), more than 80 per cent of Inc. 500 founders bootstrapped their ventures with modest funds; their median start-up capital was about $10,000.16 Entrepreneurs are known to have above-average savings rates (Quadrini 1999, 2000). In theory, the prospect of saving to overcome borrowing constraints can incentivise workers to supply high levels of effort, thereby increasing social welfare (Ghatak et al., 2001). So policies which reduce credit market imperfections may reduce welfare by dampening this effect.17 Unfortunately, there is still insufficient evidence to determine whether savings enable entrepreneurs to circumvent borrowing constraints in practice. Some simulation evidence is suggestive, though not decisive. For example,

278

Financing

Bohàc˘ ek (2006) calibrates a dynamic model which combines a Banerjee–Newman style borrowing constraint with a saving incentive, allowing entrepreneurs to accumulate wealth to escape the constraint. Bohàc˘ ek’s (2006) simulation reveals that this escape mechanism can explain high entrepreneurial savings rates and the pronounced wealth inequality observed in the US economy. However, different calibration models analysed by other researchers can also explain these stylised facts (q.v. section 9.1 above), implying an identification problem with the calibration approach. Some anecdotal evidence also suggests that some people can escape borrowing constraints by saving. This includes Yoon’s (1997) survey of Korean business owners in Chicago and LosAngeles, and Paulson and Townsend’s (2004) analysis of entrepreneurs in Thailand. For instance, Yoon (1997) observed that the respondents to his survey often started as manual, service or sales workers; accumulated capital mainly through personal savings; and then bought the business from the existing owner. Ghatak et al. (2001) cite some other examples of this kind. The notion of saving as an escape mechanism has its own limitations, however. If some individuals are so impatient that they prefer consuming while young instead of saving for the future, they might optimally choose to remain credit-constrained forever (Parker, 2000). This implies that, in some cases, borrowing constraints are better regarded as voluntary than as involuntary. A second caveat to the escape mechanism is that some employees may earn too little to permit them to build up sufficient savings to become entrepreneurs. At very low levels of income and consumption, reducing consumption in order to accumulate assets may be sub-optimal because it can seriously threaten health, production efficiency and longevity (Gersovitz, 1983). While this problem might not be widespread in developed economies, it could be important in some developing countries. Of course, savings are not the only escape mechanisms available to financially constrained entrepreneurs. Another possibility is that entrepreneurs borrow from their employees (Michelacci and Quadrini, 2005). They can do so by launching direct appeals to workers for funds in return for promissory notes; or by offering their workers stock options, whereby firms delay the cash compensation of their employees and effectively borrow from them. Or entrepreneurs can offer workers long-term wage contracts with an increasing profile. In the last of these cases, entrepreneurs effectively borrow from workers while constrained, and then compensate them later as their ventures grow from the proceeds of the borrowing. This does, however, require the entrepreneur to design commitment mechanisms which make long-term rising wage contracts ‘renegotiation proof’. One interesting implication of this line of reasoning is that financially constrained firms are younger and smaller, grow faster and pay lower wages. Another implication is that capital constraints can explain why small entrepreneurial firms have low capital/labour ratios and therefore lower labour productivity (Garmaise, 2008). All of these predictions are consistent with the evidence.18 To conclude, even if entrepreneurial participation is positively related to personal wealth and financial windfalls, this does not prove the existence of borrowing

Wealth and entrepreneurship

279

constraints. Several alternative explanations are also consistent with the evidence. And it should be remembered that borrowing constraints are problematic only if they prevent positive NPV projects from being undertaken. If they prevent negative NPV projects from going ahead, borrowing constraints can be positively beneficial. In short, the present state of the literature does not allow us to resolve definitively the vexed question of whether borrowing constraints exist, and if so, how widespread they are. Sharper tests, possibly based on bank loan application micro-data matched with detailed borrower surveys, are needed before any firmer conclusions can be drawn.

9.7 Wealth and performance in entrepreneurship

Researchers have explored at least three ways that wealth affects entrepreneurial performance and behaviour. Wealth can affect survival, investment choices and profitability. The arguments for, and evidence relating to, these three effects are now reviewed in turn. 9.7.1 Effects of wealth on venture survival As well as facilitating start-ups, it seems likely that personal wealth can also enhance a new venture’s survival prospects. As Holtz-Eakin et al. put it: ‘if entrepreneurs cannot borrow to attain their profit-maximising levels of capital, then those entrepreneurs who have substantial personal financial resources will be more successful than those who do not’ (1994b, p. 53). These authors argue that unexpected increases in wealth (such as inheritances) will make entrepreneurship more attractive if the alternative is a business operating with a sub-optimal capital stock. But in the absence of borrowing constraints, entrepreneurs will presumably operate their optimal capital stocks, so an inheritance may actually make outside options like paid employment or tax sheltering more attractive, promoting voluntary exits rather than entries. So if it turns out that inheritances are positively associated with company survival, this can be taken as relatively strong evidence of Type I credit rationing. In fact, Holtz-Eakin et al. (1994b) did find that inheritances increase the probability that self-employed Americans remain in self-employment. Similar results have been obtained by other researchers using measures of personal wealth rather than windfalls.19 On the other hand, findings that higher-skilled entrepreneurs are less capitalconstrained than the average have led some researchers to argue that human capital is the real driver of entrepreneurial survival, rather than financial capital (which is related to human capital). For instance, Cressy (1996) claims that failing to control for human capital endows personal (housing) wealth with spurious explanatory power in UK business survival models. However, a growing body of US evidence generally rejects the view that financial inputs are unimportant determinants of survival, even after human capital is controlled for. This evidence will be discussed in chapter 14.

280

Financing

9.7.2

Effects of wealth on venture investment decisions In an influential article, Fazzari et al. (1988) reported that investment by large US firms is closely related to cash flow, with greater investment sensitivity to cash flow among firms that they considered to be most vulnerable to Type I credit rationing. These findings appear to offer some support to the notion that borrowing is restricted, rendering cash flow an important source of funds. Several authors have subsequently replicated these findings.20 If this argument is true, limited access to capital might have macroeconomic implications as well. Monetary policy could influence the economy through a novel credit channel by affecting the investment decisions of entrepreneurs. Investment-cash flow sensitivity might also carry implications for economic development and export performance. Consider, for the sake of argument, two sectors. One sector is dominated by large capital-intensive firms which can self-finance investments and do not face credit constraints, while the other is dominated by small labour-intensive firms which do face credit constraints. If wealth alleviates credit constraints (Aghion and Bolton, 1997), then it follows that poor economies enjoy a comparative advantage by producing goods in the large-firm sector, while wealthier nations enjoy a comparative advantage by producing goods in the small-firm labour-intensive sector (Wynne, 2005). This might explain why so much of US export trade is labour-intensive, contrary to traditional trade theories which suggest that the USA might be expected to have a comparative advantage in capital-intensive goods. However, echoing the critique presented in the previous section of this chapter, investment sensitivity to cash flow need not reflect borrowing constraints. For example, such sensitivity can also arise if firms learn under uncertainty, as the smallest firms are often the youngest firms, which have the most imprecise beliefs and are therefore more likely to overreact to aggregate shocks affecting cash flow than larger and older firms (Li and Weinberg, 2003; Oliveira and Fortunato, 2006). In any case, as Hubbard (1998) describes it, the problem is that firms desire a loan size L∗ at the market interest rate r, but because of agency costs arising from imperfect information, banks are only willing to supply L∗ at a higher interest rate than r. As Jaffee and Stiglitz (1990, p. 847) point out, this involves ‘price rationing’ rather than credit rationing, since firms could obtain L∗ if they paid a higher interest rate.

9.7.3

Borrowing constraints and profitability Parker and van Praag (2006a) extended the Type I credit rationing model of Bernhardt (2000), outlined in chapter 7, by allowing entrepreneurs to differ from each other in terms of observed and unobserved ability. Unobserved ability is (by definition) undetected by banks which screen loan applicants, so random screening errors mean that some entrepreneurs face tighter constraints than others. Consequently, entrepreneurs make lower profits, π , the more constrained they are. On the other hand, years of schooling, sch, are associated with observed ability, which is assumed to increase entrepreneurs’ profits and to facilitate access to finance. Parker and van Praag (2006a)

Wealth and entrepreneurship

281

were able to employ a direct measure of capital constraints, con, namely the percentage of initial requested funding which was not granted by lenders: 0 ≤ con ≤ 100. This is an improvement over the indirect measures of borrowing constraints used in previous research, such as wealth, windfalls or start-up capital. Parker and van Praag (2006a) estimated the ‘triangular’ econometric model set out as equations (3.4) through (3.6) in chapter 3. This model recognises the potential endogeneity of capital constraints, con, and years of education, sch. Identifying instruments z1 for sch include father’s education and the number of siblings in the entrepreneur’s family. The identifying instrument z2 for con is the capital intensity of the industry in which an entrepreneur is active. The rationale is that screening errors – which are associated with variations in capital constraints – are more likely to occur in technologically advanced capital-intensive industries with higher proportions of intangible capital.21 Using a sample of Dutch entrepreneurs from 1995, and utilising an IV estimator, Parker and van Praag (2006aa) estimated the rate of return to education in entrepreneurship as 13.7 per cent (βˆs = 0.137). Second, βˆc = − 0.039, implying that a 1 percentage point relaxation of capital constraints increases entrepreneurs’ average business incomes by 3.9 per cent. Third, γˆs = − 1.183, so an extra year of schooling decreases capital constraints by 1.18 percentage points, all else equal. These estimates imply that schooling has an additional indirect effect on entrepreneurial performance, of βˆc γˆs = 0.039 × 1.183 = 0.046. To summarise, credit constraints appear to reduce profitability in entrepreneurship. The dependence of credit constraints on formal education makes the latter an even more effective determinant of entrepreneurial performance than was previously thought. This complements other evidence cited above claiming that access to capital is associated with enhanced entrepreneurial performance. But once again one must caution against rushing to interpret these findings as proof of inefficient capital markets, or arguing that firms which struggle to raise finance require public subsidies. Type I credit rationing in the Parker and van Praag (2006a) model is economically efficient. So failure to obtain finance can in fact amount to the best use of society’s scarce resources.

9.8

Evidence relating to Type II credit rationing

This section reviews evidence relating to outright denials of loans to entrepreneurs (i.e. Type II credit rationing), rather than access to smaller loan sizes than entrepreneurs request. As noted in chapter 7, I will concentrate on tests of equilibrium credit rationing, not ‘temporary’ or disequilibrium credit rationing arising from a temporary excess demand for credit while banks adjust their interest rates. Reflecting the emphasis in published research to date, the evidence discussed below comes from developed economies. The causes and effects of credit rationing in developing countries often simply reflect limited financial development (see, e.g., Levy, 1993; Kochar, 1997). This section has the following structure. First, I explain why several ‘popular’ indicators of credit rationing are in fact no such thing. Second, some influential micro-data

282

Financing

evidence by Berger and Udell (1992) is reviewed. Third, some miscellaneous additional evidence is considered to round out and conclude the discussion. 9.8.1 What does not constitute evidence of credit rationing

Claims by survey respondents about difficulties of raising capital and the existence of credit rationing should be treated with considerable caution. The fact that one-half of employee survey respondents claim to have seriously considered becoming selfemployed in the past, but blamed insufficient capital as the reason for not making the switch,22 does not necessarily mean that loans were unavailable to these respondents – or that these respondents were ever really very serious about becoming self-employed. Survey responses are prone to self-serving bias because entrepreneurs might blame banks for inherent shortcomings in their loan applications. Asking entrepreneurs what the price of loans should be is equally uninformative, since it is unlikely to reflect the genuine cost of funds. It might be thought that another way of establishing the existence of Type II credit rationing would be to compare survival rates of entrepreneurial ventures funded with government-backed loans that would not have been granted without the government guarantee, with survival rates of ventures that were financed purely privately. If the two survival rates are similar, one might argue that the government intervention alleviated genuine credit rationing. In fact, although there is some evidence that public-sector Loan Guarantee Scheme (LGS)-backed start-ups do have similar survival rates to purely privately funded start-ups, this conclusion does not necessarily follow (see chapter 16). The primary role of an LGS is to help fund entrepreneurs who lack collateral and/or a track record, and so are observably risky to finance. Banks refusing to finance ventures that they expect to yield insufficient returns for the risk borne cannot be construed as rationing credit. The whole point of a loan guarantee is to insure banks against most of the downside risk, turning some marginal investments into attractive lending opportunities. The evidence does suggest that these are indeed the projects that banks typically fund via an LGS (KPMG, 1999). 9.8.2 Berger and Udell’s (1992) approach To avoid the problems identified above, the results described next are based on observed micro-level behaviour rather than on subjective beliefs or co-movements of macroeconomic aggregates. This strategy introduces its own problems, because to identify Type II credit rationing directly, the researcher needs to obtain samples containing a set of ventures which are observationally identical to banks, some of which are given funds and others of which are not. Since such data are unavailable, actual tests of Type II credit rationing tend to be indirect in nature. Indirect tests of Type II credit rationing usually investigate the ‘stickiness’ of commercial loan rates. In the absence of Type II credit rationing, interest rates on commercial loans should adjust freely in response to changes in the supply of and demand for credit. But if credit rationing exists, that is no longer the case, and commercial loan rates can

Wealth and entrepreneurship

283

exhibit ‘stickiness’. This, at least, can be expected to apply to non-‘commitment loans’, where a commitment loan (CL) is a loan that a bank agrees to allocate to a firm in advance over a specified period should the latter request it. Contractually, a CL cannot be rationed or withdrawn once it has been granted, so it can be regarded as a hedge against the future hazard of Type II credit rationing. A problem with interpreting sticky loan rates as evidence of credit rationing is that alternative explanations for this phenomenon also exist. One is Fried and Howitt’s (1980) implicit contract theory, in which banks and borrowers agree to fix interest rates for long periods of time irrespective of economic conditions. Another is the ‘distressed company’ phenomenon, whereby banks prefer not to increase lending rates in line with base rates if that tips clients into bankruptcy and forces banks to incur bankruptcy costs. Both of these explanations can potentially account for Berger and Udell’s (1992) finding that up to 7 per cent of US commercial loans over 1977–88 charged interest rates at below the (safe) open-market rate. Early studies tested for loan rate stickiness using aggregate time-series data and reported mixed results (Slovin and Sushka, 1983; King, 1986). However, the highly aggregated nature of this research is a serious drawback, because it averages over heterogeneous loan types. Micro-data are to be preferred because of their greater detail and precision. Berger and Udell (1992) conducted an especially thorough analysis of the credit rationing hypothesis using a sample of one million commercial loans in the US between 1977 and 1988. Berger and Udell (1992) detected some loan rate stickiness, consistent with the credit rationing hypothesis. However, CL rates were also sticky, suggesting that reasons other than credit rationing must account for the observed stickiness. And less rather than more stickiness was observed during periods of ‘credit crunch’, which is the opposite of what one would expect from the credit rationing hypothesis. Furthermore, in times of credit crunch or low loan growth rates, the proportion of CLs was observed to decrease. The opposite would be expected to occur under credit rationing. On the basis of this evidence, Berger and Udell concluded that ‘information-based equilibrium credit rationing, if it exists, may be relatively small and economically insignificant’ (1992, p. 1071). They went on to suggest that, even if some borrowers are rationed, others take their place and receive bank loans. 9.8.3

Other evidence

A direct upper bound estimate of the extent of Type II credit rationing is the number of loans rejected by banks for any reason. If loan rejection rates are very low, then credit rationing cannot be very important. Using US data on small company borrowing over 1987–88, Levenson and Willard (2000) estimated that only 2.14 per cent of small firms ultimately failed to obtain the funding they sought. Of course, the actual extent of credit rationing will be lower than this to the extent that some of these loans were observably non-creditworthy and so deserved to be rejected. Levenson and Willard (2000) also reported that the probability of loan denial was negatively related to firm size. Hence the extent of credit rationing by value is even

284

Financing

less than 2 per cent. UK evidence from Cosh and Hughes (1994) tells a similar story. From a sample of firms surveyed between 1987 and 1989, Cosh and Hughes reported that only 3.2 per cent of firms seeking external finance failed to obtain it. Other evidence sheds further light on the Type II credit rationing phenomenon. Theories of rationing based on adverse selection suggest that bank managers prefer to deny credit to borrowers rather than to raise interest rates, even if rationed borrowers are willing to pay higher rates (see chapter 7). Evidence does show that bank managers are reluctant to raise interest rates above the bank’s ‘standard’ business rate for observably higher-risk projects. But this reluctance appears to be motivated less by concerns about adverse selection or moral hazard, than by anxiety that public relations could be harmed by an image of the bank as a ‘usurer’ (National Economic Research Associates, 1990). It may be noted that the bank managers who were interviewed in this study had no obvious motive for responding to this survey question with self-serving bias.23 On the other hand, even if Type II credit rationing is empirically unimportant, the perception of it might discourage potential entrepreneurs from approaching banks at all. Individuals who do not approach lenders for funds because they fear rejection are called ‘discouraged borrowers’ (Kon and Storey, 2003). Levenson and Willard (2000) estimated that 4.2 per cent of potential entrepreneurs are discouraged borrowers; Cowling (1998) provides a similar estimate for the UK. According to the 1989 SCF, as many as 9 per cent of US business owners stated that they thought of applying for credit but changed their mind because they thought they might be turned down (Cagetti and de Nardi, 2006). The literature on entrepreneurial finance and gender suggests that a higher proportion of discouraged borrowers are women rather than men (Coleman, 2007). There is a rational basis for the discouraged borrower phenomenon. Home ownership and the commitment of personal equity to ventures are known to be strongly linked to successful loan applications. Discouraged borrowers are less likely to own a home or to commit personal equity to their ventures, so they may be correctly anticipating what banks are looking for (Blumberg and Letterie, 2008).24 9.9

Conclusion

The chapter commenced by establishing the importance of wealth for entrepreneurship, and vice-versa. Entrepreneurs play a central role in aggregate wealth accumulation and savings. Savings are used to fund investment, both entrepreneurial and corporate, so they are needed to support economic growth. Research is only just beginning to reveal the scale of entrepreneurs’contributions to these important economic outcomes. Further research is needed in several areas, including empirical studies which can distinguish cleanly between competing explanations for the private equity premium puzzle and limited entrepreneurial wealth diversification. The need for further research is especially urgent in countries outside the United States, where most of the empirical work to date has been conducted. The chapter also analysed the possibility that borrowing constraints cause entrepreneurship and wealth to be related. Because causality between these variables

Wealth and entrepreneurship

285

could run both ways, and because unobserved characteristics might link wealth acquisition and entrepreneurship without any direct relationship between them, one should not automatically infer the existence of borrowing constraints from empirical studies which find a positive association between them. Personal wealth and windfalls are likely to be endogenous, so researchers should move on from simple probit estimations where wealth or windfalls are used as independent variables serving as proxies for borrowing constraints, and should instead isolate suitable identifying instruments. Better still, researchers should try to measure borrowing constraints more directly (see, e.g. Parker and van Praag, 2006a). Even if convincing instrumental variables for wealth are found, a positive association between wealth and entrepreneurship neither implies that wealth is necessary for entrepreneurial entry nor that credit is rationed. A range of other reasons, which have nothing to do with imperfect capital markets, can also explain this association. It is noteworthy that alternative reasons can also explain why venture survival and investment are related to wealth or internal cash flow. And even if credit rationing exists, it might be efficient, in which case public intervention would be unnecessary and potentially damaging. Although there is some evidence of Type I credit rationing (e.g. Parker and van Praag, 2006a), evidence of Type II credit rationing remains elusive. Periods of credit crunch aside, modern capitalist economies are usually awash with credit to fund entrepreneurs with good ideas, which is not to deny that credit might be withheld from aspiring entrepreneurs with flawed or poorly presented business proposals. Future research should focus on potentially more fruitful lines of enquiry, such as discrimination in entrepreneurial credit markets; the impact of enterprise education on access to finance and subsequent entrepreneurial performance; limited financial literacy among entrepreneurs; and sophisticated modelling of processes of entrepreneurial accumulation and reinvestment. 9.10 Appendices 9.10.1

Preferences for skewness and entrepreneurship Denote wealth by B and suppose returns are given by the random variable ϑ, with density function g(ϑ), whose first three central moments are denoted by µ1 (mean), µ2 (variance) and µ3 (skewness). Then a Taylor expansion gives

U (B + ϑ) =

U (B) U
(B) U

(B) 2 U


(B) 3 + ϑ+ ϑ + ϑ + ... . 0! 1! 2! 3!

So taking expectations and truncating to the third moment yields
∞ E[U (B + ϑ)] = U (B + ϑ)g(ϑ)d ϑ −∞

≈ U (B) + U
(B)µ1 +

U

(B) U


(B) µ2 + µ3 . 2 6

(9.5)

286

Financing

U (B) can be treated as a constant for present purposes. If decreasing absolute risk aversion (DARA) holds, then U

(B) < 0 < U
(B). But also, crucially, since absolute risk aversion is defined as −U

/U
, DARA implies   d −U

−U
U


+ (U

)2 ≤ 0, = dB U
(U
)2 which holds if U


≥ (U

)2 /U
> 0. Thus U


> 0 under DARA. So by (9.5) agents like odd-numbered moments (mean returns and skewness) and dislike even ones (variance, or risk). Hence individuals might participate in risky entrepreneurship even if expected returns are negative (µ1 < 0, µ2 > 0), as long as they are compensated by a sufficiently large positive skewness of returns, µ3 > 0. A mechanism of this sort might have been observed in the case of the UK’s National Lottery, the popularity of which increased after the reward structure was redesigned to increase the value of the biggest cash prizes while decreasing the values of the smaller cash prizes. 9.10.2

The Paulson et al. (2006) structural model Recall from chapter 3 that the log-likelihood function of the standard probit model is

ln L =

n 

 

zi ln (β
Wi /σ ) + (1 − zi ) 1 − (β
Wi /σ ) .

i=1

˘ denote the probability that an individual is an More generally, one can let Pr(Wi |β) ˘ Then the entrepreneur, which depends on the variables Wi via some parameter set β. likelihood function is ln L =

  ˘ + (1 − zi ) 1 − Pr(Wi |β) ˘ . zi ln Pr(Wi |β)

n  

(9.6)

i=1

The conventional reduced form approach arbitrarily sets Pr =  and assumes a linear ˘ = β
Wi /σ . specification (Wi |β) In contrast, Paulson et al. (2006) pursued a structural approach. The idea here is that a lender specifies for every potential borrower a contract stipulating allocations of consumption, effort, capital and entrepreneurial output, to be received by the borrower:  = (ζ , e, k, q). Lenders observe individuals’ schooling, schi and assets Bi , which conditions their ability xi as described by (9.9) below. Lenders are competitive and break even in equilibrium. Lenders also know the following functional forms for utility, technology and ability: U (ζ , e) =

ζ 1−γ1 eγ2 −κ , 1 − γ1 γ2

κ, γ1 γ2 > 0

Pr(q = x|e, k > 0) = k α e1−α /(1 + k α e1−α ),

00)≥0



subject to (i) lenders breaking even, (ii) the above functional forms, and (iii) adding up constraints on the probabilities. The key part of Paulson et al.’s (2006) approach is to specify additional constraints that reflect either the EJ or the AB representation of borrowing constraints. As noted above, the EJ constraint imposes on  the additional requirement that k ≤ γ B. In contrast, the AB constraint requires that e satisfies an incentive compatibility constraint such that borrowers supply the recommended effort e rather than any alternative effort. Denote the solution to the maximisation of (9.10) subject to the particular set of ˘ in (9.6) above are applicable constraints by Pr∗ (|x, B, sch; k > 0). Then the Pr(Wi |β) calculated as
∞  ∗  ˘ = Pr(Wi |β) Pr (|x, B, sch; k > 0) d φ(). (9.11) −∞

The final step is to use ML to estimate the β˘ that maximises (9.6) subject to (9.11). Evidently, imposing the EJ and AB constraints will yield different solutions of Pr∗ (|x, B, sch; k > 0) and hence different structural probabilities – and therefore different maximised values of ln L. Crucially, likelihood ratio tests can then determine statistically which of the AB or EJ constraints gains most support from the data. For example, if the likelihood suggests that it is admissible to set κ = 0 in (9.7), then moral hazard cannot be a problem; whereas EJ requires γ > 1 to be binding. Paulson et al.’s (2006) results, based on a sample of Thai data from a socioeconomic survey fielded in 1997 to 2,880 households, were as follows. Paulson et al. (2006) rejected the hypothesis that EJ constraints alone can explain the data, and found statistical evidence instead that the dominant source of financial constraints is moral hazard. These authors also obtained reduced form evidence that is consistent with AB but not EJ. AB predicts that greater wealth is associated with less borrowing and more self-finance; EJ predicts the opposite. For wealthier Thai households who declared themselves credit-constrained, net saving is higher – consistent with AB but contradicting EJ. Taken together, these findings cast doubt on the EJ conceptualisation of borrowing constraints, influential though it has been in the literature. Notes 1. See, respectively, Gentry and Hubbard (2004), Quadrini (2000), Cagetti and de Nardi (2006) and Heaton and Lucas (2000).

288

Financing

2. Nor is the predominance of entrepreneurs at the top of the wealth distribution a new phenomenon.According to Moehling and Steckel (2004), the self-employed in nineteenth-century Massachusetts also held and accumulated a disproportionate share of wealth. 3. Bradford (2003) shows that both black and white entrepreneurs enjoy greater upwards mobility and lower downward mobility in the US wealth distribution than wage-and-salary workers. 4. See Shorrocks (1988), Banerjee and Newman (1993) and Parker (2000). 5. Moskowitz and Vissing-Jørgenson (2002), Gentry and Hubbard (2004) and Bitler et al. (2005). 6. To see why, note that the optimal demand for capital k ∗ maximises xk α − (1 + r)k, yielding k ∗ = [αx/(1 + r)]1/(1−α) – which is increasing in x. 7. To obtain these results, note that bank profits are returned to depositors who receive the rate of return ρ > 0. Hence   D(B) p(B)D(B) = ρ = αD(B) 1 − (1 − B) s . R Solving for D(B), one obtains  D(B) = 1 −

8.

9. 10. 11. 12.

13.

14. 15.

16. 17.

18.

!

4ρ(1 − B) 1− αRs



Rs . 2(1 − B)

(9.12)

Now (9.12) implies (i) D
(B) < 0 so given p(B) = ρ/D(B) we have p
(B) > 0, i.e. effort is increasing in wealth; and (ii) all agents with wealth B < 1 − (αRs )/4ρ will be unable to borrow. For cross-section and panel studies, see Evans and Leighton (1989b), Bernhardt (1994), Bates (1985, 1997), Laferrère and McEntee (1995), van Praag and van Ophem (1995), Taylor (1996), Boden (1999b), Fairlie (1999), Quadrini (1999), Bruce et al. (2000), Dunn and Holtz-Eakin (2000), Johansson (2000), Fairlie and Meyer (2003), Gentry and Hubbard (2004), Henley (2004), Paulson and Townsend (2004) and Nykvist (2008). For time-series studies, see Robson (1991, 1996, 1998a, 1998b), Black et al. (1996) and Cowling and Mitchell (1997). See Wong (1986), Henley (2004) and Shane (2003, p. 152). See Taylor (2001), Uusitalo (2001) and Hurst and Lusardi (2004). See Lindh and Ohlsson (1996), Blanchflower and Oswald (1998) and Yang (2008). See Holtz-Eakin et al. (1994a, 1994b), Lindh and Ohlsson (1996), Blanchflower and Oswald (1998), Taylor (2001), Georgellis et al. (2005) and Yang (2008). According to Georgellis et al. (2005), redundancy and inheritance windfalls have a positive effect on transitions, whereas lottery wins have the reverse effect. Hurst and Lusardi were not the very first, however. To the best of my knowledge, that honour goes to Magnac and Robin (1996), in a study which used French Financial Assets Survey data. Although their instruments lack a convincing rationale, Magnac and Robin (1996) reported that instrumented wealth entered the self-employment specification positively. See Cagetti and de Nardi (2006) for a calibrated multi-period model which provides a theoretical underpinning for this empirical finding. A problem with this explanation, however, is that workers can in fact choose from an array of work hours in paid employment; they do not have to become entrepreneurs to cut their work hours. And the evidence reviewed in chapter 12 shows that entrepreneurs tend to work longer, not shorter, working weeks on average. Winborg and Landström (2000) discuss some bootstrapping strategies. ‘The fact that some unlucky agents suffer the penalty [of borrowing constraints] may, from a social point of view, be less important than the fact that the penalty motivates agents to succeed more often’ (Ghatak et al., 2001, p. 798). See Oi (1983), Brown and Medoff (1989), Oi and Idson (1999) and Garmaise (2008).

Wealth and entrepreneurship

289

19. See Bates (1990), Black et al. (1996), Taylor (1999), Quadrini (1999) and Bruce et al. (2000). In contrast, Taylor (2001) found no effect of inheritances on UK self-employment survival probabilities. 20. See Hubbard (1998) for a review. Some recent Russian evidence from Hartarska and GonzalesVega (2006) also suggests that firms with cash reserves are significantly more likely to invest than firms which lack them. 21. Indeed, UK evidence shows that entrepreneurs in more technologically intensive industries are more likely to claim they are capital-constrained – though whether these claims are justified is not investigated (Westhead and Storey, 1997). See also Moore (1994) and Oakey (1984) for similar evidence. 22. See Cosh and Hughes (1994), Moore (1994), Blanchflower and Oswald (1998), Guiso (1998), Bratkowski et al. (2000), Blanchflower et al. (2001) and Cagetti and di Nardi (2006). 23. Asking bank loan officers whether they respond to adverse selection or moral hazard is probably superior to asking surviving firms whether they think loan officers take these issues into account when making loan decisions (see Hyytinen and Väänänen, 2006, for an example of the latter approach). 24. De Meza and Webb (2006) provide an alternative interpretation of the observation that borrowers often choose not to return to the credit market following rejection of their loan applications. When unfunded projects decay rapidly, some entrepreneurs who failed to get credit might never reapply because the value of their projects is diminished. This reason is distinct from the entrepreneurs being discouraged.

Part III Performance

10 Entrepreneurship, job creation and innovation

This chapter discusses the role of entrepreneurship in creating new jobs and innovating new products. Both of these topics have their origins in research conducted in the late 1970s and the 1980s, when David Birch first claimed that small new firms acted as the engine of job creation in the economy, and David Audretsch and Zoltan Acs argued that small firms played a disproportionate role in the commercialisation of new innovations. Both topics command widespread interest because they suggest that entrepreneurship directly drives venture performance and economic growth. Much (though not all) of the empirical discussion in this chapter is framed in terms of comparisons between small and large firms. The focus on firm size – which, as noted in chapter 1, does not obviously capture the essence of entrepreneurship – is chiefly a historical legacy. It also reflects data availability. In the discussion that follows, ‘small business’ will merely be assumed to serve as a convenient shorthand for entrepreneurship. However, this focus will be complemented with a discussion about the role of individual entrepreneurs in job creation and innovation. After setting out in the first section some basic facts about entrepreneurs’ decisions to hire external labour, I will present some theory about the labour demand of individual entrepreneurs.1 This paves the way for an analysis of the empirical factors which determine job creation by entrepreneurs. The second section discusses the role of small firms in creating jobs in the broader economy. The third section turns its attention to the role of small firms in aggregate innovation activity. The final section concludes.

10.1

Job creators

10.1.1

Some basic facts about entrepreneurs’ labour demand Despite the attention policy-makers devote to the employment-creation potential of entrepreneurship, relatively few entrepreneurs create any jobs. Measuring entrepreneurship as self-employment, only 20–30 per cent of entrepreneurs in the USA, Canada and the UK employ any external workers. This ‘employer self-employed’ rate seems if anything to have declined in these countries in the 1990s.2 According to Bregger (1996), of the 21 per cent of self-employed Americans who hired any employees in 1995, one-third 293

294

Performance

had only one employee, and only one-seventh hired six or more workers. Using BHPS data over 1991–99, Henley (2005) reported that only 6–7 per cent of self-employed Britons create jobs for ten or more people. Transition rates into employer status also appear to be low. Both Carroll et al. (2000) for the USA and Cowling et al. (2004) for the UK observe that only about 7–8 per cent of self-employed sole proprietors hire any other workers over a three- to fouryear period. But sole proprietors are nevertheless more likely to employ other workers than any other labour market group, including new entrants to self-employment from unemployment or paid employment (Cowling et al., 2004). A higher proportion of self-employed people employ others in some of the countries of Continental Europe, reaching 46 per cent in Denmark and 51 per cent in Germany (Cowling, 2003). This might be because these countries lack the low-value sole-trader self-employed workforce which shows up in the left-hand tail of the self-employment income distribution in countries like the UK and the USA (McManus, 2000). According to Estrin et al. (2006), less than one-quarter of self-employees in the transition economies of Poland and Russia are employers, although this fraction reaches about one-half in Hungary and the Czech Republic. Clearly there is considerable heterogeneity in rates of hiring by entrepreneurs across countries.

10.1.2

Theoretical analysis

The theoretical literature on labour demand by entrepreneurs is sparse. To the best of my knowledge, only Jefferson (1997) and Carroll et al. (2000a) have systematically analysed the labour demand choices of entrepreneurs in a theoretical framework. Jefferson (1997) proposed a rather unusual credit rationing model, in which individuals become entrepreneurs if they are non-rationed, but can only work for entrepreneurs or draw unemployment benefit if they are rationed. Rationed workers are unable to verify the quality of projects run by entrepreneurs. Entrepreneurs can always claim to have poor projects, so they can always offer a minimal wage to employees. As a result, in equilibrium rationed individuals will prefer to become unemployed rather than work for entrepreneurs. While this model claims to explain why only a small minority of entrepreneurs hire outside labour, it is clearly unrealistic in several important respects. Carroll et al. (2000a) proposed a different model which distinguishes between two margins: entrepreneurs choose whether to become employers, and if so how many workers to hire. The model is sketched out in the chapter appendix. Carroll et al.’s (2000a) main assumption is that employees are costly to hire, but the entrepreneur’s effort is more productive in ventures where employees are hired. The main result is that a cut in the marginal income tax rate (MTR) has ambiguous effects on an entrepreneur’s decision about whether to become an employer; but if the utility function is sufficiently concave in consumption, cuts in the MTR will have a positive overall effect on the probability of hiring. The latter is precisely what Carroll et al. (2000a) found empirically using US IRS data from 1985 and 1988 – years which span the ‘tax cutting’ Tax Reform Act (TRA) of 1986. Carroll et al.

Entrepreneurship, job creation and innovation

295

(2000a) applied data on sole-trader and employer entrepreneurs to the probit model given at the end of the chapter appendix, and estimated that a 10 per cent cut in an entrepreneur’s MTR increases the mean probability they hire workers by about 12 per cent. The implied elasticity of 1.2 suggests that general income tax reductions might be a powerful way of stimulating employment creation – in marked contrast to its apparently limited effects on work hours (chapter 12) or participation in entrepreneurship (chapter 17). An alternative theoretical approach to Carroll et al.’s (2000a) might estimate a labour demand function directly. That can be done if one utilises some simple functional forms for the production and utility functions (Hamermesh, 1993). For example, van Praag and Cramer (2001) and Cowling et al. (2004) both assume a Cobb–Douglas production function which is scaled up by human capital, from which it follows directly that ‘job creators’ are predicted to have greater human capital on average than self-employed sole traders.3 Two limitations of the existing state of theoretical knowledge are over-reliance on the Cobb–Douglas production function (which is restrictive in several important respects4 ), and limited analysis of the number of employees hired by job creators. Data limitations meant that Carroll et al. (2000a) could only estimate the responsiveness of the wage bill, wl, to changes in the MTR.5 Carroll et al. (2000a) estimated that the median wage bill is significantly negatively related to the MTR, with an estimated elasticity (with respect to one minus the MTR) of 0.37.

10.1.3

Evidence Empirical research on the characteristics of job creators is rather limited in scope. Table 10.1 summarises research findings from several dedicated analyses of this issue. Possibly reflecting the small number of studies and the relatively homogeneous nature of the samples used, there is strong agreement about the effects of various factors on the propensity of entrepreneurs to hire outside labour. Human capital, in the form of education, age and experience, all have uniformly positive effects. One source of disagreement is between the two BHPS studies of Cowling et al. (2004) and Henley (2005). Whereas the former found that academic qualifications were not important determinants of job-creator status (in contrast to vocational qualifications), the latter reported that entrepreneurs with university degrees were significantly more likely to be job creators. Henley (2005) used both a superior data sample and an econometric model which fully exploits the longitudinal nature of the data, so his results should probably be preferred. Also noteworthy is that van Praag and Cramer (2001) qualified a positive impact from education by recording negative marginal effects from having an arts degree. Burke et al. (2002) reported a significant positive effect from professional training on job-creator status. Cowling et al. (2004) observed that the ranks of job creators are disproportionately filled by men; this gender difference is confirmed in other studies. Women entrepreneurs are most likely to hire workers if they run their business full-time (Burke et al., 2002).

296

Performance

Table 10.1. Characteristics of job creators Variable

Effect

Evidence

Education

+

Age Experience Male

+ + +

Parent self-employed or manager Wealth

+ + +

Industry conditions

+

Burke et al. (2000, 2002); van Praag and Cramer (2001); Cowling et al. (2004); Henley (2005); Fairlie and Robb (2007a) Mata (1996); Cowling et al. (2004); Henley (2005) Burke et al. (2002) Van Praag and Cramer (2001); Cowling et al. (2004); Henley (2005) Van Praag and Cramer (2001); Cowling et al. (2004); Henley (2005) Burke et al. (2002); Cowling et al. (2004); Henley (2005); Fairlie and Robb (2007a) See text

Notes: All are UK studies apart from van Praag and Cramer (Netherlands), Mata (Portugal) and Fairlie and Robb (USA). All studies define entrepreneurship in terms of self-employment or business ownership, and estimate a binary choice model (see chapter 3).

The greater incidence of part-time working among women business owners might therefore explain part of the gender gap in job-creator status. Having a parent who was self-employed or a manager also increases the likelihood of job-creator status. According to Henley’s (2005) estimates, a self-employed person is over one-third more likely to employ one or two employees if they had a parent who employed others. And they are three-quarters more likely to employ ten or more employees than someone without a parent who employed others. Wealth and inheritances are also associated with job-creator status, although these variables are likely to be endogenous (see chapter 9). According to Curran and Burrows (1989), male employer self-employees in Britain are more likely to own their home (89.3 per cent) than male own-account self-employees (75.6 per cent), who in turn are more likely to own their home than male employees (68.2 per cent). Finally, industry conditions and the state of the local economy can also be important factors. For example, using a sample of data on 766 Portuguese firms, Mata (1996) reported that the log employment size of new firms was significantly larger in more turbulent industries and in industries with greater minimum efficient scale. More generally, job creators are most commonly found in professional and managerial occupations or in the distribution, hotels and restaurant sector. They are significantly less likely to be found in the construction industry (Curran and Burrows, 1989). Henley (2005) reported that local unemployment rates are significantly positively associated with self-employed job creation: this suggests that employers might exploit slacker labour markets by hiring more workers – consistent with the neoclassical theory of labour demand (Hamermesh, 1993).

Entrepreneurship, job creation and innovation

10.2

297

Job creation by small firms

The publication of the influential Birch Report in the USA (Birch, 1979) stimulated a wave of research on the job-creation performance of small firms. While entrepreneurs rather than small firms are the primary focus of this book, many small firms are owned and managed by entrepreneurs, so it is appropriate to review this literature. Below, I will discuss evidence about the relative job-creation performance of small firms as a group, rather than on the determinants of employment growth in representative samples of small firms. The latter topic will be treated separately in chapter 11. Birch (1979) claimed that, between 1969 and 1976, small firms employing fewer than twenty workers generated 66 per cent of all new US jobs, and firms with fewer than a hundred employees accounted for 82 per cent of net job gains. The implication is that the small-firm sector is the primary engine of job creation in the economy. Subsequent researchers have confirmed these findings for the USA and other countries.6 Acs and Audretsch (1993) highlighted a distinct and consistent shift away from employment in large firms and towards small enterprises in the 1980s in every major Western economy. Others have challenged these claims, however. An early rebuff came from Armington and Odle (1982), whose study of employment changes over 1978–80 revealed much more modest small-business job-creation rates than Birch (1979). But as Kirchhoff and Greene (1998) pointed out, Armington and Odle’s findings might have merely reflected the unusually subdued macroeconomic conditions prevailing between 1978 and 1980. Weightier criticisms of Birch’s thesis came in the late 1980s and the 1990s. In particular, Davis and Haltiwanger (1992) and Davis et al. (1996a, 1996b) argued that the ‘conventional wisdom about the job-creating powers of small businesses rests on statistical fallacies and misleading interpretations of the data’ (1996a, p. 57). These were said to include: 1. The ‘size distribution fallacy’, whereby static size distributions of numbers employed by firm size are used to draw inferences about dynamic changes in employment shares. Measures such as ‘the net change in employment by small firms as a proportion of the net change in total employment’ are biased when firms move between size categories. They can also be misleading if the denominator of this ratio is very small over a particular period, exaggerating the scale of employment growth by small firms. 2. The ‘regression fallacy’, whereby transitory size shocks bias the relationship between employment growth and firm size. Transitory shocks imply that businesses which have recently contracted will be measured as small and will be observed to grow subsequently, while businesses that have recently expanded will be measured as large and will contract subsequently.7 3. Poor quality US micro-data, distorting the true employment creation–firm size relationship (see the previous point).

298

Performance

4. Confusion between net and gross job creation, since small firms destroy many jobs as well as creating them. 5. Sample selection bias, whereby small firms with poor job creation records die and leave the sample, biasing upwards measured small firm job creation rates. 6. Disguised job creation: some ‘small firm’ job creation may actually be done by large firms in disguise, as incumbents often expand by creating additional small establishments (Armington and Odle, 1982). As Hamermesh puts it: ‘The locus of expansion is the small plant; but the locus of decision-making is the average-sized firm’ (1993, p. 157). On the other hand, job-creation performance by small firms may be understated if growing firms undergo changes of ownership, such as acquisition, which subtract from employment growth attributable to small firms by recording a false death, and add to employment growth attributable to large firms by recording a false birth (Kirchhoff and Greene, 1998). Large firms often acquire small ones, while in counterpoint to point 1 above, small firms which are successful grow into large ones, adding to apparent job creation by ‘large’ firms. Another issue which muddies the waters is subcontracting. For example, some firms subcontract the manufacture and distribution of their products, which can create additional indirect employment that goes unrecorded – or is attributed to other firms which gain the manufacturing and distribution contracts. The sizes of the contractors’ businesses determine whether the new jobs are attributed to ‘small’ or ‘large’ firms. While the researcher interested in gauging the role of small firms in job creation can rarely do much about the quality of available data, they should always strive to fix the other problems listed above. Most researchers now eschew simplistic static size distribution comparisons, and measure net rather than gross job-creation rates. Many also circumvent the regression fallacy by measuring a firm’s growth rate relative to a sample period average size rather than relative to initial size. Recognising these points, and using a panel of data on US manufacturing plants over 1972–88, Davis et al. (1996a, 1996b) claimed, in contrast to Birch, that larger plants and firms create (and destroy) most manufacturing jobs. In addition, Davis et al. found no clear relationship between rates of net job creation and employer size. However, despite some corroborative evidence from Wagner (1995), most other researchers who have made similar corrections have re-asserted a negative cross-section relationship between firm size and net employment creation.8 Some of the disagreement between these findings may be attributable to the exclusion by Davis et al. (1996a, 1996a) and Wagner (1995) of service-sector firms from their samples. This point is especially pertinent because most new jobs created in the USA in the 1990s were in the service sector. Thus Bednarzik (2000) showed that small ventures employing fewer than twenty employees played a much greater role in job creation in services in the USA than they did in manufacturing. In short, although the precise scale of the contribution by small firms to employment creation is still disputed, the OECD (1998) could justly claim that there is now ‘general agreement’ that the share of jobs accounted for by small firms has steadily increased since the early 1970s in most developed economies.

Entrepreneurship, job creation and innovation

299

In many ways, though, the debate has now moved on. There is declining interest in the issue of job creation by small firms, and growing interest in the issue of job creation by new firms. New firms are widely believed to better represent the notion of entrepreneurship than small firms (and there are of course no ‘regression to the mean’ problems when firms are categorised by age). In contrast to the contentious evidence relating to size, that relating to age is rather clear-cut. For example, using US Census Bureau Longitudinal Business Database (LBD) data, Haltiwanger (2006) showed that net employment creation rates decrease monotonically in business age; and few systematic relationships between employment creation and firm size emerge after controlling for business age. About 60 per cent of the job creation from young firms comes from entry and about one-half of their job destruction comes from exit. The corresponding rates for mature firms are about one-half of these. Acs and Armington (2006) provide corroborating evidence: establishments less than two years old account for almost 100 per cent of net job growth, while all other age classes of establishments shed employment on average. In the words of Haltiwanger and Krizan (1999, p. 94): ‘for employment growth, it looks as though the more important factor is age and not size. Put differently, most small establishments are new. Thus, the role of small business in job creation may simply reflect the role of births and in turn young establishments.’ A word of caution is in order, however. The above findings might be a particularly American phenomenon. More modest effects on job creation from new firms have been observed in Europe, where new firm formation rates are lower on average (Bednarzik, 2000). Even if new and small firms create more jobs than large incumbents, one can question the quality of those jobs. Evidence consistently shows that the jobs created by new and small ventures are of lower average quality than those created by larger incumbents. Small firms in particular tend to employ higher proportions of part-time workers, freelancers and home workers. Their employees tend to be less educated on average, receive lower levels of training, and work longer hours with a greater risk of accidents and major injuries (Storey, 1994a; Hasle and Limborg, 2006). At the same time they enjoy lower job tenure than their counterparts in larger firms (Brown et al., 1990). One reason for the last finding is no doubt that new and small enterprises have higher closure rates than their older and larger counterparts (see chapter 14). Small firms also pay lower wages and fringe benefits (e.g. health insurance, maternity leave and retirement benefits) than large firms do.9 Brown et al. (1990) reported that workers in large companies earn over 30 per cent more on average than comparable workers in small firms. This finding appears to hold across industries and countries. It is consistent with simple efficiency wage models, although an alternative explanation is that small firms are simply less efficient in terms of total factor productivity (Alvarez and Crespi, 2003). In fact, it turns out to be difficult to explain the large-firm wage premium empirically using ‘standard’ explanatory variables (Troske, 1999). Nevertheless, the size–wage premium is an important phenomenon in the present context. When Baldwin (1998) adjusted raw measures of employment to take account of higher wages in larger

300

Performance

firms, he found that small manufacturers in Canada no longer outperformed their large counterparts in terms of growth. New firms also pay lower wages than otherwise comparable incumbents, including those of a similar size (Brown and Medoff, 2003). Brixy et al. (2007) estimated the wage gap between new firms and incumbents to be 8 per cent in Germany over 1997– 2001, though the gap diminished as the surviving new firms matured. Troske (1999) had greater success explaining the incumbent–wage premium (over young firms) than the size–wage premium. The key explanatory variables were the experience, tenure, education and demographic characteristics of the employees, as well as occupation type. The ‘new firm wage discount’ raises the interesting question of how new venture founders can attract talented employees away from incumbents, where they are better remunerated. Stock options which entitle employees to a share of future profits are one possible explanation (Caselli and Gennaioli, 2005). The value of these options does not usually show up in current wages in new start-ups. To set this last point in context, it should be noted that entrepreneurs use productivityrelated pay schemes less frequently than owners of large firms do (Cowling, 2001). Perhaps a more important attractor is the higher job satisfaction enjoyed by employees of small firms compared with large ones, at least in Europe (Clark and Oswald, 1996; Benz and Frey, 2004, 2008). As noted in chapter 4, a plausible reason for higher job satisfaction in small firms is that employees in small firms are less subordinated to complex hierarchies than employees in large firms. From a policy perspective, the objection that new firms create lower-quality jobs than large ones do should not be overdone. Policy-makers cannot simply choose between creating high-quality jobs in large incumbent firms on one hand and low-quality jobs in new ventures on the other. If anything, policy-makers face instead a choice between promoting the creation of new jobs in small firms (some of which may eventually become large employers), and promoting no job creation at all. This realisation no doubt underpins public policy efforts to encourage entrepreneurship and the creation of new firms, the above facts about differential job quality notwithstanding. I will return to the issue of public policy towards entrepreneurship in the closing chapters of this book. 10.3

Innovation by small firms

Innovation, deemed by Schumpeter to be a central aspect of entrepreneurship, is another major reason for policy interest in entrepreneurship. It is known that industries with high rates of entry by small firms tend to enjoy high rates of productivity growth and innovation (Geroski and Pomroy, 1990; Cosh et al., 1999). Furthermore, more innovative new entrants tend to enjoy superior post-entry performance (Vivarelli and Audretsch, 1998; Arrighetti and Vivarelli, 1999). Josef Schumpeter believed that large incumbents would ultimately come to dominate the innovation process by exploiting their economies of scale. Schumpeter predicted that the vast majority of R&D and innovations would eventually come to be conducted

Entrepreneurship, job creation and innovation

301

Table 10.2. Innovation and entrepreneurship Large firm innovation advantages

Small firm innovation advantages

1. Scale economies - spread high fixed R&D costs over larger output - generate economies of scope - free limited managerial attention 2. Innovation is less risky - size and market power - diversification of product lines

1. Diseconomies of scale - less bureaucracy in small firms - shorter lines of communication - greater agility and lower agency costs

3. Larger firms can afford to spend more on R&D

2. Greater incentives to innovate - easier to incentivise agents in small firms - easier to overcome entry barriers and competition 3. Diminishing marginal returns to R&D; entrepreneurs can exploit knowledge spillovers

4. ‘Efficiency effect’ favours large incumbents

4. ‘Replacement effect’ favours small entrants

by large firms, while small firms would merely become the repositories of low-level imitation: ‘relics of a bygone age’. Subsequently this view has been challenged on both theoretical and empirical grounds. Both theory and evidence are now considered in turn. 10.3.1 Theoretical arguments The first column of Table 10.2 lists several reasons why large firms might possess advantages at innovation over small firms. The second column lists some counterarguments. As Table 10.2 shows, there are four major areas of theoretical disagreement about the role small firms are likely to play in innovative activity. The first major difference relates to economies of scale, the basis of Schumpeter’s predictions mentioned above. A modern articulation of this view is found in Klepper (1996), in which early innovators rapidly achieve scale, providing greater incentives to implement incremental process innovations than their smaller rivals. Klepper’s key insight is that large firms can spread the costs of developing new innovations over a massive scale, so that innovations which yield cost reductions of a given percentage rate yield greater absolute profit margins in larger firms. These forces enable large firms to reduce costs and expand scale even more. This self-reinforcing process forces out smaller rivals and chokes off entry, eventually culminating in highly concentrated, oligopolistic market structures. Hence innovation interacts with scale to entrench large firm competitive advantage. Scale economies in production might also provide ‘economies of scope’, thereby increasing profits from innovation (Acs and Audretsch, 2003). And senior managers of large firms can more easily delegate operational tasks to more junior managers, freeing up precious managerial attention needed to identify new ideas and innovations (Gifford, 1998).

302

Performance

On the other hand, large firms can be prone to diseconomies of scale, potentially making innovation more difficult. Large firms can suffer from bureaucratic inertia, which is antithetical to innovation (Link and Rees, 1990; Freeman and Engel, 2007). Indeed, it has been claimed that incumbents are not only relatively poor at pioneering radical innovations, but also struggle to develop incremental ones (Henderson and Clark, 1990). It could be that managers in existing firms fail to spot new opportunities because they follow internal routines which blind them to new trends in the market. If they act quickly, entrepreneurs located outside large firms might therefore be well placed to develop innovative ideas untrammelled by conventional corporate thinking (Pavitt et al., 1987; Freeman and Engel, 2007). There also tend to be shorter lines of communication in small firms (Fielden et al., 2000). This facilitates easier sharing of information, by promoting tighter linkages between work colleagues and fostering greater trust. There can also be greater ease of technology diffusion between networks of small firms, especially those involved in clusters which generate opportunities for cross-organisational learning (Morgan, 1997). And new, small ventures might be more agile and responsive to opportunities created by changing demand and demographic patterns (Bannock, 1981). The second set of entries in Table 10.2 recognises that incentives to innovate are also likely to differ by firm size. Large diversified firms can not only use existing marketing channels to sell innovative products to numerous customers quickly and easily, but can also spread the risks of innovation, giving them greater incentives to develop radical innovations. On the other hand, large bureaucratic firms might find it costly to overcome agency problems and to provide the high-powered incentives required to motivate their workers to develop radical innovations (Holmstrom, 1989). Facing sufficiently high costs to incentivise effort in these cases, large firms may optimally pass over uncertain new technology development in favour of more routine ones, despite the more modest average returns of the latter (Bhide, 2000). Furthermore, incremental changes in corporations most easily satisfy objective external processes of scrutiny. And small firms may have to innovate in order to overcome entry barriers and to cope with retaliatory conduct by incumbents (Acs and Audretsch, 1989). This argument highlights important strategic and industry dimensions of small firm innovation activity. The third row in Table 10.2 draws on evidence that larger firms tend to perform more R&D (Cohen and Klepper, 1992). Larger firms usually have easier access to finance, through re-invested profits and bank loans, to fund expensive innovation. Yet there can be diminishing returns to R&D, which weigh on large firms more than small ones. That attenuates the large firm R&D advantage (Acs and Audretsch, 1991). Small firms can also compensate for limited direct R&D spending by exploiting knowledge spillovers which leak out of larger organisations (see below). Finally, Table 10.2 contrasts the so-called ‘efficiency’ effect with the ‘replacement’ effect. The efficiency effect recognises that incumbents have greater incentives to innovate and retain monopoly profits than new entrants. The latter can at best obtain duopoly profits by exploiting an innovation opportunity in the same market (Gilbert and Newbery, 1982). The urge to preserve market share driven by monopoly profit

Entrepreneurship, job creation and innovation

303

incentives might explain why corporate giants in the IT world like Intel, Microsoft and Cisco continually innovate, sometimes radically. The replacement effect, on the other hand, recognises that large incumbents might be unwilling to destroy their monopoly rents by engaging in paradigm-shifting innovations which render existing products obsolete (Arrow, 1962). That is, by definition, not a problem for new entrants. If the efficiency effect is strong enough, monopoly positions in product markets will tend to last, which might explain persistent innovation by incumbents in some industries, such as IT products (Klepper, 1996; Klepper and Simons, 2000a). But in sectors open to drastic innovations, the replacement effect is more likely to dominate (Reinganum, 1983), entailing a strong tendency to entry.

10.3.2

Evidence about innovation

The evidence relating to the role of large versus small firms, and radical versus incremental innovation, speaks to two issues. One concerns whether small firms are more or less innovative than large firms, and the other concerns whether large firms are responsible for a disproportionate number of incremental innovations while small firms are associated with more radical ones. An important preliminary question is how to measure innovation. Researchers typically utilise one or more of the following measures: R&D, the number of patents, the impact of innovations, and the commercialisation of inventions. The first two measures are skewed in favour of large firms and suffer from numerous methodological limitations (Acs and Audretsch, 2003). Among these, R&D is an input rather than an output measure; and many patents never lead to useful innovations in practice. The last two measures are perhaps the most informative, although it is important to use objective data rather than subjective self-assessments. Aggregate cross-industry evidence paints a mixed picture about the contributions of small and large firms to aggregate levels of innovative activity. Large firms certainly perform the bulk of R&D spending and patenting activity,10 though consistent with the right-hand side entry of row 3, Table 10.2, small firms enjoy greater marginal growth benefits than large firms from additional R&D spending (van Praag and Versloot, 2007). As noted above, though, R&D is an input, not an output. In terms of innovative output, smaller and younger firms appear to enjoy pronounced advantages over their larger and older counterparts, at least in some industries.11 For example, according to Scherer (1991), in the 1980s ‘small’ US firms (defined as those with fewer than 500 employees) created 322 innovations per million employees compared with 225 per million in large firms. In a similar vein, Acs and Audretsch (1990, chap. 2) reviewed four databases measuring technological innovation on the basis of their peer-reviewed ‘importance’. These authors estimated that small firms contribute around 2.4 times as many innovations per employee as large firms. The SBA (2003) estimates that small firms represent one-third of the most prolific patenting companies that have registered fifteen or more US patents. Small firms’ patents are twice as closely linked to scientific research as those of large firms, being more ‘high-tech and leading-edge’. Moreover,

304

Performance

their patents are on average more highly cited than those of large firms. In particular, new ventures hold more highly cited patents in the biotechnology industry than incumbent firms, in both the USA and France (Gittelman, 2006). Reflecting these findings, many famous radical innovations originated in small rather than large firms. Examples include the airplane, FM radio, the zipper and the personal computer, among many others (Baumol, 2007). This evidence supports the contention that small firms specialise in radical innovations while large firms focus on incremental ones.12 Further evidence of this kind comes from Prusa and Schmitz’s (1991) analysis of the software industry, in which new firms provided the majority of ‘category opening’ products in the 1980s, developing six times as many of these products in absolute terms as large firms. Prusa and Schmitz (1991) observed that existing firms had a comparative advantage in incremental improvements within existing product categories, consistent with the ‘replacement effect’ argument that incumbents avoid introducing competencedestroying technologies.13 Further buttressing this point, Audretsch (2003) observed greater displacement of incumbents by entrants in innovative industries, suggesting that new firms have a ‘more pronounced Schumpeterian creative destruction’ innovation impact compared with incumbents. On the other hand, large firms have produced many path-breaking innovations too (Acs and Gifford, 1996). A study of consumer durables and office products in the US concluded that, in these sectors at least, incumbents and large organisations introduced the majority of radical product innovations over the last sixty years of the previous century (Chandy and Tellis, 2000). Furthermore, King and Tucci (2002) chronicle how experienced incumbents successfully rode each new technological wave that hit the hard-disk-drive industry in the USA. Incumbents may not have been the first ones into the market, but their survival rates exceeded those of new entrants. Indeed, later entry by incumbents does not necessarily point to failure or irrationality (Berchicci and Tucci, 2006; Bayus and Agarwal, 2007). Baum et al. (1995) report that the shift from analogue to digital technology in the facsimile transmission business enhanced the competitiveness of incumbents and reduced the new firm formation rate. These findings are all broadly consistent with Klepper’s (1996) evolutionary perspective in which early innovators continue to grow by exploiting ever-increasing incentives to innovate, owing to ever-increasing economies of scale. The evidence from a range of US industries, including tyres, autos, penicillin and TVs, seems to confirm the innovative advantage of large, long-established incumbents. There is greater disagreement among researchers about whether start-ups are more or less likely than established firms to commercialise inventions generated by university departments. Mansfield (1991) claimed that small firms have an advantage over large firms in this regard, but Lowe and Ziedonis (2006, p. 180) detected few such differences between start-ups and established firms. Both sets of authors agree, however, that small start-ups bring commercial applications based on academic research quicker to market than large firms do. On the negative side, entrepreneurs continue to pursue unsuccessful commercialisations for longer than established firms, perhaps because of over-optimism, and thereby can end up destroying value (van Praag and Versloot, 2007).

Entrepreneurship, job creation and innovation

305

Building on work by Jaffe (1989), Acs, Audretsch and Feldman (1994) argue that new small ventures exploit knowledge which spills over from universities and large companies. Based on estimates derived from a simple econometric model,14 Acs, Audretsch and Feldman (1994) claim that knowledge spillovers are more decisive in promoting innovative activity of small firms than of large corporations. Other evidence is consistent with these findings.15 An analysis of Canadian biotech firms in the 1990s shows that entrants are attracted to incumbents’ R&D resources within a 500-metre radius, but not outside this radius (Aharonson et al., 2007). This suggests the existence either of agglomeration benefits (access to pools of labour or specialised inputs) or knowledge spillover externalities, which dissipate rapidly with distance. Aharonson et al. (2007) also observe that entry rates are significantly higher if technologically similar incumbents and firms with university alliances are situated close by. In their commanding survey of the evidence base on innovation, technological change and small firms, Acs and Audretsch (2003) conclude that economies bestowed through geographical proximity and spatial clusters might be more important for producing innovative output than ‘traditional’ scale economies, at least in some industries. In absolute terms, while it might be true that large firms have been the most important sources of innovations in the US economy, small firms have bucked the trend in several industries, including computers, process control instruments and biotechnology (Acs and Audretsch, 1990; Gittelman, 2006). That is, small firm innovative advantage in the USA tends to be in different industries to those where large firms have an innovative advantage (Acs and Audretsch, 1988; Prevenzer, 1997). As a generalisation, large firms have a comparative advantage in exploiting efficient internal ‘routinised regimes’ to develop innovations (including radical ones) in capital-intensive, concentrated and unionised industries which produce differentiated products (Acs and Audretsch, 1991). Precisely this outcome has been observed in manufacturing, for example (Tether et al., 1997; Craggs and Jones, 1998). In contrast, small firms have a comparative advantage in exploiting their ‘entrepreneurial regimes’ in industries where human capital and skilled labour are important productive factors.16 This nuanced picture does not provide unambiguous support for Schumpeter’s prediction of ever-increasing concentration of innovation in large firms. As always in discussions about the contributions made by entrepreneurs, it is important to retain a sense of perspective. At the level of the individual entrepreneur, most start-ups are in mundane non-innovative trades such as hairdressing and car mechanic businesses (Storey, 1994a). Real innovation appears to be confined to a small handful of businesses run by a few talented, visionary and determined entrepreneurs. 10.4

Conclusion

This chapter reviewed the literature relating to two important contributions made by entrepreneurs: job creation and innovation. Both topics have been extensively analysed through the lens of small versus large firms; rather less analysis has been performed at the level of the individual entrepreneur. More research is needed into the innovative

306

Performance

performance of entrepreneurial firms at the micro-level, especially research which is tied into the local and national economic and policy context. A good exemplar in this regard, which unifies data on institutions, individuals and the regional context, is Zucker et al. (1998). These authors found that although major universities and federal research support made important contributions to the sudden appearance and rapid growth of the US biotechnology industry, the growth and location of ‘intellectual human capital’ (embodied in star scientists) were the principal determinants of the growth and location of the industry. Interestingly, venture capital availability did not seem to play a driving role. With regard to the theme of the first part of this chapter, more theory and evidence is also needed about job creators and the factors which determine the number of workers they hire. Future research could usefully estimate econometric models of labour demand. These have been extensively analysed for samples of established firms, but scarcely at all for samples of entrepreneurial firms. Pursuit of this research agenda will require high-quality data on factor prices, factor quantities and production costs. This agenda is worth pursuing because of its relevance to several pressing public policy issues. For example, it would be desirable to understand how government labour market regulations bear on individual entrepreneurs’ hiring decisions. Some discussion of this question framed at the aggregate level, and at the interface between innovation, public policy and entrepreneurship, appears in the last two chapters of this book.

10.5 Appendix 10.5.1

Carroll et al.’s (2000a) model of labour demand and taxation Let ζh and ζn denote consumption and eh and en denote effort, where the h and n subscripts denote ‘hiring’ and ‘non-hiring’ status of an entrepreneur, respectively. Also, let π(w, eh ) denote the indirect profit function of an employer entrepreneur, where w is the market wage paid to employees. This function is decreasing in its first argument and increasing in its second argument. Also, let Ben denote the profits of non-hiring entrepreneurs, where B > 0 is a constant. With a common marginal tax rate of τ , consumption possibilities for both types of entrepreneur are

ζh = (1 − τ )π(w, eh )

and

ζn = (1 − τ )Ben

respectively. Letting l denote the number of hired workers, the indirect profit function π ∗ is obtained by solving the first-order condition ∂F(e, l)/∂l − w = 0 from the profit function π = F(e, l) − wl to get the labour demand function l ∗ = l(e, w), which is put back into π to obtain π ∗ = π(w, e). If the utility function is separable in consumption and leisure, one can write Vh = u(ζh ) − µ

and

Vn = u(ζn ),

Entrepreneurship, job creation and innovation

307

where µ > 0 is a fixed cost of hiring workers. Then using the envelope theorem, it follows directly that ∂(Vh − Vn ) = u
(ζh )π(w, eh ) − u
(ζn )Ben . ∂(1 − τ )

(10.1)

Eq. (10.1) can be expanded in a Taylor series around en to obtain ∂(Vh − Vn ) = u
(ζn )[π(w, en ) − Ben )] ∂(1 − τ )

" ∂F[e, l(e)] "" + u (ζn )(1 + σ )(eh − en ) , " ∂e e=en


(10.2)

where σ is the elasticity of the marginal utility of consumption. The first term on the RHS of (10.2) is the utility of additional net income of being an employer relative to a sole trader, holding effort constant over the two modes. The second term captures the fact that because effort is generally different in the two modes there are additional effects on consumption and marginal utility. To sign (10.2), assume that effort is more productive in the employer mode than in the sole-trader mode. Then π(w, eh ) − Ben > 0 and also eh∗ > en∗ . It therefore follows directly that if σ > −1, (10.2) is positive. So cutting income taxes would increase the propensity of entrepreneurs to hire labour (i.e. switch from n to h status). In their empirical work, Carroll et al. (2000a) estimated the probit equation

Pr[l88 > 0] =  ξ0 + ξ1  ln(1 − τ ) + ξ2 [l85  ln(1 − τ )] + ξ3 l85 + γ
X . (10.3) The dependent variable in (10.3) is the probability that a given entrepreneur is an employer in 1988;  ln(1 − τ ) := ln(1 − τ88 ) − ln(1 − τ85 ) is the change in the log of one minus the marginal tax rate resulting from the 1986 TRA; and l85 is a dummy variable taking the value one if the entrepreneur employed labour in 1985 and zero otherwise. The interaction term allows the response to changes in taxes to vary by employer status. Carroll et al.’s estimates of ξ1 and ξ2 are used to calculate the income tax/labour hiring elasticity discussed in the text. Notes 1. This material draws from and updates Parker (2006b). 2. See Carroll et al. (2000a), Kuhn and Schuetze (2001) and Moralee (1998), respectively. 3. The Cobb–Douglas production function is F(K, l) = δK α (1 + l)β , α, β > 0, where δ > 0 is a measure of human capital, K is physical capital and l is hired labour. It follows from this specification that ‘job creators’, i.e. those entrepreneurs for whom l ∗ > 0, have greater human capital δ than sole traders for whom l ∗ = 0. 4. The Cobb–Douglas forces the elasticity of substitution between capital and labour to be unity, and imposes an infinite elasticity of substitution between own labour supply and outside labour. 5. Letting e denote the entrepreneur’s own effort, Carroll et al. (2000a) observed that ∂l ∂l ∂e = . ∂(1 − τ ) ∂e ∂(1 − τ )

(10.4)

308

Performance

To estimate responsiveness of wl, multiply both sides of (10.4) by w, which is assumed to be constant in competitive markets. The sign of the second of the derivatives on the RHS of (10.4) is positive, but the sign of the first derivative is ambiguous, depending on whether outside and inside labour effort are complements or substitutes. 6. See Leonard (1986), Dunne et al. (1989a, 1989b), Brown et al. (1990), Acs and Audretsch (1993), OECD (1994a), Birch et al. (1997), Baldwin (1998), SBA (2001) and Acs and Mueller (2008). According to SBA (2001), small firms account for about 75 per cent of net new jobs created each year in the USA. A similar proportion of new jobs were created in the self-employment sector in Canada in the 1990s (Lin et al., 2000). 7. To see this, let Ht∗ be true employment size at time t, and Ht be observed size: Ht = Ht∗ +vt , where ∗ + u , where u is a transitory vt is a measurement error with variance σv2 . Suppose Ht∗ = Ht−1 t t shock that is independent of vt . Then the true average conditional change in employment size is zero: ∗ ) = 0. E(Ht∗ |Ht−1

However, were the conditional change in employment to be estimated using actual data, which are subject to measurement error and transitory shocks, one would erroneously find E(Ht |Ht−1 ) = −[σv2 /(σH2 ∗ + σv2 )]Ht−1 < 0, implying a spurious negative relationship between firm growth and size. 8. See e.g. Baldwin and Picot (1995), Konings (1995), Hart and Oulton (1996) and Davidsson et al. (1998). The role of small firms in job creation is if anything even more pronounced in transition economies, as large state-owned incumbents have shed jobs en masse (Konings et al., 1996). 9. See Brown and Medoff (1989), Brown et al. (1990), Ringuedè (1998), Oi and Idson (1999) and Wunnava and Ewing (2000). 10. Cohen and Klepper (1992), Almeida and Kogut (1997) and Sørensen and Stuart (2000). 11. See Scherer (1980, pp. 407–38; 1984; 1991), Acs and Audretsch (1988), Audretsch (1991), Cohen and Klepper (1996) and Klepper (1996). 12. Also consistent with these arguments is differences in the sources of innovation by firm size: ‘Entrepreneurs commercialise innovations to a larger extent, but score lower on the adoption of innovations than their [larger] counterparts’ (van Praag and Versloot, 2007, p. 377). 13. For similar evidence from the US cement, minicomputer and airline industries, see Tushman and Anderson (1986). 14. Acs, Audretsch and Feldman (1994) estimated a ‘knowledge production function’ of the form Iik = β0k + β1k ln RDi + β2k URi + β3k ln(UR.GC)ik + uik , where i indexes firms and k the type of firm (small – under 500 employees – or large); I is the number of innovations in 1982; URi is expenditure on university research which may be accessible to firm i; and RD is industry R&D that may be accessible to i. GC is a measure of geographical propinquity of university and industrial research: this captures spillover effects. Acs, Audretsch and Feldman (1994) estimated this function by tobit and found that βˆ1 was largest for large firms while βˆ2 was largest for small firms. Also, βˆ3 was positive but insignificant for large firms (= 0.033 with t-statistic 0.687) but positive and significant for small firms (= 0.111 with t = 1.965). 15. For other evidence that the presence of external knowledge sources (e.g. large R&D-intensive firms and universities) in regions increases the innovative output of firms located in those regions, see Jaffe (1989), Jaffe et al. (1993), Feldman (1994), Audretsch and Feldman (1996), SBA

Entrepreneurship, job creation and innovation

309

(2002b), Audretsch and Lehmann (2006) and Aharonson et al. (2007). According to Stephan (1996) and Audretsch and Lehmann (2006), proximity to universities which excel in social (rather than natural) science outputs is associated with greater density (and subsequent IPO performance) of start-ups. 16. Pavitt et al. (1987); Acs and Audretsch (1987a, 1987b, 1988, 1991); Marvel and Lumpkin (2007).

11 Entrepreneurship and growth

Once a venture has obtained finance and made the transition to start-up, what determines whether it grows and survives, or stagnates and exits? These issues are examined in the present chapter and in chapter 14. There is evidently more to successful entrepreneurship than simply deciding whether or not to start a new firm. Many entrants create little value and do not last long, while others grow to a considerable size and end up creating substantial wealth. Clearly growth plays a central role in any discussion of value-creating entrepreneurship. For their part, policy-makers appear to be moving away from wholesale efforts to encourage new entrants, in favour of promoting a business climate where growth-orientated businesses can more easily create jobs and wealth for the economy. Most firms start small, so to achieve these laudable goals they have to grow. It is interesting in this respect that a well-known theory of venture growth, Gibrat’s Law, effectively rules out any role for activist public policies. Gibrat’s Law states that growth rates are unrelated to venture size and age, and are effectively random. In fact, the evidence reviewed below suggests that while Gibrat’s Law approximates growth patterns for large firms, it does not do so for small ventures, which tend to grow at a faster rate. After outlining Gibrat’s Law (and some extensions), several theories of entrepreneurial growth are discussed. The second section of this chapter reviews evidence relating to these theories and identifies the salient empirical determinants of venture growth. Three groups of empirical growth determinants are identified: those associated with the founder, those associated with the firm, and those associated with the entrepreneur’s strategic choices. The basic message to emerge from the empirical literature is that while a few measurable factors seem to influence growth, by far the largest source of variation in enterprise growth rates remains unaccounted for. This might explain why Gibrat’s Law, which predicts that growth is random, has proven to be such a useful (if only approximate) representation of venture growth in the literature. The final section of the chapter moves from the level of the individual venture to the more aggregate levels of industries, regions and national economies. The basic question explored here is whether there is a relationship between entrepreneurship and aggregate economic growth. Several useful econometric growth models are introduced 310

Entrepreneurship and growth

311

as we proceed throughout the chapter; they are collected together in a table located in the chapter appendix.

11.1 Theories of venture growth

Several theories of growth are discussed in this section. The first one is Gibrat’s Law of Proportionate Effect. The second is Jovanovic’s (1982) industry selection model of entry, learning, investment and exit. In contrast to the ‘atheoretical’ model of Gibrat, Jovanovic’s theory studies growth at the level of the individual firm and goes on to derive properties of industry equilibrium. It generates a rich array of testable propositions. Similar properties are shared by a different strand of theorising linking innovation, growth and shakeouts, which is associated with the work of Steven Klepper. Both Jovanovic’s and Klepper’s writings have had a major influence on how growth is conceptualised, at the level of the individual firm as well as for entire industries. The final part of this section considers some alternative theories of the entrepreneurial growth of new enterprises. 11.1.1

Gibrat’s Law Gibrat’s ‘Law of Proportionate Effect’(Gibrat, 1931) is a well-known benchmark model of firm growth. Gibrat’s ‘Law’ states that if there is a fixed number of firms, and if firms’ growth rates are random draws from a normal distribution which are independent of firm size and previous draws, then the distribution of firm sizes will be lognormal with a variance that increases over time. An econometric specification of Gibrat’s Law appears in the first row of Table 11.2 in the chapter appendix. Early studies found some empirical support for lognormal firm size distributions and for the assumption that venture growth rates are roughly independent of firm size (see e.g. Hart and Prais, 1956; Simon and Bonini, 1958). More recent studies (Audretsch et al., 2004; Petrunia, 2008) emphatically reject Gibrat’s Law. Three assumptions underlying Gibrat’s Law appear especially questionable. One is that the population of firms is fixed. In practice, of course, firms are born and die, so this assumption cannot be tenable in a strict sense. A second assumption is that each firm faces a draw from a common distribution of random shocks. But recent evidence, discussed below, suggests that the variance of firm growth rates is higher for small than for large firms. Third, Gibrat’s Law assumes that mean growth rates are the same for all firms and for all time. Again, recent evidence reviewed below refutes this claim, showing that smaller firms tend to have higher growth rates than larger firms. In the light of these problems several refinements to Gibrat’s Law have been proposed (see Ijiri and Simon, 1977). One contribution modifies the Law to allow for idiosyncratic firm-specific components of growth and for growth rates which depend on firm size.1 Another modification adds non-linear age, a, and size, q, effects to the model (see (11.3) in Table 11.2). This permits testing of the Jovanovic (1982) model described next.

312

Performance

11.1.2

Jovanovic’s (1982) model of industry selection In a seminal paper, Jovanovic (1982) developed a model in which entrepreneurs have imperfect information about their innate abilities, which can only be learned by trying entrepreneurship. Jovanovic’s model predicts that the ventures of able and/or lucky entrepreneurs survive and grow, while those of less able and/or unlucky entrepreneurs shrink and exit. It derives endogenously the entry, growth and exit behaviour of a cohort of new ventures within an optimising framework. As will be seen later, the predictions of the model turn out to be consistent with a large body of empirical evidence about new firm dynamics. The Jovanovic model shares several features in common with some of the models of entrepreneurship discussed in chapter 2. As in Lucas (1978), individuals have heterogeneous abilities in entrepreneurship, x; but unlike Lucas, entrepreneurs do not know their abilities when they start their ventures. As in the Kihlstrom and Laffont (1979) model, entrepreneurs face a risky trading environment; but unlike Kihlstrom and Laffont, all individuals are assumed to be risk-neutral rather than risk-averse. Finally, like Dixit and Rob (1994) and Parker (1997aa), the Jovanovic model is dynamic. But in contrast to those models, there are no switching costs and entrepreneurs learn by doing. In particular, entrepreneurs learn about their idiosyncratic x, about which they are initially uncertain. For simplicity, Jovanovic assumed that all entrepreneurs start with the same prior belief about their ability, so they all enter at the same scale of operation. Entrepreneurs use a Bayesian updating rule to adjust their beliefs about their ability as information comes in. Although relatively able entrepreneurs are more likely than less able entrepreneurs to receive information which signals high ability, this is not guaranteed to occur. Thus able entrepreneurs can receive an unlucky sequence of bad draws, and erroneously come to believe that they possess low ability. Conversely, less able entrepreneurs can receive a lucky sequence of good draws, and erroneously come to believe that they possess high ability. Beliefs matter because they influence entrepreneurs’ output decisions. Entrepreneurs increase output if they infer high ability and decrease output if they infer the opposite. As time goes on, unrepresentative good and bad random draws should cancel out, and innate ability is increasingly likely to shine through. As in all occupational choice models, individuals do not have to be entrepreneurs. Entrepreneurs who believe they are able remain in entrepreneurship, and increase their output in accordance with their perceived ability. Entrepreneurs who come to believe they possess low ability exit and take a valuable outside option instead. All exits are voluntary; entrepreneurs quit when the expected value of continuation in entrepreneurship falls below some positive threshold given by the outside option. Hence there are no involuntary bankruptcies. Jovanovic’s model generates the following implications for the distribution of ability, venture growth rates, exits, industry concentration, profits and output:

J.1 Entrepreneurs who run young firms have had less time to accumulate information about their true abilities. Therefore the level and variability of growth rates are

Entrepreneurship and growth

J.2

J.3

J.4

J.5

J.6 J.7

J.8

313

highest among younger and smaller firms. Growth rates are lowest among mature surviving firms. Firms that exit never return to the market, in part because an exiting entrepreneur never obtains any more information that could change their terminal belief about their innate ability. Results J.1 and J.2 imply that (cohort) industry concentration increases over time, since initially all firms of a given cohort were the same size, at which point industrial concentration was at a minimum. Entrepreneurs who remain in business indefinitely eventually learn their true ability, whereas entrepreneurs who exit possess relatively imprecise estimates of their true abilities. The last clause of J.4 implies that some entrepreneurs who exit might have true entrepreneurial abilities which would have made entrepreneurship optimal for them. Unfortunately, these entrepreneurs received ‘unlucky’ draws from the distribution of shocks, and incorrectly (but understandably) interpreted these draws as evidence of low ability, prompting their decision to exit. The greater the noise in the learning process, the more of these ‘efficient but unlucky’closures there will be. Surviving firms are larger and older than both firms which failed and new entrants. Larger, more efficient firms earn profits as a reward for their exceptional ability. For each cohort of firms, average profits increase as the industry matures. The distribution of profits resembles the distribution of ability. The more dispersed the latter, the more dispersed are firm sizes and profitability rates. Entry occurs in every period.

As will be seen below, Jovanovic’s predictions relating to firm growth rates (J.1) receive strong empirical support. Additional predictions relating to survival will be discussed in chapter 14. At the risk of ending this section on a negative note, however, I shall conclude by mentioning a few less satisfactory aspects of the model. First, it is debatable whether in practice average cohort firm size and industrial concentration increase continually over time (in a first-order stochastic sense), as suggested by J.3 and J.6. One can think of many practical reasons, absent from Jovanovic’s model, why firms might face barriers to growth, including regulatory constraints and tax disincentives. Also, entrepreneurs have only finite attention to apportion between maintaining current projects and evaluating and adopting new ones. The opportunity cost of neglecting profitable current projects might deter firms from seeking growth and so place bounds on firms’ sizes (Oi, 1983; Gifford, 1998). While some entrepreneurs might be able to release scarce resources to exploit profitable new opportunities by closing or transferring viable existing businesses to less able entrepreneurs (Holmes and Schmitz, 1990), this is not always feasible. A second questionable feature of the model is that entrepreneurs who exit never return and ‘try their luck again’ (J.2). As noted in the appendix to chapter 1, between 20 and 30 per cent of entrepreneurs at any given point in time are ‘serial entrepreneurs’. Many entrepreneurs who have run an unsuccessful venture the first time around are keen to

314

Performance

try again (Stam et al., 2008). In fact, re-entry by failed entrepreneurs can be sustained in a modification of Jovanovic’s model by replacing the assumption of an inexhaustible supply of potential entrepreneurs with an alternative assumption stipulating a limited supply (Brock and Evans, 1986, pp. 60–3). Yet the possibility remains that entrepreneurs want to re-enter because their unsuccessful entrepreneurial experience nevertheless enabled them to perform ‘active’ learning (which enhances their abilities and hence future expected returns in entrepreneurship) rather than the ‘passive’learning envisaged by Jovanovic (Stam et al., 2008). More will be said about the distinction between active and passive learning in chapter 12. Third, the next section notes that entry eventually ceases in the mature stages of most industry life cycles. This refutes prediction J.8. Fourth, Jovanovic’s model is not the only theory to generate several of the predictions listed above, although it is economical in generating so many predictions from such a simple theoretical framework. Below I briefly consider several other theories which have also influenced our thinking about the entrepreneurial growth process at the levels of both the individual firm and the industry.

11.1.3

Innovation, growth and shakeouts Klepper (1996) and Klepper and Simons (2000a) offer an alternative theory of the growth of enterprises. These authors analyse the evolution of innovative industries by building on ‘stage’models of product life cycles and the theory of shakeouts, articulated in Gort and Klepper (1982) and Klepper and Graddy (1990). Unlike Jovanovic (1982), Klepper’s (1996) theory has surviving entrepreneurs not learning more, but innovating more. Suppose that owing to some innate differences in entrepreneurial talent or access to resources, some entrepreneurs can innovate while others can only imitate. Costly R&D is needed to generate innovations. Innovations reduce average production costs and increase output and profits for the innovators. Larger ventures optimally undertake the most R&D because they can spread the lower resulting average costs over a larger scale of output. The greater output and lower average costs increase competitive pressure on the imitators. Next period, all imitator firms which can survive the growing competitive pressure catch up and adopt the innovation themselves. Average costs therefore fall across the board and aggregate output rises. For a given industry demand, greater output decreases the market-clearing price. This process continues over time with further rounds of innovation, imitation, rising output and falling prices. An interesting set of predictions about industry structure follows if prices decrease by more each period than average production costs decline following the initiation and imitation of last period’s innovations:

• Innovation promotes ongoing entry and growth. But price–cost margins shrink over time, eventually choking off entry by imitators, and ultimately even entry by new innovators. Entry eventually ceases altogether.

Entrepreneurship and growth

315

• There is ongoing exit of high-cost firms and later entrants (which are smaller on average). A shake-out of producers occurs, despite growing industry output (which, for example, reached 20 per cent per annum in the early years of the auto industry). • The entrepreneurs who were the earliest innovators grow rapidly and ultimately end up dominating the industry. These innovators have lower average costs and so are always larger than imitators who entered at the same time. Earlier entrants have a lower hazard of exit at older ages. Thus size, age, growth, innovation and survival are all interrelated. Innovation drives growth, which increases the gains (and hence incentives) from further innovation. The pressure of competition then leaves the fastest-growing and most innovative ventures to dominate the market, eventually culminating in an oligopoly (Klepper, 2002). Like Jovanovic (1982), this theory also carries important implications for firm exit, which will be discussed in chapter 14. This model takes early entrants to be the innovators. This implies a form of firstmover advantage in entrepreneurial innovation. The reason is exploitation of economies of scale. In fact, Geroski (2003) identified this as just one of four sources of first-mover advantage. Others include monopolisation of scarce inputs; lock-in of consumers who are reluctant to change products once they become accustomed to them; and enhanced brand identity and status from being first to market. According to Geroski (2003), these benefits generally outweigh second-mover advantages such as learning from the mistakes of early movers and free-riding on their efforts. While shakeouts have been observed in several industries, they are not ubiquitous. In both the USA and Germany, for example, the laser industry has evolved differently from traditional manufacturing industries, exhibiting sustained entry and neither shakeouts nor first-mover advantages for early entrants (Buenstorf, 2007). It is not yet clear why this industry has exhibited such different development patterns from autos and other manufacturing industries. 11.1.4

Other theories of growth

Other models also predict that smaller firms will have higher and/or more variable growth rates. For example, Segal and Spivak (1989) analysed an entrepreneur who chooses how much profit to take out of her business and how much to re-invest in it. Re-investment of profits increases output (which is subject to random shocks), enabling small firms to rapidly attain the industry’s minimum efficient scale (Audretsch, 1991). An entrepreneur’s firm fails and closes if output reaches zero, perhaps following an unlucky sequence of successive adverse shocks. In that event, the owner pays an intangible dissolution cost – perhaps the loss of reputation. Now consider an entrepreneur who enters the market with an output that is positive but sufficiently close to the zero boundary to make the prospect of failure a real possibility. Clearly, this entrepreneur has a particularly strong incentive to reinvest profits in order to decrease the risk of costly failure, compared with a manager of a larger enterprise producing output that is further from the boundary. It therefore follows that small firms have higher output growth

316

Performance

rates than larger firms, since re-investment of profits promotes growth. An interesting by-product of Segal and Spivak’s (1989) model is the prediction that large firms have average growth rates that converge to a constant – consistent with Gibrat’s Law. Another reason why small new firms may grow faster than their older larger counterparts is attributable to sunk costs (Cabral, 1995). There is often an option value of waiting before making sunk cost investments. Consequently new firms will optimally start operations at a small scale and only exercise an expansion investment option if circumstances prove to be favourable. In contrast, large firms optimally invest more initially and therefore have lower subsequent growth rates. In support of this theory, Geroski (1995) reported that most new firms start out with output less than the industry minimum efficient scale (MES). The Industrial Organisation literature has also spawned several other models of optimising behaviour under conditions of risk which give rise to negative relationships between firm age, growth and exit. In an important contribution which diverges from Jovanovic (1982), Ericson and Pakes (1995) supposed that each firm knows the value of a parameter which determines the distribution of its profits. But the value of that parameter responds dynamically to the investments made by the firm and its competitors. Ericson and Pakes (1995) predict that small firms optimally start with a relatively low level of investment, and can never compete with larger firms which continue to invest and scale up output and so entrench a competitive advantage over them. Similarly to Klepper (1996), most entrants die young. Survivors invest, grow and learn until they reach a ‘coasting’ point. Pakes and Ericson (1998) discuss empirical strategies to distinguish their model from that of Jovanovic (1982), and argue that different models may apply in different industries.2

11.2

Evidence about the growth of entrepreneurial ventures

Many new enterprises do not grow. Even among those which do, rapid growth is difficult to achieve and maintain. Most start-ups are imitative businesses in mature industries serving local markets, which necessarily have limited growth potential (Aldrich, 1999). Only about one in seven firms generates sustained profitable growth. It is noteworthy that a mere handful of new firms enjoying spectacular growth rates create the majority of new jobs. In the context of the small firms literature, Storey (1994a) estimated that 5–10 per cent of UK manufacturing businesses which started in the 1970s provided 40–50 per cent of total employment in their cohort ten years later. David Storey refers to these firms as the ‘ten-percenters’. Likewise, only five per cent of American businesses grow their employment by 15 per cent per annum or more over five-year time spans (Zook and Allen, 1999; Barringer et al., 2005). Bhide (2000, p. 13) cites work by Birch and Medoff claiming that 4 per cent of all US firms generate 60 per cent of all new jobs in the US economy. These fast-growing firms are called ‘gazelles’. British data show that most gazelles cannot sustain their rapid growth rates (Parker, Storey and van Witteloostuijn, 2005).

Entrepreneurship and growth

317

Understanding the determinants of growth holds out the promise of understanding ongoing value creation by entrepreneurs. However, in any discussion about the growth of entrepreneurial firms, it is imperative to define clearly at the outset how growth is to be measured. The first subsection below considers the relevant issues. The one after summarises evidence from multivariate regression analyses about the relationship between entrepreneurial growth performance gi among a sample of firms indexed by i, and three broad groups of explanatory variables Xi : founder characteristics, firm attributes and entrepreneurial strategies. The ‘standard form’ of these regressions defined over a sample of ventures indexed by i is given in row 4 of Table 11.2. At no point will I discuss so-called ‘stages of growth models’ which are popular in many business studies courses (e.g. Churchill and Lewis, 1983). These simple descriptive constructs assume that ventures develop their organisational structures in a predictable fashion from birth through several pre-defined growth stages until maturity. As several critics have pointed out, they assume an excessive pre-determination and homogeneity of development, and understate the role of the entrepreneur (Bhide, 2000; Davidsson, Achtenhagen and Naldi, 2006). The latter criticism can also be applied to ‘evolutionary’ theories of enterprise development (Nelson and Winter, 1982) which attribute growth to luck and exogenous routines rather than to the purposeful choices of entrepreneurs and other decision-makers. 11.2.1

Definitional and measurement issues Before conducting an empirical analysis, the researcher first needs to address several basic questions about what is meant by ‘growth’. For example, is the researcher interested in tracking an entrepreneur’s growth performance over time, or in following the growth of a particular venture, regardless of the sequence of ownership during its lifetime? Over what time period is growth to be measured? Should it be measured across ventures spanning different industries or should a more restricted and homogeneous sample be chosen? Is the researcher interested only in ‘organic’ growth or is he or she prepared to count growth by acquisitions as well? What type of data (e.g. panel or cross-section) and empirical methods are to be used to estimate the determinants of growth? Should ‘the growth rate’ be defined in relative or absolute terms (see below)? And what quantity is growth measured for: sales, employment or something else? Answers to most of these questions depend upon the context. Rightly or wrongly, and for a plethora of reasons which researchers rarely record, most of the evidence cited below is drawn from studies which analyse the growth rates of a set of ventures each of which could be associated with one or more individual entrepreneurs. These studies typically measure growth between one and five years from the date of first sampling (depending on data availability), and usually sample ventures from a range of industries. Shorter time spans for assessing growth reduce the risk that ventures change so much (e.g. by acquisition) that one is no longer measuring the same entity; but in these cases outcomes can be distorted by transitory idiosyncratic shocks. Also,

318

Performance

there can be a long time lag before entrepreneurial growth occurs. For example, only one-tenth of new firms between 1976–86 grew in employment terms in their first four years, but over one-half had grown within eight years (Phillips and Kirchhoff, 1989). Recommendations are easier to make regarding some of the other issues. The majority of growth in new and small enterprises is organic, whereas that in established large firms is often acquisition-based (Davidsson, Delmar and Wiklund, 2006). Organic growth is usually regarded as more interesting than growth by acquisition, although surprisingly few empirical studies actually differentiate between them. And despite the superiority of panel over cross-section data, and of multivariate econometric analysis over bivariate comparisons, some studies still report results based on the latter rather than the former. There appears to be greater disagreement in the literature about whether growth should be measured in relative or absolute terms; and which variable should be the subject of growth. Let q(t) be a performance variable (e.g. revenue) and write q(t) = q(t) − q(t − 1). Relative growth (the ‘growth rate’) is q(t) ∂q(t)/∂t ≈ , dt→0 q(t) q(t − 1)

g =  ln q(t) = lim

as can be readily seen by manipulating the growth process q(t) = q(0) exp{gt}.Absolute growth is simply q(t). ‘Small’ firms (i.e. with small values of q(t − 1)) tend to record higher relative growth but lower absolute growth than large firms. It is sometimes argued that relative growth rates overstate the actual growth performance of small relative to large firms, because of the small denominator for small firms. However, the same is true of GDP growth rates for poor developing countries and few objections are made in that context. Furthermore, the use of relative measures is consistent with the theories outlined earlier. Most of the studies cited below use relative rather than absolute measures of growth. Variables whose growth rates are often studied include profitability, revenues (or sales turnover), returns on assets and the number of employees. Each measure has its strengths and weaknesses. Sales growth precedes employment growth in the newest ventures, making it an obvious measure of choice in samples of new firms (Davidsson, Delmar and Wiklund, 2006). And sales growth rather than employment tends to be the indicator of performance that many entrepreneurs use themselves (Barkham, Gudgin et al., 1996; Robson and Bennett, 2000). Moreover, ‘sales’ is probably superior to ‘assets’ as a growth variable because few new ventures have much in the way of tangible assets, especially those in the service sector. On the other hand, policy-makers are usually more interested in employment growth than in sales growth. Davidsson, Delmar and Wiklund (2006) observe that for a broad range of enterprises, employment growth appears to be more strongly correlated with other growth measures than sales growth is. Rather than selecting one growth variable, some researchers use a compendium of measures in an attempt to obtain a more rounded understanding of growth. An alternative strategy is to aggregate them into a portmanteau measure. The latter however can be hard to interpret. With these preliminary remarks in mind, I now discuss the determinants of growth at the level of the individual venture.

Entrepreneurship and growth

319

11.2.2

Evidence about determinants of venture growth Storey (1994a) proposes the following three groups of determinants of entrepreneurial venture growth:

I. Founder characteristics II. Firm attributes III. Strategies of the entrepreneur This classification will be used to frame the empirical results which follow. These results are drawn from multivariate econometric analyses of reduced-form growth models. Reduced-form models are prone to the problem that some individual characteristics and firm attributes might affect growth indirectly through hard-to-measure entrepreneurial strategic choices (Wiklund, 2007). This is a type of endogeneity problem, which will be found to recur several times below. Table 11.1 provides a selected summary of prominent findings from the literature. Under I, the discussion in chapter 4 would probably lead one to expect human capital to improve venture growth prospects. As the first panel of Table 11.1 shows, this is broadly (but not uniformly) what emerges from previous research on education and experience. Digging into the education variable, Almus and Nerlinger (1999) reported higher employment growth in new ventures whose founders have degrees in technological subjects and business. With respect to experience, managerial experience appears to have a stronger positive effect on venture growth than previous entrepreneurial experience in the same industry sector – while the effects from founder’s age and previous time spent in self-employment are mixed (Storey, 1994a). Mixed effects have also been found for gender, training, membership of a business network and having a self-employed parent (see the entry in Table 11.1 under ‘social capital’). Several authors have cited positive effects from entrepreneurs having growth motivations at the outset.3 But the evidence about motivations is not uniform. For example, Birley and Westhead (1994) found no such effects. Perhaps growth aspirations are best interpreted in a ‘negative’ way: i.e. one would be surprised if entrepreneurs who did not intend or desire to grow in the future ever did so. It is important in this light to remember that firms can sometimes do better by not growing. For example, some small firms retain flexibility by relying on interdependent networks of external businesses offering specialised services (Saxenian, 1994) – benefits which could be lost or compromised if they attempted to ‘make rather than buy’ by growing internally instead. These considerations might explain why, for example, less than one-half of small firms in the UK regard employment growth as an objective, even in economically buoyant times (Hakim, 1989b). Scase and Goffee (1982) claimed that some entrepreneurs do not hire workers if they suspect this will destroy a personal client-based service; if they believe that hired labour is unreliable; or (among some artisans) if hiring entails spending one’s time as a ‘businessman’ rather than as a ‘craftsman’. Other entrepreneurs fear losing the family atmosphere of a small firm and dislike having to rely on external capital with its perceived (or actual) loss of control.

320

Performance

Table 11.1. Determinants of entrepreneurial venture growth I. Founder characteristics Variable Education

Effect +

0 Experience

+

Age

?

Male gender Training Social capital

+ 0 ? ?

Growth motivation

? ?

Evidence Cooper et al. (1991, 1994); Westhead (1995); Westhead and Cowling (1995); Reynolds and White (1997); Almus and Nerlinger (1999); Kangasharju and Pekkala (2002); Akoten et al. (2006); Tamásy (2006) Schutjens and Wever (2000); Haber and Reichel (2007); Capelleras and Greene (2008) Cooper et al. (1991, 1994); Kangasharju and Pekkala (2002); Koeller and Lechler (2006) Cf. Storey (1994a); Barkham, Hart and Hanvey (1996); Capelleras and Greene (2008) Cooper et al. (1991) Kalleberg and Leicht (1991); Capelleras and Greene (2008) Cf. Skuras et al. (2005) and Foreman-Peck et al. (2006) Network membership: Cf. Bosma et al. (2004) and Foreman-Peck et al. (2006) Parent se: Cf. Cooper et al. (1991) and Tamásy (2006) Cf. Storey (1994a) with Birley and Westhead (1994) II. Firm attributes

Variable

Effect

Evidence

Initial firm size Firm age Venture team size

+

Limited liability

+

(Reinvested) profits

?

Numerous: see text Numerous: see text Eisenhardt and Schoonhoven (1990) Cooper et al. (1991, 1994); Barkham, Hart and Hanvey (1996); Reynolds and White (1997); Schutjens and Wever (2000); Tamásy (2006); Westhead and Howarth (2006) Koeller and Lechler (2006) Storey (1994a); Almus and Nerlinger (1999) Capelleras and Greene (2008) Cf. Reid (1993) and Koeller and Lechler (2006) III. Strategies of the entrepreneur

Variable Sources of finance Personal savings Formal processes

Effect + + +

Evidence Westhead and Cowling (1995); Capelleras and Greene (2008) Barkham, Gudgin et al. (1996); Koeller and Lechler (2006); Foreman-Peck et al. (2006)

Entrepreneurship and growth

321

Table 11.1 (cont.) III. Strategies of the entrepreneur Business plans

+ 0

External assistance

+ ?

Schutjens and Wever (2000); Masuerl and Smit (2000) Orser et al. (2000) Reid and Smith (2000); Tamásy (2006); Haber and Reichel (2007); Capelleras and Greene (2008) Links to external firms: Almus and Nerlinger (1999) Outsourced retail: Koeller and Lechler (2006) Financial and advisory assistance: Cf. Reid and Smith (2000); Haber and Reichel (2007); Capelleras and Greene (2008)

Notes: ‘Effect’ summarises the consensus about the sign of the relationship, with ‘?’ indicating mixed or inconclusive results.

Turning to II, a widely documented finding is that growth rates are decreasing in venture size among ventures of the same age, and are decreasing in venture age among ventures of the same size. These findings have been obtained using several of the specifications outlined in the first part of Table 11.2.4 In addition, younger firms have more variable growth rates (Brock and Evans, 1986; Bates, 1990). These results all support Jovanovic’s (1982) prediction J.1. Occasionally, however, researchers identify data samples in which long-term growth is found to be higher among ventures which commit more resources (including hiring of employees) at the time of launch (e.g. Schutjens and Wever, 2000). This outcome is more likely to occur when the study omits controls for venture quality, since one would expect entrepreneurs to invest more resources in higher-quality ventures. One index of quality is high technology. In this regard, Almus and Nerlinger (1999) reported significantly higher employment growth rates among new technology-based firms (NTBFs) which, unlike non-NTBFs, created jobs on net in Germany over the period 1989–96. One concern is that several of the growth determinants considered so far might be endogenous. This concern deepens when discussing other variables under II such as the size of the founding team, the company’s legal status, and its profitability. As noted in chapter 4, multiple founders may have access to a broader, more heterogeneous pool of skills and experience and may give each other technical and psychological support (Eisenhardt and Schoonhaven, 1990). One would therefore expect venture team size to enhance growth. As the second panel of Table 11.1 reveals, this conjecture is largely borne out by the evidence. But the number of founders is a choice variable and is likely to be positively correlated with the quality of a venture, making this variable endogenous. Although there is some evidence that limited-liability and multiple-establishment ventures (Variyam and Kraybill, 1992) also have higher growth rates than ventures which do not, these variables are also prone to endogeneity. Surveying previous work, Davidsson, Achtenhagen and Naldi (2006) concluded that a link between profitability and growth is still not well established. It is important

322

Performance

to distinguish, however, between trading profits and retained profits. The difference between them comprises directors’remuneration and taxation. Watson (1990) reported a weak relationship between employment growth and trading profits, but a much stronger relationship between employment growth and retained profits. This is consistent with the notion that retained profits are re-invested for expansion. Finally under II, average growth performance varies systematically by industry. For example, using data on 1,053 US entrepreneurs tracked over a three-year period, Cooper et al. (1991) recorded significantly below-average growth rates among ventures located in retail and personal services sectors. Foreman-Peck et al. (2006) reported significantly above-average growth rates for Welsh SMEs located in the financial services sector. Strategy variables (which come under III) are more commonly used by business studies researchers than by economists. Table 11.1 includes a selected set of relatively ‘objective’ determinants of this kind. These include strategies for accessing multiple sources of finance; use of formal information management processes (e.g. computerised accounts); business plans and use of external assistance. Despite occasional claims that business plans and planning in general are associated with superior venture performance (Shane, 2003, p. 223), Table 11.1 shows that the econometric evidence on this issue is mixed. Indeed, Bhide (2000) observed that 41 per cent of the Inc. 500 founders in his sample had no business plan at all at start-up. Bhide speculated that the expected costs of formal planning for these entrepreneurs outweighed the expected benefits. In contrast, other forms of external linkages appear to promote growth, including connections with other firms and outsourcing of product distribution (see Table 11.1). Franchising is widely regarded as a facilitator of rapid growth in industries where opportunities can be codified, especially at early stages of the venture life cycle (Martin, 1988; Michael, 1996). Investors sometimes balk at the size of funding requests from entrepreneurs wanting to ‘get big fast’, but franchisees can help entrepreneurs spread start-up costs and risks in an economical way.5 By making franchisees residual claimants and by including termination clauses, franchisors can reduce agency costs while enabling self-selection of able franchisee managers (Norton, 1988). A franchising strategy can therefore promote growth-oriented entrepreneurship. The findings relating to entrepreneurs’ strategies under III should be interpreted with considerable caution. Unusually promising ventures might adopt particular strategies to promote growth that would be entirely inappropriate for more run-of-the-mill firms (which probably constitute the majority of start-ups). Related to this point, it has been recognised that strategy has to fit the context to be effective (Parker and van Witteloostuijn, 2009). For example, for small NTBFs, the founding team’s technological experience is an important determinant of the success of a product differentiation strategy (Shrader and Siegel, 2007). If fit to context matters more generally, few generalisations about appropriate strategies may be possible. The problem is compounded by the possibility that several aspects of strategy could be important but hard to measure. These considerations might hold the key to explaining why strategy variables are generally found to have only weak effects in venture growth regressions (Storey, 1994a).

Entrepreneurship and growth

323

In a lively descriptive account, Bhide (2000) discusses some of the hard-to-measure determinants of enterprise growth which have largely eluded researchers working in the econometric tradition. Bhide reports that most of the highly successful Inc. 500 firms started as marginal businesses with little in the way of human capital, unique product offerings or even resources. He argues that the effective execution of an ordinary idea creates more value than haphazard development of an exceptional idea; he went on to speculate that gazelles are scarce because the entrepreneurial qualities needed to start a venture – a capacity for opportunistic adaptation, a tolerance of ambiguity, perceptiveness and salesmanship – are not the same as the qualities needed to grow a venture rapidly. Bhide conjectured that the latter qualities include ambition, a tolerance of risk, imagination, a talent for creative synthesis, and a capacity to abstract general principles from specific situations and to formulate sound long-term plans. Few individuals make the cut for starting a venture; even fewer make the further cut of being able to grow successfully post start-up. To conclude, the state of knowledge about entrepreneurial growth at the level of the individual venture remains rather limited. That might seem surprising given the simple organisational structures, lack of organisational inertia and relatively uncomplicated strategies utilised by many new ventures (Shrader and Siegel, 2007). Part of the reason for the limited progress on this front is no doubt attributable to the pronounced heterogeneity of entrepreneurs and their ventures (Santarelli and Vivarelli, 2007). This might explain why empirical studies can usually explain no more than a small fraction of the variation in enterprise growth rates. For example, Westhead and Cowling (1995) found that of sixty-seven potential explanatory variables (embracing personal and firmspecific characteristics, strategies, environmental factors and financial structure, inter alia), only a handful – initial size, the owner’s education and sources of financing – significantly explained firm growth. Findings like these are not encouraging and place serious limitations on the ability of researchers to formulate concrete advice for policy-makers keen to promote new venture growth. The limited findings obtained to date should also caution against targeting assistance by ‘picking winners’. We still know very little about the drivers of growth in most enterprises and cannot confidently predict who the future winners will be. Nor is it obviously a good idea to target government support on ‘high-flyer’ ventures, which are presumably best placed to thrive without it. It is asking too much of a government to acquire the necessary information to target individual ventures for such support, even supposing that it can be relied upon to act effectively when dispensing selective assistance. Another problem is that government policies designed to promote growth and survival may have perverse effects. Even well-intentioned targeted policies designed to enhance venture growth could encourage non-targeted firms to change their behaviour in order to exploit any assistance on offer. And assistance can dilute entrepreneurs’ incentives to take the necessary actions to succeed in the marketplace on their own merits. Perhaps general policies which enhance the stock of human capital, and increase the incentives to innovate, are more promising ways of promoting entrepreneurship and

324

Performance

growth. After all, as we saw earlier, more educated entrepreneurs tend to run firms that grow faster. The tax system might be another effective policy instrument. There is evidence that higher US marginal income tax rates significantly and substantially decrease small firm growth rates (measured in terms of business receipts) and investment expenditures (Carroll et al., 2000b, 2001). More generally, Acs and Armington (2006) contend that policies designed to support higher rates of entrepreneurial and regional growth should focus on the ‘seeding’ of entrepreneurship – by promoting entrepreneurship as an attractive occupational choice; enhancing the ease of commercialisation of new technology; and encouraging the production and transfer of knowledge spillovers. The next section builds on these points by discussing government policy issues in the context of aggregate economic growth and entrepreneurship. 11.3

Entrepreneurship and economic growth

The evidence reviewed above showed that smaller and newer enterprises experience higher average growth rates than their larger and older rivals. This implies that entrepreneurship and growth are positively related at the level of the individual firm. The present section outlines evidence suggesting that they may be linked at higher levels of aggregation as well, including industries, regions and the national economy. However, throughout the review that follows, and for reasons which I will go on to explain, the reader should always interpret the putative relationships in terms of statistical associations rather than in terms of causality. It was noted in the previous section that researchers can typically explain only a small fraction of the variation in growth rates among enterprises. It might be thought that efforts to predict growth at the aggregate level must therefore be even more unsuccessful as disparate heterogeneous sectors are added together. Indeed, to the extent that firm ‘growth’ occurs by acquisition rather than by the creation of new value, aggregate growth might even be lower than that measured at the level of the firm. On the other hand, if knowledge spillovers exist, aggregate growth might amount to more than the sum of its individual parts. And more generally, macroeconomic structure might emerge in an aggregate context which is hard to detect at the level of individual ventures. To flag the main finding in advance, the evidence below suggests that entrepreneurship and aggregate economic growth are positively related. This finding holds for several measures of aggregate entrepreneurship. These include: industry shares of small firms; the number of new businesses; the number of all businesses; and the self-employment rate. One exception is the construct known as ‘turbulence’, defined as the sum of entries and exits in industries or regions. Turbulence has been linked to the Schumpeterian process of creative destruction. Although some individual studies link turbulence to regional productivity rates and growth,6 Caves (1998) concluded from his review of the evidence that turbulence makes little difference to industry rates of productivity growth.

Entrepreneurship and growth

325

The evidence tells a reasonably consistent story regarding the relationship between rates of new firm formation and economic growth at the following three levels: industry, region and national economy. This evidence is important because it addresses a concern that jobs created by new and small firms are often of low quality and so start-ups do little to stimulate economic growth (Harrison, 1994; Brixy et al., 2007). In fact, a more relevant construct than job quality in this respect is productivity. At the level of the industry, there is some evidence that newer and smaller firms boost industrial productivity. For example, in an analysis of US Census Bureau LBD data from 1975–2001, Haltiwanger (2006) found that the net entry of establishments accounted for roughly 30 per cent of productivity growth in US manufacturing over a ten-year horizon, and for nearly 100 per cent in the retail industry. Therefore the contribution to overall industry productivity growth from entrepreneurial entry and selection appears to be substantial. In their overview of the evidence about productivity by firm size, van Praag and Versloot (2007, p. 372) concluded that ‘entrepreneurs may lag behind in the levels of productivity, but they are catching up to the production efficiency of the control group due to a higher growth rate’. With some exceptions (e.g. Peneder, 2008) this verdict is probably emerging as something of a consensus view. Other evidence suggests that the formation of new independent firms is associated with significantly higher rates of regional income growth and in-migration (Davidsson et al., 1994, 1995). These associations are often stronger than those relating to expansions among existing establishments, illustrating the economic importance of new firms. Nor is it only entries by small firms which generate these effects. Carree and Thurik’s (2003) review of the evidence suggests that regions possessing industry size distributions containing above-average small-firm shares enjoy above-average rates of income growth.7 However, few studies delve into the underlying reasons why industries, regions and nations containing higher proportions of small firms enjoy higher rates of economic growth. Several reasons have been suggested: small firms are innovative, promote competition, create jobs and facilitate further entrepreneurship.8 However, we do not know the contribution, if any, made by these factors individually. Likewise, we still know relatively little about why small firms have become more important relative to large ones in many developed economies since the 1970s.9 Another issue is the relationship between regional firm formation rates and rates of growth of regional employment. As noted in chapter 10, while new entrants can in principle create new jobs by exploiting new technology and responding deftly to changes in consumer demand, few entrants create many new jobs in practice. That is because most entrants are imitative rather than innovative. Thus van Stel and Storey (2004) cite figures showing that only 5.5 per cent of the stock of UK employment in 1989 was located in firms created in the previous two years. At the same time, entrants create competition which forces incumbents to shrink, exit or improve their productivity. The overall effect on employment is theoretically ambiguous, and depends

326

Performance

on the relative strength of these offsetting forces. Empirical data are needed to determine which prevails on net. Early empirical research on this issue detected small and/or insignificant net effects of new venture creation on regional employment growth rates (Audretsch and Fritsch, 1996; Fritsch, 1997). However, these studies looked only at short time periods to identify the employment effects of new venture creation. Recent research highlights the importance of time lags when measuring the impact of new venture formation. There are both direct (immediate) and indirect (delayed) effects of venture creation on employment growth. The indirect effects can take six years and more to work through. When lagged as well as contemporaneous start-up rates are included in employment growth regressions, net positive effects of start-ups on regional employment growth are usually observed.10 A robust temporal pattern in employment growth has been detected in several of these studies. A typical example, taken from Fritsch and Mueller (2008), is illustrated as Figure 11.1. This pattern has been interpreted in terms of new ventures initially creating jobs in a region, followed by regional job losses as the stiffer competition created by the entrants begins to displace older firms. Net regional job creation eventually swings back into positive territory as stronger incumbents respond to competition from effective entrants by increasing their productivity, increasing overall employment on net. Similar patterns have been observed at the aggregate cross-country level too (Carree and Thurik, 2008). Not every single industry or region exhibits the temporal pattern illustrated in Figure 11.1. The pattern is most pronounced in regions with dense populations and high levels of productivity (van Stel and Suddle, 2008; Fritsch and Mueller, 2008). It has also been observed to be especially pronounced in the German services industry

Impact of new bu siness formation on employment change

0.20 0.15 0.10 0.05 0.00 –0.05 –0.10 t

t–1

t–2

t–3

t–4

t–5 Start-

t–6

t–7

t–8

t–9

t–10

up rate

Figure 11.1 The lag structure of the impact of new business formation on regional employment change (taken from Fritsch and Mueller, 2008, p. 19)

Entrepreneurship and growth

327

(Fritsch and Mueller, 2004; Fritsch and Weyh, 2006) and in Dutch manufacturing (van Stel and Suddle, 2008). In contrast, an absence of negative displacement effects has been observed in the USA, where firms with between 20 and 499 employees are responsible for most of the positive employment growth from new venture creation (Acs and Mueller, 2008). Recall from earlier chapters that the knowledge spillover theory attributes a special role to regional agglomerations of knowledge-intensive activities which are conducive to entrepreneurship. In a regional context, this is referred to as ‘entrepreneurship capital’, EK.11 EK is a function of local levels of knowledge creation (Audretsch and Keilbach, 2007b), and is usually proxied in applied work by regional rates of new firm formation, i.e. by EKj for a region j. Assume an aggregate Cobb–Douglas production function which includes as productive inputs EKj and R & Dj as well as regional aggregate capital and labour stocks, Kj and Hj . Then the role of EKj can be assessed by estimating regression (11.5) in Table 11.2. Using German regional data, Audretsch and Keilbach (2004a) estimated the effect of regional start-up rates on regional GDP, [α4 in (11.5)] to be positive and significant, and five times larger than the effect from regional R&D (α3 ). These results appear to be robust to various methodological refinements.12 It is noteworthy that technology- and knowledge-intensive enterprise start-up rates, EK, spur higher regional growth rates in Germany, whereas non-technology-based or non-knowledge-intensive start-up rates do not (Mueller, 2007). This suggests that the effects of entrepreneurship capital derive from knowledge spillovers arising from technology-intensive ventures. Furthermore, Acs et al. (2004) detect a prominent role for the interaction between entrepreneurship capital and R&D, rather than for R&D or entrepreneurship capital alone. Similarly to Mueller (2007), this implies that entrepreneurship does not affect growth unless it is associated with knowledge spillovers. Findings of significant spatial (regional) autocorrelation in entrepreneurship capital (Audretsch and Keilbach, 2007b) can be taken to imply the existence of geographical spillovers of EK to neighbouring regions. Whether this requires a focus on national rather than local policies to promote entrepreneurship is open to debate. Current research on knowledge spillovers and entrepreneurship capital can be criticised for being overly reduced-form in nature. The deep structural factors connecting entrepreneurship with economic growth remain hidden. Some authors have tried to address this problem, though with mixed success. For example, Audretsch, Bönte and Keilbach (2008) estimate an aggregate model linking innovation efforts, technical knowledge, entrepreneurship capital and economic performance as endogenous variables. Their findings suggest that innovative activities promote regional economic performance in Germany both directly and also indirectly through increases in entrepreneurship capital. However, this article does not estimate structural parameters relating to agents’ preferences and constraints. Nor does other work, by Berkowitz and DeJong (2003, 2005), who estimate structural equations for regional economic growth and the number of small firms, and report a significant positive relationship between the two.13 Their estimates suggest that an additional 2.3 legally

328

Performance

registered enterprises per 1, 000 inhabitants in post-Soviet Russia are associated with an increase of 2.0–3.4 percentage points in real rates of regional economic growth. Contrary to these rather positive findings, it should be noted that in some economies where competition is limited, entrepreneurship can actually retard growth – especially when entrepreneurship takes the form of family businesses (Bertrand and Schoar, 2006). Nepotism, a selfish desire to turn the business into a family legacy, and rigid inheritance norms can lead to poor decision-making which inhibits the growth and survival of family firms. To the extent that family firms play an important role in the economy, this can translate into slower growth and inhibit development at the aggregate level. Bertrand and Schoar (2006) provide some simple cross-country regression evidence based on four waves of the World Values Survey, which shows that countries in which the institution of the family is generally regarded as more important than the average have below-average rates of GDP per capita. In a similar vein, and using the same data set, Morck and Yeung (2004) add intriguing evidence that countries in which large firms are controlled by powerful mercantile families not only have a lower average material standard of living, but also provide worse infrastructure, health care and education; implement more irresponsible macro-economic policies; and are more inegalitarian. Morck and Yeung (2004) attribute these dismal outcomes to the strong political connections enjoyed by powerful family firm owners, who engage in ‘rent-seeking’ behaviour (‘destructive entrepreneurship’: see chapter 12) which reduces economic growth. This implies that entrepreneurship can be negatively associated with economic growth. The evidence for developed nations conducted at the highest level of aggregation (the nation state) is of a different nature. Some nation-level studies use the number of businesses as a proxy explanatory variable for entrepreneurship.14 Others use as explanatory variables one or more of GEM’s measures of ‘total entrepreneurial activity’ (TEA), defined as the proportion of a nation’s workforce either engaged in starting a business or operating one which is less than forty-two months old. Estimates derived using GEM data from thirty-six developed and developing countries find a positive association between TEA and economic growth, which is strongest in developed countries (van Stel, Carree and Thurik, 2005).15 This might be because there is more ‘high potential’ innovative entrepreneurship in developed economies which stimulates growth (Wong et al., 2005). Alternatively, growing markets can exhibit less vigorous competition and support more emerging niches (Hause and du Rietz, 1984). Despite van Stel, Carree and Thurik’s (2005) use of lags, however, the possibility of omitted and unobserved variables affecting both TEA and economic growth means that one cannot blithely interpret these results in a causal sense. An alternative measure of entrepreneurship that is sometimes used in growth regressions is the share of small firms in the nation’s stock of businesses. Audretsch et al. (2002) report a significant positive association between national GDP growth rates in Europe over 1990–94 and growth in the small firm share of value-added relative to large firms. Macro panel-data regressions from US states 1986–95 also uncover a positive relationship between rates of state productivity growth and the employment share of small businesses (Robbins et al., 2000).

Entrepreneurship and growth

329

Another popular empirical strategy is to run regressions linking GDP, GDP per capita, or GDP growth rates with aggregate national self-employment or business ownership rates, sometimes including fixed ‘country’effects (see chapter 3, section 3.5). Examples of this approach, which are usually based on OECD data, abound.16 Most of these studies report statistically significant positive associations between economic growth rates and self-employment rates. On a variation of this theme, several researchers suggest there exists an ‘optimal’level of entrepreneurship, s∗ , that maximises economic growth.17 Too little entrepreneurship, s, reduces competition, diversity and learning, while too much entrepreneurship reduces economies of scale. These studies suggest not a simple monotonic positive relationship between s and g but instead a non-linear function g(·) (see (11.7) in Table 11.2, where growth is increasing in the first argument and decreasing in the second). An increase in s does not necessarily increase economic growth: whether it does so also depends on s relative to s∗ . Empirical estimates of a linearised form of (11.7) with additional control variables suggests that on average, a 5point deviation in |s − s∗ | implies a growth loss of 3 per cent over a period of four years; and higher values of |s−s∗ | reduce subsequent levels of s in a gradual ‘error-correcting’ fashion (Carree et al., 2002). Obviously, this estimation and inference depends on correct estimation of s∗ , which is something researchers adopting this approach assume they have done correctly. I have spent some time analysing a range of different approaches to measuring entrepreneurship–growth relationships. As we have seen, published research on this topic generally claims to find evidence that entrepreneurship enhances economic growth. This claim is made at the level of the individual firm (smaller firms grow faster than larger ones); at the level of the industry and region (industries and regions with larger shares of small firms or new entrants grow faster); and at the level of the national economy (the standard of living grows faster in more entrepreneurial economies). However, more than is usual in applied entrepreneurship research, the results from these studies must be taken with a generous pinch of salt. They are all vulnerable to unquantifiable problems of omitted variable bias, aggregation bias and endogeneity bias, their occasional use of time lags notwithstanding. Most of the studies in this area have weak or non-existent micro-foundations, and rely on reduced-form econometric equations which lack a structural interpretation. One is also left wondering whether ‘publication bias’ has silently militated against the reporting of insignificant relationships between rates of entrepreneurship and economic growth. A causal interpretation of a growth–entrepreneurship relationship is especially problematic because ample evidence suggests that causality might run the other way, from economic growth to entrepreneurship. For example, economic growth appears to have a strong effect on regional firm formation rates,18 while at the national level, business ownership rates are sometimes expressed as a function of countries’ states of development.19 Some authors refine this argument and claim that entrepreneurship does not possess a simple monotonic decreasing relationship with GDP per capita, but is L (or U) shaped, falling as GDP per capita rises and then levelling off or rising at high levels because advanced technology is complementary with entrepreneurship.20 Thus

330

Performance

Wennekers et al. (2005, p. 306) estimate that entrepreneurship is negatively related to national income but positively related to its square; they go on to interpret this in terms of ‘necessity’ entrepreneurship declining as GDP rises from low levels and ‘opportunity’ entrepreneurship increasing as GDP rises at high levels. Future research can improve on the current state of knowledge and practice in several ways. It should provide firmer micro-foundations for econometric analysis and discriminate between different sources of growth, rather than merely conjecturing what might underlie observed aggregate relationships. These are unlikely to be easy tasks. 11.4 Appendix

Table 11.2. Useful empirical growth models Name

Equation

Eq. #

Reference

Gibrat’s Law (GL)

ln qi t = β + ln qi t−1 + ui t

(11.1)

Gibrat (1931)

Generalised GL (1)

ln qi t = βi + γ ln qi t−1 + ui t

(11.2)

Petrunia (2008)

Generalised GL (2)

ln qi t − ln qi t−1 = β + γ1 ln qi t−1 +γ2 [ln qi t−1 ]2 + γ3 ln ai t−1 +γ4 [ln ai t−1 ]2 + γ5 ln qi t−1 ln ai t−1 +ui t

(11.3)

Brock and Evans (1986, chap. 6)

‘Standard form’

gi = γ
Xi + ui

(11.4)

Entrepreneurship capital

ln GDPj = α0 + α1 ln Kj + α2 ln Hj +α3 ln(R&Dj ) + α4 ln EKj + uj

(11.5)

Audretsch and Keilbach (2004a, 2004b)

Cross-country

gjt = β0 + β1 TEAjt−1 +β2 [TEAjt−1 × Yjt−1 ] + Xjt
γ + ujt

(11.6)

Van Stel et al. (2005)

(11.7)

Carree et al. (2002, 2007)

Optimal rate

g = g(s, |s − s∗ |)

Notes: qit denotes a measure of size (e.g. number of employees, turnover or assets) of firm i at time t. In all cases, ui t is a mean-zero stochastic disturbance term with a distribution that is identical across i and t.

Notes 1. This implies βi = β and γ < 1 in (11.2) in Table 11.2. 2. Alternative models specify firm-specific productivity levels that follow a particular stochastic process (Hopenhayn, 1992); or model endogenous risk-taking and endogenous borrowing constraints vis-à-vis optimal financial contracting to obtain similar results (see e.g. Hopenhayn and Vereshchagina, 2003; Clementi and Hopenhayn, 2006). 3. See Storey (1994a), Davidsson, Achtenhagen and Naldi (2006) and Delmar and Wiklund (2008). Wiklund and Shepherd (2003) report that education and experience impact growth most for those whose growth aspirations are also high. 4. See Brock and Evans (1986), Evans (1987a, 1987b), Hall (1987), Cooper et al. (1989), Dunne et al. (1989a), Audretsch (1991), Davis and Haltiwanger (1992), Variyam and Kraybill (1992), Reid (1993), Mata (1994), Barkham, Hart and Hanvey (1996), Hart and Oulton (1996), Almus

Entrepreneurship and growth

5. 6.

7. 8. 9.

10.

11. 12.

13.

331

and Nerlinger (1999), Audretsch et al. (1999), Goedhuys and Sleuwaegen (2000), Akoten et al. (2006), Moreno and Casillas (2007) and Petrunia (2008). Geroski (1995), Sutton (1997) and Audretsch et al. (2004) review the early literature. A good case study illustration of this point is afforded by the London Business School case Regus (A). See Davidsson et al. (1994, 1995), Reynolds (1999) and OECD (2003). OECD (2003) attributes 30 per cent of total productivity growth in OECD countries to entries and exits. Audretsch and Fritsch (1996, 2002) uncover different results relating to turbulence in Germany in different decades. Although others have observed a U-shaped relationship between average firm size and employment growth at the industry and regional levels (Johansson, 2005; Shaffer, 2006). Audretsch et al. (2002), Acs and Armington (2006) and Carree and Thurik (2006). Some observers have suggested reasons such as deregulation and privatisation (Phillips, 1985). Others have speculated that large firms are increasingly concentrating on their ‘core competencies’ leading to spin-offs and divestments (Carlsson, 1989). Alternatively, growing incomes and globalisation have increased the demand for variety in consumption goods, which small firms are better placed to supply (Jackson, 1984; Carree and Thurik, 2006). But none of these conjectures relates directly to the theories of growth and entrepreneurship set out in chapter 2. Most studies focus on European countries: see Wagner (1994), Audretsch and Fritsch (2002), Fritsch and Mueller (2004, 2007, 2008), van Stel and Storey (2004), Baptista et al. (2008), Carod et al. (2008) and van Stel and Suddle (2008). There are also US studies: see Acs and Armington (2004b, 2006) and Acs and Mueller (2008). Audretsch and Fritsch (2002) estimated that German start-up rates in the 1980s had larger positive effects on German regional employment growth in the late-1990s than early-1990s start-up rates did. Thurik et al. (2008) provide international evidence of long lags (around eight years) in the self-employment-unemployment relationship. See Audretsch and Keilbach (2004a, 2004b, 2007b), Audretsch (2007b) and Audretsch, Bönte and Keilbach (2008). These include panel-data fixed effects (see chapter 3); university–industry relationships as measured by grants from industry to university researchers (Mueller, 2006a); use of GDP growth rather than GDP in levels as a dependent variable (Audretsch, 2007b); the inclusion of spatial autocorrelation (Audretsch and Keilbach, 2007b); and systems estimation (Audretsch, Bönte and Keilbach, 2008). Berkowitz and DeJong measured entrepreneurship in region j, denoted by sj , as the number of legally registered small private enterprises in j. Denoting the economic growth rate in region j by gj , Berkowitz and DeJong estimated the following recursive two-equation model by 2SLS: sj = α0 + α1 ’Xj + α2 Zj + u1j

(11.1)

gj = β0 + β1 Xj + γ sj + u2j .

(11.2)

Here, Xj are a set of covariates which affect both endogenous variables, while Zj is a set of identifying instruments which affect sj but not gj . The recursive structure of (11.1) and (11.2) is assumed and full simultaneity is not tested. Berkowitz and DeJong (2005) reported estimates of γ which were significant, positive and substantial: see the text. 14. This measure doubles as both a measure of entrepreneurship and a measure of competition which can promote total factor productivity and hence economic growth (Nickell, 1996; Nickell et al., 1997). 15. Van Stel et al. (2005) regressed growth of GDP per capita on lagged TEA, as well as lagged TEA interacted with national GNP per capita, Y . Van Stel et al.’s (2005) pooled time-series crosssection econometric model is given as (11.6) in Table 11.2. Positive estimates βˆ1 and βˆ2 in (11.6)

332

16.

17. 18. 19. 20.

Performance were reported, suggesting that TEA does cause economic growth, and that the size of the effect depends on the level of per capita income, being greater in richer, i.e. developed, countries. See e.g. Audretsch and Acs (1994), Dennis (1996), Robson (1996, 1998b), Thurik (1999), Blanchflower (2000), Audretsch and Thurik (2000, 2001b), Braunerhjelm and Borgman (2004), Acs et al. (2004), Parker and Robson (2004) and Acs and Armington (2006). Carree et al. (2002, 2007), van Stel and Carree (2004) and Martinez (2005). See e.g. Audretsch and Fritsch (1994a), Keeble and Walker (1994), Reynolds (1994), Carree and Thurik (1996) and Audretsch and Keilbach (2007a). See e.g. Acs, Audretsch and Evans (1994), Schultz (1990) and Fölster (2002). See e.g. Carree et al. (2002), Wennekers et al. (2005) and Verheul et al. (2006).

12 Entrepreneurial effort

An important determinant of entrepreneurial performance is the effort that entrepreneurs devote to their ventures. Effort is needed as a productive input to help ventures survive and grow. Because expenditure of effort is typically costly, entrepreneurs must decide carefully how much effort to supply. Entrepreneurs must also decide whether to expend their effort on productive, socially unproductive – or even destructive – activities. Baumol (1990, 1993a) famously distinguished between productive and unproductive entrepreneurship. Baumol’s central premise is that although the entrepreneurial spirit is fairly uniformly distributed across different places and periods of time, institutions and cultural norms – which describe the ‘rules of the game’ – can either encourage entrepreneurs to engage in socially productive activities or give them incentives to engage in socially unproductive or destructive behaviour by generating profits at the expense of others. Baumol (1990, 1993a) cites several examples of unproductive and destructive entrepreneurship from the ancient world, early China and medieval Europe. He observes that the materially unproductive activity of abstract puzzle solving attracted greater social status in medieval China than wealth creation did. In the same period, enterprising but destructive Europeans sought booty through wars, banditry and piracy. The allocation of effort to manipulate the economic environment rather than to engage in trade and the production of wealth goes by the name of ‘rent seeking’. The first section of this chapter explains how modern-day entrepreneurs also respond to incentives to devote effort to rent-seeking activities, and analyses the economic implications of this behaviour. It turns out that effective enforcement of well-defined property rights is essential for encouraging entrepreneurs to allocate their effort to productive rather than to unproductive or destructive activities. The sections which follow explore effort devoted to productive entrepreneurship. The second section analyses how much work effort entrepreneurs devote to their ventures, and the determinants of that effort. Previous analytical studies of entrepreneurs’ work effort fall into two broad categories. One strand of research describes how entrepreneurs allocate their time between the various tasks involved in running a venture. Not surprisingly, these descriptive studies find differences between tasks performed at various stages of venture development. Early-stage entrepreneurs chiefly focus on customers 333

334

Performance

and sales, and engage in little or no delegation, while later-stage entrepreneurs specialise and focus on administrative and strategic challenges.1 I will instead explore a second theme, which tries to explain the chosen quantity of work hours.2 What is to be gained from studying the determinants of entrepreneurs’ work effort? One answer is a better understanding of some aspects of public policy. For example, it will be seen below that on average entrepreneurs work longer hours than employees – so policies designed to promote entrepreneurship should take account of this when targeting women, for example, who tend to have more onerous domestic commitments than males (Hundley, 2001a; and see chapter 6). Alternatively, if entrepreneurs work harder in response to larger expected incomes, then there might be grounds to support income tax cuts as part of a social-welfare-enhancing entrepreneurship promotion policy (Parker, 1999a). However, one certainly cannot simply assume that entrepreneurs will always supply longer hours in response to greater returns to their hourly effort. This consideration also motivates the discussion below. The third section of the chapter deals with ageing and the retirement decision of entrepreneurs. This is an especially topical issue in view of the ongoing debate about promoting entrepreneurship among retired employees, as a way of jointly stimulating output and relieving pressure on national public pension systems. Finally, the last section of the chapter treats a particular form of effort associated with entrepreneurial learning.

12.1

Unproductive and destructive entrepreneurship

12.1.1 Conceptual issues When politicians and commentators eulogise entrepreneurship, they usually have in mind the creation and growth of businesses in a benign market environment characterised by free and fair competition; unambiguous, unassailable and well-protected property rights; and transparent and impartial legal institutions which uphold those rights and treat all contracting parties fairly. Well-protected property rights help make entrepreneurship and innovation attractive. Secure property rights prevent ad hoc expropriation of the fruits of entrepreneurship, which is usually necessary to make costly investments worthwhile. A stable legal framework promotes planning, resource acquisition and co-ordination, all of which take time to put in place. In practice, some economies lack the institutions needed to: oversee free competition; uphold property rights; and enforce contracts by guaranteeing fair resolutions of legal disputes. In such cases, entrepreneurs can have incentives to apply their talents and effort to restrict competition rather than to create new wealth. When this happens, we talk about entrepreneurs engaging in ‘rent seeking’ activities. In these cases, entrepreneurs can make private gains at the expense of others, to the detriment of overall social welfare. For example, elementary microeconomic theory shows that oligopolists seeking to maximise profits have incentives to collude where possible and to restrict competition. In that way, the producers all make greater profits at the expense of the consumer.

Entrepreneurial effort

335

The discussion below will focus on two types of rent-seeking behaviour: the purposeful restriction of competition by incumbents, and the appropriation of payoffs by entrepreneurs and government officials, usually in the form of mafia-style extortion and bribes, respectively. Both types of activity impact adversely on entrepreneurship. Holmes and Schmitz (2001) present several examples from a wide range of industries showing how entrepreneurs have used the regulatory process to block rivals in order to limit competition. Examples of such behaviour include small and inefficient banks lobbying for restrictions on bank branching in US states, in attempts to restrict competition; efforts by small European and Japanese retailers to lobby for legal restrictions on maximum store floor space; incumbent trucking companies exercising undue influence by persuading the US Interstate Commerce Commission to deny licences to new competitors; and pressure by trade unions and small firms to change city building codes to restrict the spread of labour-saving innovations in the US construction industry. Holmes and Schmitz (2001) cite evidence showing that these efforts have all reduced output and productivity. In the context of international trade, Holmes and Schmitz (2001) argue that reducing tariffs can force entrepreneurs to redirect resources from unproductive to productive entrepreneurship. They conceive of productive entrepreneurship as involving the use of resources to innovate and profit from selling a superior product. Unproductive entrepreneurship, on the other hand, involves expending resources to block the potential innovations of domestic rivals, for example by using a country’s regulatory and legal processes to restrict competition. To see the benefits of free trade in this context, suppose that initially a domestic monopolist is protected by high tariffs. The monopolist has an incentive to engage in blocking activities to protect its supernormal profits (see also Grossman and Helpman, 1991). This entrepreneur might be using an inefficient technology relative to entrepreneurs in other countries, but that does not concern them because tariffs deter or prevent domestic consumers from accessing the foreign alternative. But if domestic tariffs on imports are reduced, domestic consumers can now buy foreign goods, and entrepreneurs can no longer gain from blocking domestic rivals. Instead, they have to innovate – shifting resources away from blocking towards productive entrepreneurship – in order to compete with foreign rivals. Note that a country can unilaterally reduce domestic tariffs and still realise net benefits. Naturally, the benefits are even greater if foreign countries reduce their tariffs too, because by innovating the domestic monopolist will be rewarded with the prospect of substantial exports in the global market, further increasing their incentive to reallocate effort from unproductive to productive entrepreneurship. An indirect implication of Holmes and Schmitz’s (2001) argument is that globalisation – encapsulated in cross-border trade and competition, and extended by the World Trade Organisation’s tariff reduction initiatives – can increase productivity, and hence average living standards, by replacing unproductive with productive entrepreneurship. However, this argument has still not prevented incumbents from lobbying their governments to introduce legal regulations which affect their competitors. Consider, for example, regulations which impose high fixed costs on all firms, including their own,

336

Performance

Payoff

␤⬘ ␣ ␥

Entrepreneurs

␤ Rent-seekers

n* = (␣⫺␥)/ ␤

n**

No. of rent-seekers

Figure 12.1 Payoffs to entrepreneurship and rent-seeking in Murphy et al.’s (1993) model

such as administrative compliance costs relating to banking and pensions. Because larger firms can spread these costs over greater output, they can thereby put their smaller rivals at a competitive disadvantage and increase their profits (see chapter 17). The prevention of entry by new firms, even in different industries from those of incumbent entrepreneurs, also benefit the latter by reducing aggregate labour demand and hence the wages that the incumbents have to pay their employees.3 Another form of rent-seeking behaviour involves direct appropriation of surplus from productive individuals. Murphy et al. (1993) provide an illuminating analysis of this problem in a model of free occupational choice between three occupations. These occupations are entrepreneurship, which generates a payoff of α per person; subsistence agriculture, which generates a payoff of γ < α per person; and rent-seeking (extortion), which offers a maximum payoff of β per person. Entrepreneurship returns can be appropriated by rent-seekers, so if there are n rent-seekers, returns in entrepreneurship fall to α − nβ. For simplicity, suppose that returns from subsistence agriculture cannot be appropriated by rent-seekers. This situation is illustrated in Figure 12.1, where payoffs in entrepreneurship and rent-seeking are plotted (as bold lines) against the number of rent-seekers. Figure 12.1 shows that once the number of rent-seekers reaches n∗ = (α − γ )/β, entrepreneurs rationally switch out of entrepreneurship and into subsistence agriculture, where they earn γ . As the number of rent-seekers rises beyond n∗ , rent-seekers begin to crowd each other out, since there is a fixed pie to divide between them all. This reduces their returns to (α − γ )/n < β. In the case with the two bold lines illustrated in Figure 12.1, there is a simple equilibrium. Everyone does best in productive entrepreneurship because α > γ > β. There is therefore no rent-seeking or subsistence agriculture in this equilibrium. One can think of this as a case where property rights are well protected, since β is lower even than subsistence returns γ . However, if property rights are very poorly protected,

Entrepreneurial effort

337

so β is replaced by a higher value β
> α, then the equilibrium changes. Rent-seekers’ return function is now given by the dotted line in Figure 12.1 and there is now a unique equilibrium with numerous rent-seekers n∗∗ > n∗ . In this economy, everyone earns the subsistence income γ , which is less than entrepreneurial output, α. Rent-seeking has therefore reduced productivity and living standards.4 Taking β as an inverse measure of property rights protection, higher values of this parameter clearly do not affect technological output possibilities, but they do help determine which of the above cases obtains. A lower value of β increases the prospects of the good equilibrium and decreases the prospects of the bad one. If an economy starts out in the bad equilibrium, it needs to be jump-started out of it, by improving property rights protection or by promulgating anti-rent-seeking ideology to ensure that β falls below γ (Murphy et al., 1993). This improvement needs to be radical because the equilibria discussed above are stable and by inspection of the second equilibrium in Figure 12.1 (the dotted rent-seeking payoff), living standards will be unchanged (at γ ) by minor improvements in property rights. This might explain why it is so difficult for economies characterised by rent-seeking to rid themselves of this drain on their productivity. Continuing with the theme of entrepreneurs as rent-seekers, Morck and Yeung (2004) argue that in many countries, including developed ones, oligarchic entrepreneurs running family businesses find it easier than other economic agents to engage in successful rent-seeking practices. Oligarchic entrepreneurs can build trust with officials to gain favours because of their longevity, small number, ability to pre-commit to outcomes, discretion, and power to punish. They sometimes also have blood ties to senior political figures. This can culminate in ‘a positive feedback trap where oligarchic family control, political rent seeking, and poverty all perpetuate each other’ (Morck and Yeung, 2004, p. 404). Note that it is not an absence of trust in society that is responsible for this welfare loss, but rather the inappropriate presence of trust between entrepreneurs and officials which lies at the heart of the problem (see also Bertrand and Schoar, 2006). A further problem is that once oligarchic family firms are entrenched, they can be hard to remove. Not only can they restrict entry by lobbying for formal entry barriers, but as wealthy financiers in their own right they can use their influence in the credit market to withhold finance from potential rivals. In different circumstances, entrepreneurs can be preyed upon by non-entrepreneur rent-seekers. Corrupt officials can prosper by using the power of the state to enable favoured entrepreneurs to earn rents and block new competitors. In return, the officials tax part of the favoured entrepreneurs’ monopoly profits via bribes. The officials win from this state of affairs; would-be entrepreneurs and the general public unambiguously lose; and whether the favoured entrepreneurs become better or worse off depends upon the size of the bribes demanded. If government officials make enough money from rent-seeking and if rent-seeking activities pervade all areas of government, it can become well nigh impossible to root out corruption, even by well-intentioned politicians (Bardhan, 1997; Foellmi and Oechslin, 2007). One solution might be to rotate bureaucrats across different jobs and to reduce the number of bureaucrats

338

Performance

needed to approve new entrepreneurial projects (Lambert-Mogiliansky et al., 2007). Credit market imperfections can also generate rents which officials expropriate via bribes. As above, the least wealthy entrepreneurs lose if bribes are so high that they deter entry to all but the wealthiest who can afford them. At the same time, restricted entry reduces demand for capital and hence the interest rate, benefiting the richest entrepreneurs. This private gain can even outweigh for rich entrepreneurs the higher costs of the bribes they have to pay. In a general discussion of institutional frameworks, rent-seeking and entrepreneurship, Frye and Shleifer (1997) distinguish between the ‘invisible hand’ of a benevolent state that guarantees property rights but minimises intervention in markets, and the ‘grabbing hand’ of a state in which government does not enforce property rights rigorously but intervenes via corrupt officials who can expropriate value using government power. Anticipating a grabbing hand, individuals are discouraged from becoming entrepreneurs. The problem can be worsened if governments tax only productive output, for then rent-seeking and the output loss it causes worsens, and becomes an increasing function of the tax rate (Lu, 1994).

12.1.2 Evidence There is a wide array of evidence about the incidence of rent-seeking and its adverse effects on the economy. What follows is meant to be an illustrative rather than an exhaustive overview of the literature. The emphasis will be on the interface of rentseeking with entrepreneurship specifically, rather than an analysis of corruption more generally. The transition economies of central and eastern Europe have been widely studied with respect to the security of property rights vis-à-vis entrepreneurship. For example, Johnson et al. (2000, 2002) measured the security of property rights in several east European countries in terms of individuals’ perceptions of the integrity of property rights and the effectiveness of courts. A central question is whether weak property rights affect entrepreneurial behaviour. Johnson et al. (2002) estimated that entrepreneurs who perceived their property rights to be the least secure re-invested only 32 per cent of their profits, whereas entrepreneurs perceiving property rights to be the most secure reinvested 56 per cent of their profits. Re-investment behaviour is crucial because it drives employment and sales growth in these countries – and hence aggregate economic growth. Johnson et al. (2002) also reported that established firms earn higher profits in countries with weak property rights and ineffective courts. Since high profits should be eroded by the entry of competitors, these institutional deficits seem to be discouraging entry, too. In contrast, Johnson et al. (2002) found little evidence of limited access to external entrepreneurial finance in the countries they studied. This suggests that policies designed to ease financial constraints will make only limited improvements in places where courts are used infrequently, and where governments absorb entrepreneurs’

Entrepreneurial effort

339

time and money in bribes, taxes and greater costs of doing business. An important policy implication emerging from Johnson et al.’s work is that secure property rights (including the control of corruption) is a necessary condition for entrepreneurship to thrive. Additional evidence supports this view. Frye and Shleifer (1997) observe that small enterprises in Warsaw are more productive than their counterparts in Moscow partly because their tax and regulatory burden is lower. Djankov, Miguel et al. (2005) report that perceptions of corruption and an unfavourable attitude towards entrepreneurship among government officials and the general populace decrease both the probability that Russians become entrepreneurs, and the length of time they spend in entrepreneurship. Privatisation of state assets has been a common feature of the eastern European transition economies, and it has also provided ample opportunities for rent-seeking behaviour. Berkowitz and Holland (2001) claim that local small-scale privatisations in Russia over 1996–99 deterred small enterprise formation, whereas large-scale privatisations conducted at the federal level encouraged it. Their reasoning was that local privatisations may have enabled business insiders to form alliances with local institutions, who lobbied to block the reform process at the local level. This is politically much harder to do at the national level, where privatisation might have increased the transparency of the regulatory environment, thereby enhancing small enterprise formation. Ideally, governments should not only tackle impediments to investment such as corruption and mafia groups, but also should implement responsible macro-economic policies. McMillan and Woodruff (2002) argue that governments in transition economies differ markedly with respect to their choices of enterprise-friendly microeconomic and macroeconomic policies – which in turn have a substantial impact on the spread of entrepreneurship in these economies. This issue is all the more important because it seems that the success or failure of a transition economy can largely be explained in terms of the performance of its entrepreneurs. In many of these countries, new firms create the bulk of all new employment – much more so than incumbent state firms or their newly privatised counterparts, which often shed jobs (Bilsen and Konings, 1998). If anything, developing countries are even more prone to rent-seeking than the transition economies. Parente and Prescott (1999) argue that monopoly rights enjoyed by coalitions of input suppliers and sustained by government approval keep poor countries poor by forcing producers to use inefficient technologies. Parente and Prescott (1999) estimate that eliminating this form of monopoly power could increase output threefold in these countries without any change in inputs. Separately, and consistent with the earlier discussion about the influence of powerful oligarchic families in financial markets, there tends to be less investment in innovation in countries where inherited plutocratic wealth is a large fraction of GDP (Fogel et al., 2006). These are invariably poor developing countries. But property rights and the efficiency of courts matter in developed countries too. Property rights protection can be a double-edged sword, as it can protect powerful oligarchic entrepreneurs who use the resulting profits to successfully block market entry by others. Democracies might be able to circumvent this problem by dispersing

340

Performance

power to reduce the influence of oligarchic entrepreneurs. Yet democracy is prone to the drawback of populist governments imposing punitive redistributive taxes on entrepreneurs. Arguably, there is a trade-off between democratic and oligarchic systems with respect to entrepreneurship and growth (Acemoglu, 2008). When it comes to the specific question of addressing credit misallocation caused by imperfect property rights protection, policy recommendations can be subtle. On the one hand, Jappelli et al. (2005) argue that improvements in judicial efficiency in enforcing debt contracts when borrowers can default opportunistically are likely to reduce credit constraints and increase lending. The reason is that, by increasing borrowers’ incentives to repay, creditors are more likely to extend credit in the first place. On the other hand, Zazzaro (2005) points out that stricter enforcement of debt contracts under conditions of asymmetric information might actually worsen credit allocation to entrepreneurs, as it weakens banks’ incentives to screen borrowers. In contrast, improvements in accounting standards (e.g. tightening accounting and reporting rules) reduce the costs of screening and hence improve both credit allocation and social welfare. Mirroring the different theoretical positions on this issue, the available evidence is mixed. Using panel data from ninety-five Italian provinces, Jappelli et al. (2005) found that significantly fewer business loans were extended in provinces with longer trials or larger backlogs of pending trials. Judicial efficiency correlates positively with the volume of business lending and negatively with proxies for credit constraints. Using a different sample of Italian data, however, Guiso et al. (2004) report that a measure of judicial inefficiency, namely the number of years it takes to reach a firstdegree judgement in the province, has a negative but insignificant effect on Italian firm-creation rates. It would seem that the way judicial efficiency is measured affects empirical estimates of its impact on entrepreneurship. At a more aggregate level, cross-country regression evidence suggests that countries with better institutions (e.g. more secure property rights, a tradition of contract enforcement, and fair judiciaries and governments) enjoy greater venture capital investment per capita, more patents per capita, faster rates of sole proprietorship growth and higher establishment growth rates (Sobel, 2008).5 Furthermore, international trade restrictions such as tariffs deter entry (Sobel et al., 2005). One possible public policy response might be to shrink the size of government. Rather contentiously, shrinkage might even extend to publicly funded pro-entrepreneurship programmes, if these programmes ‘encourage entrepreneurial individuals to devote effort toward figuring out how to obtain the transfers, rather than devoting those efforts toward satisfying consumers and creating wealth’ (Sobel, 2008, p. 653). More will be said about pro-entrepreneurship programmes in the final chapters of this book.

12.2 Work effort

The remainder of this chapter analyses productive entrepreneurship. The present section is divided into two parts. The first briefly outlines some basic facts about entrepreneurs’ hours of work. Here, entrepreneurs are usually measured as the self-employed. The

Entrepreneurial effort

341

second subsection then provides an extended discussion of the determinants of these work hours.

12.2.1

Hours of work Most self-employed Americans, Canadians and Britons work full-time rather than parttime.6 Defining ‘full-time’ to be thirty hours or more per week, 80–90 per cent of self-employed males work full-time in these countries, compared with about 50 per cent of their female counterparts. There is mixed evidence about whether the selfemployed are more or less likely than employees to work part-time (Aronson, 1991; Bregger, 1996). But it is fairly well established that self-employed males work longer per week on average than employees do (Carrington et al., 1996; Ajayi-obe and Parker, 2005); and self-employed employers work the greatest number of weekly hours of all (Curran and Burrows, 1989; Ajayi-obe and Parker, 2005). Table 12.1 summarises some illustrative findings from the UK. It can be seen that on average own-account self-employed males work nearly three more hours per week than employees do, while male self-employed employers work more than ten more hours per week than their employee counterparts. Similar patterns are observed for female employers versus employees, while own-account self-employed women work fewer hours per week on average, reflecting their greater tendency to practice selfemployment on a part-time basis. In a similar vein, detailed time-diary evidence from the Finnish Time Use Survey data from 1999/2000 points to an eight hours’ average difference between the work hours of Finnish self-employees and paid-employees (Hyytinen and Ruuskanen, 2007). About 20 per cent of this difference is attributable to weekend working by the self-employed, with the remaining 80 per cent reflecting a longer working week. Other evidence shows that self-employed Americans work a similar number of weeks per year to employees – except during recessions, when the self-employed reduce their weekly work hours while employees work fewer weeks (Carrington et al., 1996). According to these authors, annual work hours in the USA are no more cyclical for male self-employed workers than for male employees. As with employees, average weekly self-employed work hours have been declining over time in most OECD countries, and the work hours gap between the self-employed and employees has been narrowing in many of them.7 Part-time working is sometimes regarded as an indicator of labour market flexibility. Entrepreneurs could choose to venture part-time for several reasons. One is a wish to operate a lifestyle business on a small scale. Another is a wish to retain a current job in paid employment until the entrepreneur has a better idea about whether their start-up effort is viable. In the presence of sunk costs, for example, part-time entry might be a useful way to manage risk. Swedish evidence from Wennberg et al. (2007) supports this idea, with part-time self-employees being especially likely to transition into full-time self-employment. Multiple job-holding can be regarded as another indicator of labour market flexibility. Many entrepreneurs having second jobs in paid employment do so at the start of their

342

Performance

Table 12.1. Average work hours in Britain by gender and occupation

Employees

Self-Employed: Own Account

Self-Employed: Employers

Year

Male

Female

Male

Female

Male

Female

1991

44.22 (13.03) [2448]

30.92 (13.59) [2503]

47.36 (17.25) [354]

27.47 (18.25) [130]

56.04 (16.95) [167]

47.24 (21.16) [62]

1995

44.27 (12.16) [2137]

31.50 (14.08) [2336]

47.04 (17.60) [315]

27.86 (15.58) [118]

56.25 (15.73) [137]

40.53 (16.35) [55]

2000

43.80 (11.71) [3702]

32.11 (13.14) [3938]

46.72 (16.67) [471]

27.33 (16.81) [218]

54.55 (17.27) [205]

42.51 (19.60) [91]

Notes: Standard deviations are in parentheses, and numbers of observations in square brackets. Data: BHPS, 1991, 1995, and 2000. Source: Ajayi-obe and Parker (2005).

spell in entrepreneurship. Lèvesque and Minniti (2006) mention that the owners of America’s fastest-growing private sector firms listed in the Inc. 500 magazine retained their wage jobs for an average of four months after starting up their new ventures. And using the Russian Longitudinal Monitoring Survey, Guariglia and Kim (2006) report that Russian workers who moonlighted (i.e. had a second job) as self-employed represented 26.5 per cent of entrants to full-time self-employment. These results may suggest a positive role for moonlighting as an incubator for new business ideas, or as a means of acquiring the necessary resources for a move into full-time entrepreneurship. In fact, although GEM data show that most nascent entrepreneurs hold another job while they organise their start-ups, only a minority of individuals who make a full transition to entrepreneurship continue to hold multiple jobs. British data taken from the 1987/88 UK LFS indicate that only 5 per cent of established self-employees had second jobs (60 per cent of which were also in self-employment: Hakim, 1988). US evidence also points to a small, static, role for multiple job-holding in self-employment (Aronson, 1991). Recent British evidence suggests that multiple job-holding is declining among the self-employed. According to BHPS data, the proportion of workers holding multiple jobs decreased from 8.36 per cent of employees and 11.42 per cent of the self-employed in 1991, to 7.20 per cent and 8.07 per cent by 2000, respectively (Ajayi-obe and Parker, 2005). The largest decreases in multiple job-holding were for self-employed women taking second jobs as employees (6.25 per cent in 1991, down to 2.92 per cent in 2000); and for self-employed men taking second jobs in self-employment (8.84 per cent in 1991, down to 5.31 per cent in 2000). It is unclear as yet whether these findings reflect

Entrepreneurial effort

343

q(h) = payoff in entrepreneurship

Payoff

wh = payoff in paid employment

0

h*

H*

Entrepreneurial work hours, h

Figure 12.2 Work mixing in entrepreneurship and paid employment

a secular trend or just the effects of the business cycle, as 1991 was a recession and 2000 a boom year. Why do so few entrepreneurs mix work time between entrepreneurship and paid employment? The occupational choice model of entrepreneurship under risk where time can be apportioned between multiple occupations (Parker, 1997a) suggests that work mixing can help entrepreneurs to diversify risk. It turns out that a necessary condition for work mixing is an entrepreneurial production function which is concave in entrepreneur effort, h (Lévesque and MacCrimmon, 1997). This is illustrated in Figure 12.2, where the paid employment wage w is fixed but the entrepreneurial ‘wage’ [given by the gradient of the production function q(h)] declines in h, owing to diminishing marginal returns to effort. In this case, risk-neutral income-maximising entrepreneurs optimally allocate h∗ hours of effort to entrepreneurship, while allocating H ∗ − h∗ hours to paid employment (given a total work effort of H ∗ ). It follows obviously that if q(h) is linear or strictly convex in h, entrepreneurs’ work effort ends up at a corner solution of either h∗ = 0 or h∗ = H ∗ , depending on the sign of q(H ∗ ) − wH ∗ . Of course, concavity is a necessary condition for work mixing, not a sufficient one. For example, steeply sloped concave q(h) functions which exceed wh for all h are also consistent with an all-ornothing choice, namely h∗ = H ∗ hours in entrepreneurship. As noted above, this is the modal outcome in developed economies. 12.2.2

Explaining entrepreneurs’ work hours

There is now an extensive literature explaining the labour supply behaviour of employees (see Blundell and MaCurdy, 1999, for a review). That literature is partly concerned with measuring the impact of wages on employees’ hours worked. This issue has obvious policy relevance because income taxes reduce wages and hence have potentially important labour disincentive effects.

344

Performance

It is natural to seek to extend this analytical framework to entrepreneurs. This exercise is likely to be especially valuable given the close relationship between entrepreneurs’ effort and output, especially at different stages of the business cycle (Rampini, 2004). However, only relatively few studies have attempted this task to date, in which entrepreneurs are invariably measured as the self-employed. Part of the reason for the limited analysis of this issue until now might be that sample sizes tend to be small for entrepreneurs compared with employees. But there are also intrinsic difficulties involved in analysing labour supply for entrepreneurs which will become apparent below. The discussion below will commence with a theoretical discussion of optimal workhour choices. Then some empirical findings will be presented. Much, but not all, of the empirical literature explores the effects of entrepreneurs’ wages on their work hours. Careful attention is paid in what follows to the precise definition and measurement of an entrepreneurial ‘wage’. One of the puzzles emerging from this body of research is the observation that many entrepreneurs work longer hours for lower hourly wages than employees do.8 Some possible resolutions of the puzzle are proposed in the course of the discussion. Standard theoretical labour supply models for employees (Blundell and MaCurdy, 1999) typically assume that the gross wage available to workers of a given ability is invariant to work-hours. (As we will see shortly, this is unlikely to be a tenable assumption for entrepreneurs.) Standard employee labour supply models decompose work-hours responses to wage changes into two components: a substitution effect and an income effect. The former describes how workers increase their work effort to exploit the higher return from working, holding spending power constant; the latter describes the effects of greater spending power on the work–leisure choice. The substitution effect is strictly positive in work-wage space; the income effect is ambiguous in sign, but is negative if individuals demand extra leisure as their income rises. If the substitution effect dominates the income effect, then work hours are increasing in the wage rate (i.e. upward sloping in wage-hours space). But if the income effect becomes dominant at higher wage rates, then an individual’s labour supply schedule can bend backwards, indicating that work hours eventually decline as wages rise. There are two broad categories of theoretical labour-supply model: static and life cycle. All the models which have been estimated for entrepreneurs have been, to the best of this author’s knowledge, of the static variety.9 Typical empirical specifications which are commonly estimated for a cross-section sample of individuals j = 1, . . . , n include the semi-log and double-log specifications hj = β0 + β1 ln wj + β2
Xj + vj

(12.1)

ln hj = β0 + β1 ln wj + β2
ln Xj + vj ,

(12.2)

where hj is j’s chosen work hours; wj is their wage; Xj is a vector of personal characteristics specific to j; and vj is a disturbance term. A practical advantage of (12.2) is that

Entrepreneurial effort

345

β1 measures the elasticity of work hours with respect to the wage; this is commonly referred to as the ‘uncompensated wage elasticity’. Several remarks are in order before discussing the estimation of labour-supply specifications like (12.1) and (12.2). First, it is common for these specifications to be estimated for entrepreneurs without any motivating discussion in terms of utility maximisation. While this is not a problem per se, it is perhaps indicative of an underdeveloped body of theory relating to entrepreneurs’ labour supply choices. Second, these specifications are very restrictive, as they assume that work-hours responses to wages are the same for all income levels and are also the same across all individuals.10 Third, the inclusion of leisure in the utility function [as in (12.12) in note 9] does not alter the standard economic conditions of market efficiency and equilibrium. In the long run, utility maximisation and profit maximisation converge for the marginal entrepreneur (Formby and Millner, 1985).11 There are several practical problems which researchers must address before estimating a work-hours model for a sample of entrepreneurs. Perhaps the biggest problem is that, unlike employees who face a fixed wage, the self-employed ‘wage’ must be imputed from profits and work hours. Hence the wage is not exogenous and needs to be instrumented (see chapter 3) with variables which are related to the wage but not the error term vj in an equation like (12.1) or (12.2).12 An endogenous wage is not the only problem which must be overcome when an empirical labour supply function is estimated for entrepreneurs. As will be explained in chapter 13, measurements of self-employed profits are prone to under-reporting biases and inclusion of returns to capital as well as labour.13 Furthermore, work hours are sometimes reported with error. One way of tackling measurement error is to use panel data, where observations are available across both entrepreneurs j and time t. The following specification of entrepreneur work hours is robust to unobserved heterogeneity of entrepreneurs (Parker, 1999b) and measurement error of work hours and incomes (Pissarides and Weber, 1989):14  ln h∗jt = β1  ln wjt∗ + β2
 ln Xjt + jt .

(12.3)

Here asterisks denote actual (and possibly mismeasured) data values. Ajayi-obe and Parker (2005) estimated this equation using Instrumental Variables (IV – q.v. chapter 3; and see the relevant entry in Table 12.2). An alternative to instrumenting entrepreneurs’ wages is to impute a ‘shadow’ wage. To see how, consider the wage that an entrepreneur j could earn in paid employment, wˆ j . This can be estimated using the methods outlined in chapter 3, and used in place of wj in (12.1) and (12.2). Feinberg (1987) used this approach, and argued that as wˆ j increases, entrepreneurs who can mix their work hours in entrepreneurship and paid employment should eventually devote less time to their enterprise and more to outside paid work. This implies that the uncompensated wage elasticity in entrepreneurship could turn negative.15 In practice, it is probably farmers in developing countries who can most easily mix time working in their primary business and as a hired hand for others. But here the evidence points to positive rather than negative wage elasticities.16

346

USA: NBER-TH, 1968; educated male self-employed with non-working wives; n = 281

USA: PSID, 1970–71; self-employed with annual non-labour income under $10,000; n = 68

UK: GHS cohort averages 1973–1986; self-employed males; n = 98

Feinberg (1987)

Rees and Shah (1994)

Data source and sample selection

Wales (1973)

Study

0 (insignificant)

OLS; R2 = 0.52

Semi-log

Declared work hours and computed log wage

Number of children under 5 (−); proportion of cohort married (+)

Non-white (+); construction industry (−)

0.23

SUR; R2 = 0.24

Linear

Hours of work; self-employed wage is imputed from hired managers’ regression

N/A (structural model)

−0.25 to 0.20

IV

Double log

Log of income to leisure ratio; net wage instrumented by hourly gross wage and annual non-labour income

Other significant variables

Uncompensated wage elasticity

Estimation method and R2

Functional form

Dependent variable and wage variable

Table 12.2. Estimates of the determinants of entrepreneurs’ work hours

347

USA: New York City cabdrivers 1988–1994; n = 1044 (maximum)

USA: physicians’ income survey, 1984–85; male physicians; n = 588

UK: BHPS, 1991–2000; working-age male self-employed with positive incomes and work hours; n = 1964

Camerer et al. (1997)

Thornton (1998)

Ajayi-obe and Parker (2005)

0.06

−0.04

NL2SLS

IV; R2 = 0.08

Non-linear

Double-log, first differences [equation (12.3)]

Hours worked; shadow wage imputed from a production function ( = marginal product); instrumented by assistant wage, age, and others First difference of log weekly work hours; change in log wage instrumented by age, education and length of current job spell

−1.31–0

IV; R2 = 0.15–0.48

Double-log

Hours worked; wage is revenue divided by hours worked, instrumented by summary statistics from wage distribution of other drivers

Non-labour income (−); employer status (+)

N/A (structural model)

Night shift dummy (−)

348

USA: PSID, 1968–1993; male self-employed working-age household heads; with at least three consecutive years of positive income data; n = 1379

Parker et al. (2005)

OLS and IV: R2 = 0.23

GLS-IV; R2 = 0.80

Linear

Double-log

Self-reported work hours per week

Log of annual work hours; log real wage instrumented by lagged wages and individual fixed effects

Estimation method and R2

Functional form

Dependent variable and wage variable

0 (insignificant)

N/A

Uncompensated wage elasticity

Income risk (+)

Entrepreneurs’ ownership share (+); firm size (+); age (+ but concave); equity (+); and wealth (−)

Other significant variables

Notes: n = sample size; IV = Instrumental Variables; SUR = Seemingly Unrelated Regression; NL2SLS = Non-linear 2-Stage Least Squares; GLS = Generalised Least Squares

USA: SCF, 1989–2001; self-employed members of households holding business equity; n = 3368

Data source and sample selection

Bitler et al. (2005)

Study

Table 12.2 (cont.)

Entrepreneurial effort

349

Table 12.2 summarises the main empirical findings relating to the determinants of entrepreneurs’ work hours. Most of these take entrepreneurs to be self-employed individuals working in a range of sectors, though some (e.g. Camerer et al., 1997, with taxi-drivers; and Thornton, 1998, with physicians) consider specific self-employed occupations. Table 12.2 omits studies based on farmers (see the previous note for references), who are certainly self-employed but who are less obviously regarded as entrepreneurs in the conventional sense. Table 12.2 reveals little agreement among the studies about the size of entrepreneurs’ uncompensated wage elasticity. But if attention is focused only on studies which employ IV estimation (see chapter 3) and utilise at least 1,000 observations, the evidence points to a negative elasticity. This is of course consistent with entrepreneurs working long hours for low hourly returns. This finding has policy implications too, because a negative wage elasticity apparently rules out the possibility that governments can stimulate entrepreneurial effort and output by cutting rates of income tax. Chapter 17 will explore the theme of taxation and entrepreneurship in greater depth. Several resolutions of the entrepreneurial labour supply puzzle mentioned earlier have been proposed. Some of them can be ruled out fairly quickly. One is dominant income and wealth effects among entrepreneurs who, being relatively well off (see chapter 9), can afford to take more leisure than employees. An immediate problem with this argument is that entrepreneurs work longer, not shorter, hours than employees (see Table 12.1). Second, while it has been proposed that entrepreneurs like working, perhaps because they possess a pronounced hard work ethos or ‘Protestant work ethic’ (Bonnett and Furnham, 1991), Ajayi-obe and Parker (2005) report that the self-employed are significantly less satisfied with their work hours than employees are. This implies that entrepreneurs (like employees) prefer leisure to work. A third unsatisfactory explanation of the entrepreneurial labour supply puzzle is that concavity of entrepreneurs’ production function (as in Figure 12.2) implies the highest computed ‘wage’ is associated with the smallest number of work hours – the negative relationship observed in the data. However, this is just an artefact of ‘division bias’ and cannot resolve the puzzle. A fourth problematic explanation of the entrepreneurial labour supply puzzle is that entrepreneurs face some indivisibility in the production set which mandates high work hours. But one of the hallmarks of entrepreneurship is flexibility; and it is unclear what such an indivisibility could be and why entrepreneurs could not use a different technology instead. A fifth proposal is a compensating differential in entrepreneurship such as a desire for independence (see chapter 4). While this can possibly explain why entrepreneurs work for lower returns than employees do, it leaves unanswered the question of why entrepreneurs work longer hours in entrepreneurship. Sixth, being more optimistic (chapter 4), entrepreneurs might believe they have more to gain from exerting effort while accepting low returns now in the expectation of higher returns later (which do not materialise). However, I know of no direct evidence as yet which supports this conjecture. This leaves two plausible alternative explanations. In the first, Camerer et al. (1997) argue that taxi drivers respond to busier conditions (and hence higher hourly wages)

350

Performance

by working shorter rather than longer hours because they ‘take one day at a time’, having a target income for each day, and stopping work as soon as they achieve it. This argument can explain the puzzle in principle, because these workers respond to low wages by having to work longer hours in order to reach their target income. The target income idea is rooted in behavioural economics and stands in direct opposition to neoclassical ‘life cycle’ theories of labour supply, according to which workers practise intertemporal substitution, working harder today if they anticipate lower returns (e.g. fewer customers) tomorrow. An alternative explanation, explored by Parker, Belghitar and Barmby (2005), is that entrepreneurs respond to greater income risk inherent in entrepreneurship by working longer hours to make the deterministic part of their incomes larger, despite facing a lower average wage. A non-pecuniary benefit would be sufficient to keep them in participating in entrepreneurship despite an inferior risk–return combination compared with paid employment.17 However, this result depends on risk affecting incomes in entrepreneurship additively (Parker, Belghitar and Barmby, 2005). In fact, evidence obtained by Parker, Belghitar and Barmby (2005) using data on 1,379 self-employed American household heads over 1968–93 from the PSID confirms that entrepreneurs who face greater risk work longer hours. Parker et al. measured entrepreneur-specific risk as the rolling standard deviation of an entrepreneur’s log wages in contiguous previous time periods. This variable had a large positive and statistically significant effect on work hours, and was found to be the only statistically significant variable explaining American entrepreneurs’ labour supply. So far, most attention has focused on the impact of entrepreneurial payoffs on work effort. Yet other determinants of work effort are also of interest. Although few sociodemographic covariates appear to systematically influence entrepreneurs’ work hours (see the final column of Table 12.2), an alternative approach which emphasises moral hazard considerations has yielded some interesting insights. Thus Bitler et al. (2005) ask whether a high ownership share in a business is necessary to incentivise entrepreneurial effort. Their estimates, based on a large sample of SCF data, suggest that it is, with a large and statistically significant effect on work hours. Bitler et al. (2005) also observed that an entrepreneur’s equity and the size of their business (measured as the log of the number of employees) are significantly positively related to work hours, while log wealth has a significant negative effect – presumably capturing an ‘income’ effect. Incentive effects with regard to work hours depend on economic conditions and are likely to differ between entrepreneurs and managers. De Fraja (1996) predicts that entrepreneurs will work harder when times are bad – unlike managers, who do the opposite. The reason is that leisure is a normal good for entrepreneurs, so less income in bad times implies less leisure, i.e. more work. Whereas in the standard managerial principal-agent problem where ownership is separated from control, managers are incentivised to work hard by the promise of higher salaries when output is high (see part 2 of chapter 2’s appendix). This occurs more often in good times. There are various areas where further research is needed to improve our understanding of entrepreneurs’ labour supply. One involves the specification of a dedicated

Entrepreneurial effort

351

life-cycle model of entrepreneurial participation and work effort. A dynamic life-cycle model has not yet been formulated and estimated specifically for entrepreneurs, despite the fact that life-cycle models incorporate information from future time periods as well as the present and so are superior (at least in principle) to the static models of labour supply discussed above.18 Life-cycle models also enable policy-relevant labour supply elasticities to be estimated, since tax reforms typically impact on both current and future incomes (Blundell and MaCurdy, 1999, sec. 4.5). One suspects that life-cycle models have not yet been constructed for entrepreneurs at least in part because future incomes are difficult to predict for individuals who do not even know how long they are likely to remain in entrepreneurship. The closest we have got to a complete lifecycle model is a dynamic analysis of the entrepreneurial retirement decision (see the next section). This is unfortunate, as such a model might help researchers to determine whether entrepreneurs exploit their apparent operational flexibility and seize opportunities for intertemporal substitution, opportunities which might be more limited for employees locked into more rigid labour contracts. Another task for future researchers is to apply state-of-the-art estimation methods to models of entrepreneurial labour supply, perhaps in the form of natural experiments, to improve estimates of the uncompensated wage elasticity. It will no doubt be valuable to incorporate non-pecuniary factors into these models, in view of the findings in chapter 4 about their impact on the attractiveness of entrepreneurship – as well asAtrostic’s (1982) evidence about their importance for explaining employees’ labour supply behaviour.

12.3 Ageing, retirement and entrepreneurship

At around the time of start-up, entrepreneurs often work very long hours in order to establish and grow their enterprise. Once established, new enterprises then tend to require somewhat less strenuous effort to maintain and manage. The implication is that younger entrepreneurs work harder than older ones do, since a higher proportion of the former implement start-ups than the latter, who have moved on to consolidate their earlier achievements (see chapter 4). Younger entrepreneurs might also work longer hours than older ones because they have greater energy and stamina. Certainly selfemployed Americans of retirement age work fewer hours on average than their younger counterparts do (Fuchs, 1982; Honig and Hanoch, 1985). A further reason why entrepreneurs’ labour supply may decline with age is that entrepreneurs’ returns from expending costly effort decreases over time if effort reveals valuable information about innate ability. A model with this property, by Frank (1988), is outlined in the first part of the chapter appendix. As shown there, as an entrepreneur approaches retirement, the value of their information about future returns in entrepreneurship declines to zero. This gives entrepreneurs weaker incentive to supply costly effort which reveals information as they age. Some independent evidence supports Frank’s (1988) theory. Parker (2006a) estimated a simple autoregressive model of work hours using data on 716 self-employed

352

Performance

Britons from the 1999 and 2000 waves of the BHPS. Parker (2006a) estimated an autoregressive coefficient of 0.84, which was significantly less than 1. This suggests that on average entrepreneurs supply declining work hours as they age. Splitting the sample into younger and older entrepreneurs, the autoregressive coefficient was larger for older entrepreneurs, in accordance with the notion of convergent learning in Frank’s model. However, Frank’s (1988) model stops short of explaining the age at which entrepreneurs retire. I now turn to an analysis of this issue. It is well established in the UK and USA that among workers who have passed retirement age, entrepreneurs are more likely to continue participating in the workforce than employees are.19 In both countries, around one-third of the workforce aged over sixty-five is self-employed.20 This proportion appears to be relatively stable over time. Among the very oldest segment of the workforce, self-employment rates are even higher (Fuchs, 1982). In addition, European entrepreneurs are substantially less likely to declare that they wish to retire than employees are (Hochguertel, 2005a). Several explanations of these phenomena can be proposed: 1. Older workers are from cohorts in which self-employment was once more common than it is now. 2. Anticipating a greater return to their labour effort, there is a greater willingness among optimists to retire later (Puri and Robinson, 2005). Since entrepreneurs are more over-optimistic than non-entrepreneurs are (see chapter 4), this might explain entrepreneurs’ later retirement ages. 3. Employees face a statutory retirement age (as in e.g. the UK), and sometimes also early retirement incentives embodied in occupational pension rights. Neither factor is applicable to entrepreneurs. There might also be bias by employers against hiring older workers, which encourages such workers to set up on their own if they wish to continue working. Some other individuals may work beyond the statutory retirement age because they have insufficient pension rights (see below). And an entrepreneur who has devoted their life to creating and growing a business might be unwilling to give it up altogether for non-pecuniary reasons, perhaps choosing to work fewer hours rather than relinquishing control entirely. 4. Entrepreneurship offers flexibility. Some employees switch into entrepreneurship as they approach retirement, especially in their fifties (Zissimopoulos and Karoly, 2007). Around one-third of American retirees subsequently return to work, often in ‘bridge jobs’ (Beck, 1986), and nearly one-quarter of American employees who continue to work do so by switching into self-employment. Unlike career selfemployed individuals, this is generally in different occupations and industries from their pre-retirement work; is often part-time in nature; and is often paid at a lower level than the previous career job (Iams, 1987; Doreinger, 1990). The pronounced tendency of older workers to transition into part-time work in self-employment (as opposed to full-time work in paid employment) suggests that self-employment might be a means of obtaining work hours flexibility (Zissimopoulos and Karoly, 2007).

Entrepreneurial effort

353

Fuchs (1982) and Bruce et al. (2000) estimated the switching rate from paid employment to self-employment to be around 2 per cent per annum for older male American employees. The comparable figure is less than 1 per cent in Britain (Parker and Rougier, 2007). As noted above, switching into self-employment in later life might occur because self-employment is an easier way for individuals to partially retire. For example, Quinn (1980) observed that the work hours of older self-employed individuals display markedly greater variation than do those of older employees. However, the extent of switching should not be overstated: most selfemployed Americans over 65 have been self-employed for many years rather than newly arrived from paid employment (Iams, 1987). The distinction between established older self-employees and newcomers to self-employment turns out to be an important one in empirical work (see below). Entrepreneurs do not face a mandatory retirement age, and do not usually receive employer- or government-provided benefits. Instead they have to provide their own pension and health-care coverage, and determine their date of separation from the workforce freely. Hence life-cycle saving is likely to be an integral aspect of the entrepreneurial retirement decision. The formal life-cycle model outlined in the second part of the chapter appendix predicts that the extra income which an entrepreneur can earn by deferring retirement by one extra period (called the ‘accrual’) will have a negative impact on the probability that they retire, whereas an entrepreneur’s stock of lifetime wealth advances the date of retirement. Parker and Rougier (2007) estimated this life-cycle model by applying probit estimation to a sample of 197 older self-employed individuals taken from the 1994 British Retirement Survey. These data enable lifetime wealth and the accrual to be measured for older self-employed people. The significant determinants of the probability of selfemployed retirement were found to be the accrual, with a negative sign as predicted; age, with a positive effect; and self-employment status six years earlier, with individuals who were self-employed six years earlier being significantly less likely to retire than those who had recently switched into self-employment. Less than 1 per cent of employees and virtually no retirees were observed to switch into self-employment in later life. The switches that did occur seem to have been motivated less by attempts to treat self-employment as a bridge job or ‘stepping stone’ to full retirement, than by self-employment being an ‘occupation of last resort’ among less affluent workers with job histories of weak attachment to the labour market. This echoes Evans and Leighton (1989b), who reported that poorer wage workers – the unemployed, lowerwage workers and those with a history of job instability – are especially likely to become self-employed in the USA. Parker and Rougier (2007) also found that having a working spouse delays an entrepreneur’s retirement date. This chimes with survey work summarised by Hochguertel (2005b), showing that many European entrepreneurs claim they are awaiting retirement in order to enjoy life and to retire jointly with their spouses. Unlike employees, they are less likely to cite awaiting retirement because of becoming eligible

354

Performance

for a pension. Partly this is because entrepreneurs are less likely to be covered by a pension plan than employees are (Karoly and Zissimopoulos, 2004; Parker and Rougier, 2007). Furthermore, entrepreneurs might use retirement as a means of risk reduction: Heaton and Lucas (2000) observed a reduction in risk-taking among older wealthy business-owning households, who substitute away from private business ownership and towards stocks, bonds and cash. In early work, Quinn (1980) and Fuchs (1982) found that personal wealth and eligibility for social security retirement benefits increased the probability that older self-employed Americans retire. In contrast, Parker and Rougier’s (2007) more comprehensive measure of lifetime wealth was statistically insignificant. More recently, Zissimopoulos and Karoly (2007) updated Fuchs’ (1982) early work, using five waves of the US Health and Retirement Survey over 1992–2000 to study the transitions from paid employment of full-time workers aged 51–67. Zissimopoulos and Karoly (2007) reported that for older men these transitions were likelier to be to selfemployment (compared with retirement or inactivity) the greater were respondents’ education, wealth, inheritances and wages in paid employment. Transitions to selfemployment were also more likely for men with health conditions which limit work (in contrast, Parker and Rougier, 2007, detected no such effects in their British sample). None of social security wealth, age, ethnicity or marital status affected the transition probabilities analysed by Zissimopoulos and Karoly (2007), although having health insurance and private pensions reduced them. As in Fuchs (1982), occupations matter: male executives, salesmen and labourers were significantly more likely to transition to self-employment than administrative support clerks. Salesmen were 63 per cent more likely to make the transition than the average. For older women, Zissimopoulos and Karoly (2007) reported fewer significant covariates. Like men, though, they observed positive effects on transitions into selfemployment from: a health condition which limits work; negative effects from own or spouse health insurance and private pension; and insignificant effects from social security wealth. But unlike men, female transitions to self-employment were negatively associated with age, marital status and having a working husband; and Zissimopoulos and Karoly (2007) found insignificant effects for women from their wages in paid employment, education or inheritances. Having now documented the evidence, two issues remain to be discussed. One concerns the effects of retirement on family firms, since the retirement of a family business owner has implications for family business succession. The other relates to the policy implications of entrepreneurial retirement behaviour. Family succession might be motivated for several reasons, including: • • • •

Generating non-pecuniary benefits if founders have dynastic preferences. Cementing the firm’s reputation with customers. Promoting far-sighted investment behaviour. Enabling valuable informal human capital to be transmitted to successors from an early age.

Entrepreneurial effort

355

• Facilitating trust and capital pooling. • Strengthening kinship networks which consolidate and leverage advantageous political connections (Bertrand and Schoar, 2006). Another explanation of within-family business succession is weak legal protection of entrepreneurs who sell their firm to managers hired from outside the family. Burkart et al. (2003) argue that founders are more likely to pass on a business to their heirs when outside managers find it easier to expropriate value from the business. This risk entails costly monitoring by the founders. So Burkart et al. (2003) predict that owners will be more likely to keep businesses under close family control in economies that lack strong legal protection of investors – even though outside managers may be able to run the firm better, and generate a higher sale price to the founder. Indeed, crosscountry evidence from La Porta et al. (1999) and Claessens et al. (2000) shows that dynastic family firms are especially widespread in countries with weak legal protection of shareholders. Turning to policy issues, some commentators have contended that entrepreneurship among older workers – so-called ‘third age entrepreneurship’ – might become an increasingly important phenomenon in the future. For example, it seems likely that bridge jobs will increase in importance as individuals enjoy longer and healthier lives. The institutional backdrop is that government policies in the USA and elsewhere in the last twenty years have made continued work among older people an increasingly attractive option, for example by dismantling pension earnings penalties (Bruce et al., 2000).21 However, the modest switching rates among older workers into entrepreneurship cited earlier cast doubt on the likely effectiveness of ‘third age entrepreneurship’ policies. In any case, questionnaire responses reveal that at most only 15 per cent of 50– 75-year-old Britons express any interest in self-employment (Curran and Blackburn, 2001). While acknowledging the limitations of survey claims about future possible behaviour, this figure is likely to be an upper-bound estimate of the true level of interest, because respondents do not have to bear immediately the costs of switching into entrepreneurship. Nor is it obvious that the importance of self-employment among older people is set to increase dramatically in the near future. For their part, Parker and Rougier (2007) observed that most British employees who switch into entrepreneurship in later life have histories of low incomes and multiple job spells. Since these ‘switchers’ are more likely to become self-employed only for short periods before retiring, the value-added of their businesses is also questionable. These findings therefore cast doubt on the potential for generating meaningful increases in labour force participation rates and wealth creation via self-employment schemes for older workers. Finally, there is no shortage of opportunities for further research to enrich our understanding of entrepreneurs’ venture development efforts in later life, and of their retirement behaviour. One challenge is to model work hours jointly with actual and expected future entries into and exits from entrepreneurship, through the medium of a truly dynamic, forward-looking life-cycle framework.

356

12.4

Performance

Entrepreneurial learning

Entrepreneurs not only supply time to their ventures, but also expend effort on search, evaluation of opportunities and learning from events. A large business studies literature discusses identification and exploitation of opportunities by entrepreneurs, and seems to conclude that this represents the essence of entrepreneurship (e.g. Shane and Venkataraman, 2000). This section in contrast takes a different tack, and focuses instead on effort devoted to active learning by entrepreneurs about their environment. The model of Frank (1988), described in the previous section, already introduced the notion of active learning, whereby entrepreneurs actively and deliberately explore and learn about their business environment (in that case by adjusting their labour supply), to improve their productivity and enhance their survival prospects. In contrast, passive learning has firms simply receiving information without expending effort. These firms process information which arrives from the marketplace in order to determine whether they should continue trading, or whether they should leave the market if their innate productivity appears to be too low to generate positive expected profits in the future. A ‘classic’ model of passive learning in entrepreneurship is that of Jovanovic (1982), discussed in the previous chapter. An interesting implication of active learning is that entrepreneurs can be ‘made, not just born’ – or, more precisely, they can ‘make’ themselves successful. This represents a kind of ‘bootstrapping’ behaviour, and bears on the efficiency of entry, since innovating entrepreneurs who anticipate that others can learn from and improve on their new technology might face disincentives to enter a market in the first place (Jovanovic and Lach, 1989). While refusal to enter for this reason is privately rational, the economy as a whole could suffer from socially sub-optimal levels of entry. More generally, active learning is costly, and it is not always privately optimal for entrepreneurs to invest in it. Specifically, if the expected marginal benefit from investing in learning (i.e. higher productivity) is outweighed by the marginal costs of investing, then it may be optimal for entrepreneurial firms to ‘coast’ instead (Ericson and Pakes, 1995).22 The available evidence suggests that in practice entrepreneurs perform both active and passive learning. According to Canadian evidence from Baldwin and Rafiquzzaman (1995), improvements in the average post-entry performance of cohorts of new firms are mainly attributable to passive learning, whereby poor firms select out of the market. Among the survivors, labour productivity and profitability tend to grow very little over the first decade, suggesting that relatively little active learning takes place. In contrast, Turkish evidence from Taymaz (2005) points to both effects being present, with the elimination of low-growth firms (passive learning) together with continued growth among survivors (active learning). Taymaz (2005) found little difference in the extent of active learning by initial entry size, though he estimated that new entrants are less efficient than incumbents are, and that the average efficiency differential is wider among small entrants. Other evidence suggests that younger and less experienced entrepreneurs perform the most learning.23 Sapienza et al. (2005) explored the (self-assessed) effort that

Entrepreneurial effort

357

ninety Belgian owner-managers devote specifically to learning about domestic and international markets. Regression analysis revealed that learning efforts are associated with a younger age at date of first internationalisation. De Clercq and Sapienza (2005) asked 298 US VC fund managers to assess how much they learn from their portfolio companies. They found that those fund managers with less prior investment experience and less trusting relationships with their portfolio companies learn significantly more than average; so do VCs managing better-performing companies. One likely reason why younger entrepreneurs learn more, noted above, is diminishing returns to learning, leading to convergence of beliefs among older entrepreneurs as learning is performed (Parker, 2006a). Other factors are also likely to be at work, however. For example, using data on Italian manufacturing firms, Guiso and Schivardi (2005) report that entrepreneurs who grew up in areas with more firms when they were twenty-five years old earned higher profits later. Guiso and Schivardi’s (2005) explanation is that learning is facilitated by exposure to other firms. Research into entrepreneurial learning in the economics of entrepreneurship continues apace. I anticipate that researchers will pay greater attention to this issue in future work, both theoretically and empirically.24 12.5 Appendices 12.5.1 Frank’s model of entrepreneurs’ labour supply and ageing An entrepreneur works for a finite time span T , at which point they die. In each period the entrepreneur must choose whether to continue in entrepreneurship (zt = 1) or to exit (zt = 0). If zt = 1, then they choose work hours, ht ∈ [0, h]. The disutility of supplying labour is given by c(ht ), where c(·) is an increasing and convex function. The entrepreneur’s revenue is xht +t . Here x is ability, which is unknown to the entrepreneur at the time of entry at t = 1. The entrepreneur’s prior belief is that x is distributed with mean µ1 and precision (defined as the reciprocal of the variance) v1 . Also, t is a normally distributed random influence (e.g. ‘luck’), which is distributed normally with mean zero and precision v . The entrepreneur performs Bayesian learning about x by observing successive realisations of xht + t , which at time t is normally distributed with mean xht . The entrepreneur’s beliefs in t + 1 are characterised by a distribution with mean µt+1 and precision vt+1 , where

µt+1 =

vt µt + v ht (xht + t ) vt + v h2t

vt+1 = vt + v h2t .

(12.4) (12.5)

Evidently, if ht = 0 then µt+1 = µt and vt+1 = vt . Thus if the entrepreneur expends no effort, she obtains no new information about her entrepreneurial ability. If zt = 0, the individual takes an outside option (e.g. paid employment) that yields utility wt . Let Ut be the entrepreneur’s discounted expected utility given that they entered initially and choose all future z and h values optimally; Ut ≡ 0 for t > T . Et

358

Performance

is the expectations operator conditioned on the information available at time t, and β = 1/(1 + δ) ∈ (0, 1] is the discount factor. In each time period after entry at t = 1, we have Ut = max {wt (1 − zt ) + [Et (xht + t ) − c(ht )]zt + βEt Ut+1 } zt ,ht

t = 2, 3, . . . , T . (12.6)

Proposition (Frank 1988) Entrepreneurs optimally supply less labour as they age. Proof The model is solved by backwards induction. At T , an entrepreneur remains in the industry (i.e. zT∗ = 1) if wT ≤ µT h∗T − c(h∗T ), where h∗T is optimal terminal effort, which satisfies µT = ch (h∗T ). Given this solution, the period T − 1 problem is UT −1 =

max {wT −1 (1 − zT −1 ) + [ET −1 (xhT −1 + T −1 )

zT −1 ,hT −1

−c(hT −1 )]zT −1 + βwT (1 − zT −1 ) + βET −1 UT∗ zT −1 ,

(12.7)

where UT∗ is the maximal payoff in T conditional on zT −1 = 1, and using an auxiliary result that entrepreneurs do not re-enter once they exit, so that zT −1 = 0 ⇒ zT = 0. If the entrepreneur selects zT −1 = 1, then the maximising choice of hT −1 , namely h∗T −1 , satisfies the first-order condition µT −1 − ch (hT −1 ) + β

∂[ET −1 UT∗ ] = 0, ∂hT −1

(12.8)

where the derivative in this equation is obviously positive. Now for the presumed pattern of continuation to be optimal, we must have µT −1 ≥ µT = ch (h∗T ). But this then implies ch (h∗T ) ≤ ch (h∗T −1 ) − β which implies h∗T −1 > h∗T .

∂[ET −1 UT∗ ] , ∂hT −1




The logic behind Frank’s proposition can be seen by comparing the period T objective with that for T − 1 (the latter is (12.7) in the proof). Only part of the payoff to supplying effort is the production of current output. Greater effort also generates returns in the form of valuable information about the future, where the term β∂[ET −1 UT∗ ]/∂hT −1 in (12.8) can be interpreted as the marginal information value of an extra unit of effort. As the entrepreneur approaches retirement, the value to them of information about future returns in entrepreneurship declines to zero. Consequently, this gives the entrepreneur less incentive to supply costly effort as she ages.

Entrepreneurial effort

359

12.5.2 Parker and Rougier’s model of the entrepreneurial retirement decision The life-cycle problem of a self-employed individual aged t who plans to retire at age R and expects to live until age T can be written as 
R #
T max V = exp{−δ(s − t)} U (ζ (s), 0) ds + exp{−δ(s − t)} U (ζ (s), 1) ds {ζ (s)}s=T s=t }

t


subject to

R T

exp{−r(s − t)} ζ (s) ds = B(R)

t

where ζ (·) is consumption through time, s is an age index, U (·, ·) is an ‘instantaneous’ utility function defined on consumption and (binary) retirement status respectively, δ and r are discount and interest rates, respectively, and B(R) is current total lifetime wealth, which is a function of retirement age. Lifetime wealth is the value of financial, business and housing assets plus the present value of all state pension (social security), personal pension and labour incomes that an individual commands over the remainder of their life. For an individual who is considering retirement, set R = t. Then the solution to the above problem takes the form " ∂V ∗ "" = −∇(ζ ∗ (t), t) + λ∗ BR (12.9) dR "R=t where ∇(ζ ∗ (t), R) := U (ζ ∗ (t), 1) − U (ζ ∗ (t), 0) is utility in retirement relative to work; λ∗ := dV ∗ /dB ≥ 0 is the marginal utility of wealth; and BR := dB/dR at R = t. We would usually expect that ∇(ζ ∗ (t), R) > 0, i.e. the same amount of consumption is more enjoyable when combined with leisure time (i.e. in retirement). Because of earnings opportunities in work, we would also expect that BR > 0, i.e. deferring retirement by one period enhances the current stock of lifetime wealth by the amount of extra income generated in that period. Note that BR (connoted by the term ‘accrual’) is not generally 0 for retired individuals, as it is ‘the incremental change in my lifetime wealth were I currently to be working and decided not to retire immediately but to defer for a small amount of time’. It is true that this value may be easier to compute for those currently working, but it is also clear that those who have retired and can return to employment ought also to be making a calculation of this kind. Since both terms in (12.9) are likely to be positive, there are likely to be individuals for whom it is beneficial to retire (negative sign), and others for whom it is beneficial to continue to work (positive sign). Consequently it seems reasonable to model the probability that the individual is retired at age t as some monotonic increasing function of the negative of (2), written as " ∂V ∗ "" ∗ ∗ ∗ π = Pr[R ≤ t] = (v ) where v = − (12.10) dR "R=t where (·) is the probit conditional mean function. Write v ∗ ≈ f (X ) − λ∗ BR , where f is an unknown function which depends upon the unknown utility function, and is

360

Performance

taken to be a function of a vector of covariates X . It is convenient to use the following specific functional form for a sample of entrepreneurs indexed by i: vi∗ = α1 + γ1
Xi − (β0 − β1 Bi )BRi + i

(12.11)

where λ∗ = λ(B) = β0 − β1 B, and where i is an uncorrelated Gaussian error term. Eq. (12.11) can be estimated by standard probit methods. A testable prediction of the life-cycle model is that the accrual has a negative effect (i.e. β0 > 0), and lifetime wealth a zero or positive effect (i.e. β1 ≥ 0), on the probability that a given individual is retired. Notes 1. See Cooper, Woo and Dunkelberg (1989), McCarthy et al. (1990), Cooper, Ramachandran and Schoorman (1997) andAldrich (1999).At even earlier stages of venture creation, Choi et al. (2008) argue that entrepreneurs should switch effort from exploration to exploitation of an opportunity as their ignorance about the opportunity decreases. The less novel a venture proposition, the longer an entrepreneur ought to spend in exploration mode. 2. More generally, the focus will be on the quantity rather than the quality of time supplied. From a theoretical point of view, effort quality is hard or impossible for outsiders to observe, entailing moral hazard considerations. These considerations have already been touched upon in other parts of this book, and those discussions will not be repeated here. 3. As an anti-free-trade argument, it might seem that protecting large firms at early stages of development might be beneficial if this policy helps to relax credit constraints. But as Acemoglu et al. (2006) point out, this policy becomes increasingly costly for the economy as it approaches the technological frontier at which point market entry by able entrepreneurs takes over as the engine of growth. 4. A third case is also possible, where γ < β < α. Then two stable equilibria exist: one at n = 0 (everyone earns α) and one where everyone earns γ . 5. At a more informal level, Murphy et al. (1991) noted a significant positive correlation between economic growth and the proportion of engineering college majors (taken to be a socially productive occupation), and a significant negative correlation with the proportion of law majors (taken to be a socially unproductive, rent-seeking, occupation). Murphy et al. (1991) claimed that this demonstrates that occupational choices have important effects on economic growth. They also argued that factors facilitating trade and competition, such as lower communication and transportation costs, favour productive over unproductive entrepreneurial activities. 6. See, respectively, Bregger (1996), Kuhn and Schuetze (2001) and Ajayi-obe and Parker (2005). 7. See OECD (1986), Aronson (1991), Rees and Shah (1994), Moralee (1998) and Table 12.1. 8. For evidence that this is so, see Aronson (1991), Carrington et al. (1996), Hamilton (2000), Ajayi-obe and Parker (2005) and Parker, Belghitar and Barmby (2005). 9. Normalising the amount of available time to unity, the simplest static model has the following structure: max U (ζ , 1 − h) h

subject to

ζ ≤ wh + I ,

(12.12)

where ζ is consumption and h is work hours [so 1−h is leisure hours], subject to an atemporal budget constraint where consumption is bounded by labour (wh) and non-labour (I ) income. Then for particular utility functions U (·, ·), taken to be increasing in both arguments, it is possible to derive from the first-order conditions for this problem a work-hours specification h = h(w, I ; X ), where

Entrepreneurial effort

361

h is increasing in w, decreasing in I , and depends also on a vector of individual characteristics, X . Blundell and MaCurdy (1999, Appendix A) list various utility functions corresponding to particular specifications of h(·). 10. For this reason, some authors have estimated non-linear work-hours functions directly from the first-order condition of (12.12), using a reasonably flexible utility function such as Generalised Stone-Geary (e.g. Wales, 1973; Thornton, 1998). 11. The marginal entrepreneur is required to use factor inputs as efficiently as non-owner-managed firms in order to survive under competition. Only for infra-marginal entrepreneurs, i.e. those with access to lower costs or higher net incomes, can utility maximisation and profit maximisation diverge, leading to persistent positive rents in the long run (see also Feinberg, 1980). More generally in fact, one could argue that the neoclassical labour supply model is more appropriate for entrepreneurs than for employees, since entrepreneurs have greater flexibility to adjust their work hours than employees do. This is reflected in the greater volatility of daily and weekly work-hours of the self-employed compared with employees. The latter work for employers who smooth wages and adjust to adverse shocks by allowing employees to work fewer weeks per year in recessions (Carrington et al., 1996). 12. To see the problem, observe that entrepreneur j’s operating profit is πj = pj [q(hj )]q(hj ) − cj ,

(12.13)

where the output price pj is a decreasing function of output q(hj ) – which is an increasing function of work hours hj . Here cj are input costs. Then the wage is defined as wj := πj /hj , which is in general a function of work hours hj . Hence simple OLS estimates based on wj are subject to ‘division bias’ (Rogerson and Rupert, 1993) and should be avoided. 13. Accurate measurement of entrepreneurs’ wages is further complicated by the inclusion of income from second jobs in paid employment. The last problem is usually handled by excluding multiple job-holders from data samples. 14. To see why, consider the following fixed-effects version of (12.2): ln hjt = β0j + β1 ln wjt + β2
ln Xjt + vjt ,

(12.14)

where β0j are entrepreneur-specific fixed effects, and where (12.14) cannot be estimated as it stands because the true values of h and w are both potentially measured with error. Denoting measured values of these variables with asterisks, suppose hjt = h∗jt exp{ξjt }

⇒

πjt = πjt∗ /φj

⇒

ln hjt = ln h∗jt + ξjt and $ ∗% πjt e−ξjt wjt = πjt /hjt = h∗jt φj

(12.15) (12.16)

where ξjt are mean-zero i.i.d. errors and φj ∈ (0, 1) is a profit under-reporting factor specific to entrepreneur j. Then it follows from (12.16) that ln wjt = ln wjt∗ − ln φj − ξjt . Putting this expression and (12.15) into (12.14), we obtain ln h∗jt + ξjt = β0j + β1 ln wjt∗ − β1 (ln φj + ξjt ) + β2
ln Xjt + vjt . First differencing this equation and writing the new disturbance term jt := vjt − (1 + β1 )ξjt , we obtain the final measurable specification (12.3) in the text. 15. In Feinberg’s words, ‘So we find a possible backward-bending labour-supply curve for the owner manager, with hours depending not on his own wages (which are not defined) but on the market wages of hired managers of comparable talents’ (1987, p. 214). 16. Lopez (1984), Jacoby (1993) and Thornton (1994).

362

Performance

17. To see this, suppose that while employees can benefit from wage smoothing by their employers, receiving a constant wage wE , entrepreneurs cannot diversify their risk, which takes the form of a wage shock  where E = 0. To illustrate, suppose that entrepreneurs’ indirect utility function takes the additively separable form V ∗ = U1 (wh∗ + ) + U2 (1 − h∗ ) + ν ,

(12.17)

where ν > 0 is a non-pecuniary advantage to entrepreneurship, w is entrepreneurs’ wage, and U1 (·) and U2 (·) are concave increasing functions. Assume also that ν is sufficiently large to ensure occupational equilibrium with wE > w. The first-order condition determining entrepreneurs’ optimal work hours h∗ is wE∂U1 /∂h∗ − ∂U2 /∂h∗ = 0. Then it is known (Rothschild and Stiglitz, 1971) that a mean-preserving spread (MPS) in  will increase or decrease h∗ as ψ() := w∂U1 /∂h∗ − ∂U2 /∂h∗ is convex or concave in . Differentiating twice with respect to  yields ∂ 3 U1 (·) ∂ 2 ψ() =w , 2 ∂ ∂y3

18. 19.

20. 21.

22.

23. 24.

(12.18)

where y := wh∗ + . The RHS of (12.18) is positive if ∂ 3 U1 (·)/∂y3 > 0, in which case ψ() is convex, and work hours increase with the degree of risk. Since this condition on U1 (·) is satisfied by utility functions embodying decreasing absolute risk aversion, this case seems a reasonable one. Entrepreneurs effectively ‘self-insure’ by choosing longer work hours, thereby making the deterministic part of their income larger. Rees and Shah (1994) outlined a theoretical life-cycle model for entrepreneurs, but went on to estimate a static empirical model instead. See Parker (2004, pp. 199–201) for details. For UK evidence, see Moralee (1998), Parker (2003a) and Parker and Rougier (2007); US evidence includes Quinn (1980), Fuchs (1982), Iams (1987) and Zissimopoulos and Karoly (2007). Iams (1987), Moralee (1998) and Bruce et al. (2000). The self-employed appear to exploit tax-based retirement savings incentives, too. For example, Power and Rider (2002) estimated that the tax-price elasticity of contributions to tax-deferred retirement savings plans in the USA by the self-employed is –2.0. Other possible reasons for limited active learning include constraints on entrepreneurs’time, attention and resources (Ravasi and Turati, 2005); and the use of crude self-confirming heuristics in decision-making which arise because of cognitive limitations (Busenitz and Barney, 1997). However, entrepreneurs’presumed use of heuristics has not gone without challenge in entrepreneurship research (Wadeson, 2006). See de Clercq and Sapienza (2005), Sapienza et al. (2005) and Guiso and Schivardi (2005). Four theoretical contributions by the author analyse learning under various assumptions about rationality, including adaptive expectations (Parker, 2006a), Bayesian updating (Parker, 2007c; Alvarez and Parker, 2009) and learning with cognitive biases (Parker, 2009a).

13 Entrepreneurs’ incomes and returns to human capital

A key dimension of entrepreneurial performance is financial. Financial performance is usually measured in terms of income or profitability. Income net of costs is equivalent to profits for entrepreneurs who do not incorporate their business, whereas incorporated entrepreneurs pay themselves an income out of their operating profit. To simplify the exposition in this chapter, the two terms income and profits will be used more or less interchangeably unless specifically stated to the contrary. It is commonly held that profits are the primary goal of entrepreneurs, in much the same way that employees seek jobs paying the highest wages. In view of the occupational choice model of entrepreneurship outlined in chapter 2, one obvious and important question is the extent to which entrepreneurs earn higher incomes than employees do. Naturally, if individuals are risk-averse, the volatility of incomes also bears on the attractiveness of entrepreneurship compared with paid employment. For example, volatility might introduce a risk premium into entrepreneurial returns. The volatility of incomes is related to, but distinct from, the inequality of incomes in entrepreneurship and paid employment. The objective of this chapter is to review the state of knowledge about these issues. The chapter also discusses the individual-specific determinants of entrepreneurial incomes. A topic of particular interest is the role that skill plays in enhancing entrepreneurs’ incomes, especially its embodiment in formal education. Entrepreneurs’ incomes can be measured in several different ways. Consider the following identity: Net profit ≡ Revenue – Costs ≡ Draw + Retained earnings. Net profit from running an enterprise is a widely used measure of self-employment income. An alternative is Draw, the amount of money drawn from the business on a regular basis by the owner. This represents the consumption-generating value of the business and as such may be less prone to income under-reporting than profit (see below). A less frequently used third measure is Draw augmented by the growth in business equity. Hamilton (2000) provides a detailed discussion of these different measures in relation to US data sources. 363

364

Performance

A serious challenge facing any researcher who seeks to bring data to bear on income or profits in entrepreneurship is to measure these constructs accurately. The first two sections of this chapter describe several important measurement issues, and propose some constructive solutions. It should be recognised at the outset that, by working for themselves, entrepreneurs find it easier than employees to evade income tax. Consequently, entrepreneurs who answer survey questionnaires might be reluctant to report accurately their incomes to interviewers, despite assurances by the latter that survey information is confidential and is never divulged to the tax authorities. These and other sources of measurement error in reported entrepreneurial incomes are discussed in some detail. The main message emerging from this discussion is that while there is no need for data limitations to deter researchers from using entrepreneurial income data altogether, it is imperative that researchers recognise the problems that do exist; rectify them where possible; and exercise due caution when interpreting empirical findings based on income data. The third section documents international evidence about entrepreneurs’ average incomes relative to those of employees. Invariably, data considerations mean that entrepreneurs are usually defined as the self-employed, since the income tax code and most nationally representative sample surveys distinguish sources of income based on an employee/self-employed dichotomy. This section critically evaluates the evidence pertaining to whether individuals receive a premium or pay a discount for the privilege of participating in entrepreneurship. A related question concerns the volatility and inequality of payoffs, since relatively high levels of risk in entrepreneurship imply that risk-averse individuals would need to receive a risk premium to make entrepreneurship an attractive choice. The fourth section reviews theory and evidence about the volatility and inequality of entrepreneurial incomes. The last two sections outline several theoretical and methodological issues which arise when analysing the determinants of entrepreneurs’ incomes using ‘earnings functions’. I conclude by summarising evidence about the determinants of this measure of entrepreneurial success. 13.1

Measurement issues: tax evasion and income under-reporting

Unlike entrepreneurs’ profits, conventional employment incomes usually offer little or no scope for tax evasion, because of third-party reporting and tax withholding by employers. Entrepreneurs have greater scope to evade tax because they are responsible for reporting their own incomes to the tax authorities. Not surprisingly, direct and reliable evidence about the extent of income under-reporting to the tax authorities by entrepreneurs is rather hard to come by. In this respect, data gathered by the US Internal Revenue Service’s (IRS’s) Taxpayer Compliance Measurement Program (TCMP) is of special interest. The TCMP data were compiled by teams of auditors who analysed thoroughly the tax affairs of random samples of employee and entrepreneur taxpayers on a case-by-case basis. The TCMP provides the most comprehensive direct measures of income tax non-compliance in

Entrepreneurs’ incomes and returns to human capital

365

the USA, or indeed any other country. It ran from 1963 to 1988, but was postponed indefinitely in 1995 on the basis of cost and complaints about its intrusiveness. A modified successor, the National Research Program, is currently being operated by the IRS (Slemrod, 2007). Numerous researchers have utilised TCMP data to estimate the degree of tax evasion. The TCMP tells a reasonably consistent story over time. As Feldman and Slemrod (2007) explain, whereas over 99 per cent of employee income is reported accurately to the IRS, non-farm proprietors report on average only 50–70 per cent of their true incomes to the IRS; the figure is even lower (at just over 40 per cent) for partnerships and ‘S corporations’. There are also pronounced industry variations in under-reporting rates, with self-employed taxicab drivers being the most likely to under-report their incomes to the tax authorities (Kesselman, 1989). Regarding the impact of incomes and marginal tax rates on entrepreneurs’ tendency to evade taxes, Joulfaian and Rider (1998) estimate that a 10 per cent increase in the marginal rate of income tax increases self-employed income under-reporting to the IRS by 5 per cent. Joulfaian and Rider (1996) also estimate that tax evasion is significantly negatively associated with the tax audit rate, with an elasticity of approximately −0.70. The cost of tax evasion by entrepreneurs is substantial. According to the 1988 TCMP, a full 30 per cent of unreported tax revenue comes from sole proprietors, who comprise only 5.5 per cent of the tax returns filed (Feldman and Slemrod, 2007). More recent 2001 IRS data suggest that under-reporting by non-farm proprietors (of 57 per cent) costs the IRS $68 billion per annum (Slemrod, 2007). This amounts to more than onethird of total estimated income under-reporting to the tax authorities for the individual income tax. Somewhat lower rates of self-employed income under-reporting to the tax authorities of around 20 per cent have been detected in the UK, although these estimates are based on less comprehensive data (Macafee, 1982, p. 155). Smith (1986) proposed an even lower income tax evasion rate among the self-employed of 14 per cent, and suggested that preventative policy actions in the construction sector can be a particularly effective way of reducing under-reporting rates. It is important to distinguish between under-reporting of incomes to the tax authorities and under-reporting of incomes to sample survey interviewers. The overwhelming majority of researchers use data on the latter rather than the former. However, entrepreneurs who illegally over-claim business tax deductions, or under-report gross incomes to the tax authorities, have reasons to be wary of revealing accurate income data to third parties such as survey interviewers. This is despite assurances by the latter to interviewees that all data are treated in strict confidence and cannot be divulged to the tax authorities. As the discussion below will reveal, household survey income under-reporting by entrepreneurs is possibly the most pervasive and serious limitation of public-use micro-data. It means that the raw data reported to interviewers understate, perhaps substantially, the true incomes received by entrepreneurs. Consequently, when making income comparisons between entrepreneurs and non-entrepreneurs, it is necessary to somehow estimate the survey income under-reporting rates for entrepreneur

366

Performance

respondents, and apply a correction factor to their declared incomes in order to yield more accurate values of their true incomes. Unfortunately, individuals who underreport their incomes cannot simply be identified as those who spend more than they earn, because some people may genuinely dissave and not under-report, while others may under-report and not dissave (Dilnot and Morris, 1981). A more reliable approach pioneered by Pissarides and Weber (1989) estimates expenditure equations for entrepreneurs (typically defined as the self-employed) and employees, and then inverts them to estimate ‘true’ incomes. An expenditure equation relates individuals’ spending to their incomes. Three assumptions underlie this approach: (i) all occupational groups report expenditures accurately to the survey interviewers; (ii) employees report their incomes accurately although entrepreneurs might not; and (iii) the income elasticity of consumption is the same for members of all occupations with the same characteristics. Regarding (i), food expenditure (Pissarides and Weber, 1989) and charitable contributions (Feldman and Slemrod, 2007) are both regarded as unlikely to be misreported by either occupational group. Under some further technical assumptions, Pissarides and Weber’s approach enables the researcher to calculate lower and upper bounds for individual-specific under-reporting factors (see Pissarides and Weber, 1989; and Schuetze, 2002, for details). The evidence derived from applications of this approach suggests that survey income under-reporting by entrepreneurs is of major economic importance. For example, using 1982 UK Family Expenditure Survey (FES) data, Pissarides and Weber estimated the average self-employment survey income under-reporting rate to be 55 per cent.1 Bluecollar workers were estimated to conceal from survey interviewers 51–64 per cent of their incomes, compared with 28–54 per cent for white-collar workers. Subsequent work by Baker (1993), using UK FES data over 1978–91 and the same methodology, generated somewhat lower self-employed income under-reporting rates of 20–50 per cent. Baker estimated that under-reporting rates differ between occupational and industry groups. Baker (1993) estimated lower under-reporting rates amongst blue-collar than white-collar self-employed workers, in contrast to Pissarides and Weber. Before presenting estimates from other countries, note that more recent estimates suggest that the extent of income under-reporting in the UK might be even greater than Pissarides and Weber’s and Baker’s (already high) estimates. For example, Lyssiotou et al. (2004) extend the Pissarides and Weber (1989) approach from a single consumer demand equation to a system of consumer demand equations. These authors relax assumption (iii) above, by allowing entrepreneurs to have different preferences and hence different consumer demands from employees; they claim that this avoids a potential source of bias in the estimates of under-reporting. Lyssiotou et al. (2004) show that restricting preferences to be homogeneous as in Pissarides and Weber (1989) leads to under-estimation of income under-reporting factors. Using UK data from 1993, Lyssiotou et al. (2004) estimate blue- (resp., white-) collar under-reporting rates of 118 (resp., 64) per cent. This implies that households whose heads are in blue-collar self-employed jobs report only 46 per cent [= 1/(1 + 1.18)] of their incomes to

Entrepreneurs’ incomes and returns to human capital

367

survey interviewers, while their white-collar counterparts report only 61 per cent of their incomes. Even these higher under-reporting rates might be downward-biased, if more volatile incomes cause entrepreneurs to save more than employees (Quadrini, 1999) – reducing their observed consumption. Lyssiotou et al. (2004) estimate that entrepreneurship-related black-economy activities in the UK amounted to 10.6 per cent of GDP in 1993 – nearly double Pissarides and Weber’s (1989) estimate of 5.5 per cent (see note 1). Estimates of under-reporting rates in other countries (other than the USA: see below) appear to be lower. For households obtaining at least 30 per cent of their income from self-employment, Canadian Survey of Consumer Expenditure data suggest that the under-reporting rate lies between 11 and 23 per cent (Schuetze, 2002). This is somewhat higher than an earlier estimate for Canada of 12.5 per cent (Mirus and Smith, 1996).2 According to Schuetze (2002), the highest under-reporting rates in Canada (of 38–53 per cent) are found in construction and service occupations, in which cash transactions are frequent and opportunities to evade tax are correspondingly high. Schuetze (2002) observes that under-reporting rates tend to decline with age and to vary from year to year, implying that in practice researchers should use multiple years of data when trying to estimate under-reporting rates (see also Baker, 1993). Also, under-reporting rates are lower in couples where both spouses are self-employed, compared with couples where only one spouse is. Income under-reporting factors can be estimated not only from survey expenditure and income data, but also from actual tax return income data, using data on charitable donations included on tax returns for the purpose of claiming tax deductions (Feldman and Slemrod, 2007). Assuming that the source of one’s income is unrelated to one’s charitable inclinations and that the ratio of true income to taxable income does not vary by income source, Feldman and Slemrod (2007) estimated a compliance rate of self-employment income in the USA of 55–71 per cent. Non-farm small business and farm incomes have even lower compliance rates; and Feldman and Slemrod (2007) estimated that many negative or zero reported incomes actually correspond to positive true incomes. To summarise, it is possible to estimate how much entrepreneurs under-report their incomes to the tax authorities and (most importantly for applied researchers) to survey interviewers. These two forms of income under-reporting are quite distinct, and it is not sensible to try to draw comparisons between them. Nevertheless, it is worth noting that the Pissarides–Weber approach might enjoy several advantages over TCMP data, since unlike TCMP it accounts for over-expensing as well as income under-reporting. Also, it is unlikely that TCMP tax auditors can trace every source of unreported income, especially those derived from cash transactions. And up-to-date TCMP data are unavailable to researchers anyhow. The central message for applied researchers who work with data on entrepreneurs’ incomes is that they should treat these data with considerable caution and apply under-reporting corrections to raw data whenever possible. As the estimates of income under-reporting rates cited above imply, failure to do so can bias downwards

368

Performance

substantially comparisons of average incomes relative to employees, an issue which will be explored further below. 13.2

Other measurement problems

As well as income under-reporting in sample surveys, there are at least five other problems associated with the practical use of sample data pertaining to entrepreneurs’ incomes. These include: 1. 2. 3. 4. 5.

High non-response rates by entrepreneurs to survey questions about incomes The treatment of incorporated vis-à-vis unincorporated businesses Truncation of income data by survey administrators The treatment of fringe benefits The separation of returns to capital from returns to labour I will now consider each of these problems in turn.

1. High non-response rates by entrepreneurs to survey questions about incomes. This problem can be quite pervasive and can substantially bias downwards estimates of entrepreneurs’absolute and relative returns. For example, Devine (1995) calculated that self-employed income non-response rates were about 30 per cent (CPS data, 1976–91), compared with 17–19 per cent for wage and salary workers. In related work, Devine (1994a) estimated non-response rates of 25 per cent for the incorporated and 40 per cent for the unincorporated self-employed. There are several possible reasons for survey non-response. One is mistrust of survey interviewers by entrepreneur respondents for the reasons given in the earlier discussion. A second is that richer people (of whom a disproportionate number are entrepreneurs – see below) have a higher marginal valuation of time, so participate less in household surveys. Third, lacking payroll payslips and up-to-date accounts, many entrepreneurs simply do not know their incomes with any great accuracy. According to Meager et al. (1994), this factor accounted for two-thirds of missing British income cases in the 1991 BHPS – rather than refusal to co-operate with the survey. It is reasonably straightforward to correct self-reported income data for survey nonresponse. If reliable data on the frequency distribution of entrepreneurs’ incomes are available, the sample can be re-weighted to restore representativeness and thereby deal with the problem. 2. Different ways of treating owners of incorporated businesses. Incorporated selfemployed individuals are treated for tax purposes as employees of their company. They earn higher incomes than the unincorporated self-employed on average – about twice as much in the USA (SBA, 1986) and somewhat less than twice as much in the UK (Ajayiobe and Parker, 2005). Therefore their exclusion from samples of entrepreneurs may bias downwards calculations of the average income of the self-employed. On the other hand, their inclusion in samples which ask about incomes might mix up individuals declaring profit (the unincorporated self-employed) with individuals declaring ‘Draw’ (the incorporated self-employed). For this reason, it can be sensible in practice to define income as Draw for both types of entrepreneur.

Entrepreneurs’ incomes and returns to human capital

369

3. Truncation of income data by survey administrators. Some data-set administrators truncate very high earnings to some maximum number. This is called ‘top coding’. At the other end of the income distribution, negative incomes are sometimes coded as zero. Clearly, the first practice imparts a downward bias to average income figures, while the second causes upward bias. Since entrepreneurs are disproportionately likely to be found in both tails of the income distribution, a bias of unknown sign and scale ensues. Ideally, of course, neither form of truncation would be applied to the raw data. 4. Ignoring employer-provided fringe benefits in income comparisons. This biases downwards any relative income advantage to employees relative to entrepreneurs. Some of these benefits can be substantial, especially occupational pension schemes and employer contributions to private health care plans in countries lacking publicly funded health systems (Holtz-Eakin et al., 1996; Wellington, 2001). 5. Self-employment incomes include returns to capital as well as returns to labour. National Accounts experts have long argued about how best to disentangle these returns. The lack of agreement about how best to do so might partly explain why relatively few researchers choose to separate them in practice (Christiansen, 1971; Carrington et al., 1996). A very simple method for deducting returns to capital is to subtract the interest rate multiplied by the sum of business wealth and real estate (Fairlie, 2004). Often, however, the researcher lacks data on business wealth. In such cases, Headen (1990) proposes a different remedy.3 On the other hand, entrepreneurial income data typically exclude public and private sales of businesses which are also part of the total return to entrepreneurship. This leads to a downward bias in entrepreneurial incomes, offsetting the upward bias caused by including returns to capital. Unfortunately, few researchers in practice address these data problems in much depth. Income under-reporting has received the most attention, but even here the coverage has been patchy. Of the various sources of bias in estimates of relative entrepreneurial incomes discussed in this and the previous section, income under-reporting is likely to be the greatest source of bias (downwards) in absolute magnitude, followed by omission of fringe benefits (upwards), with non-response bias (downwards) coming in third. The offsetting biases relating to capital make the sixth source of bias hard to evaluate, while the other two sources of bias are likely to be relatively minor. Therefore, on balance, studies which fail to deal with these sources of bias (especially income underreporting) and take the data at face value are probably likely overall to understate relative entrepreneurial incomes. Hence results derived from them should be treated with considerable caution, a point to which we shall return below. 13.3

Evidence relating to entrepreneurs’ relative incomes

Economic theory predicts that workers will move between occupations to exploit income differences until occupational incomes are equalised. This prediction is qualified when one takes into account the heterogeneity of job characteristics and individual preferences, including non-pecuniary compensating differentials, occupation-specific

370

Performance

risk and over-optimism. As will be seen below, few researchers have reported equality between average incomes in entrepreneurship (usually measured as self-employment) and paid employment. Most relative income comparisons for entrepreneurs have been performed using US data. The evidence here is inconclusive, with a roughly equal split of studies finding higher and lower average incomes in self-employment compared with paid employment.4 These studies usually control for personal characteristics but make few if any adjustments for income mismeasurement. Fairlie (2005a) points out that these relative income comparisons can be sensitive to whether income is measured in dollar amounts or in logs. Male entrepreneurs are often found to earn more than male employees in the former case but roughly the same in the latter case. Two studies are of particular interest. The first is by Hamilton (2000), who analysed median rather than mean incomes to avoid distortions arising from the pronounced skew in the distribution of self-employment incomes (see below). Using 1984 SIPP (Survey of Income and Program Participation) data, Hamilton estimated that on average all individuals except those in the upper quartile of the self-employment income distribution would have earned more, and enjoyed higher future income growth rates, if they had switched into paid employment. For individuals in business for ten years the median earnings differential was 35 per cent in favour of paid employment. Hamilton argued that these findings were unlikely to reflect selection effects, since the wage distribution of employees switching into self-employment resembled that of employees remaining in paid employment. Also, these findings were found to hold using all three definitions of self-employment income described at the start of the chapter. However, this study did not correct for income under-reporting by the self-employed, which can be substantial, as noted earlier. Furthermore, simple cross-section comparisons like this one cannot capture an important incentive to become an entrepreneur, which is to accumulate assets rapidly. That is, lower measured incomes can be compensated by higher lifetime wealth (Bohàc˘ ek, 2006). The second study of interest, by Fairlie (2005a), also did not correct for income underreporting yet still found higher mean and median incomes for the self-employed, the differences being $14,042 and $4,979, respectively (NLSY data). The self-employment income advantage was found to hold also for young men (though not young women) from disadvantaged families, defined as families where both parents had less than a high-school education (Fairlie, 2005b). This might suggest that self-employment is a route out of poverty for disadvantaged youth. It can be important to distinguish between self-employees who employ others and those who work on their own account. Alba-Ramirez (1994) for Spain, Prottas and Thompson (2006) for the USA, and Ajayi-obe and Parker (2005) for the UK all observed that employees earn less on average than the employer self-employed but more on average than the own-account self-employed. If the former are regarded as more ‘entrepreneurial’ than the latter, then the evidence is again consistent with the notion that entrepreneurship provides higher returns on average than paid employment.

Entrepreneurs’ incomes and returns to human capital

371

Some evidence points to a downward trend in relative entrepreneurial incomes. Using US Social Security data from 1951 to 1988, Aronson (1991) showed that a 48 per cent income advantage to the self-employed in 1951–4 had dwindled to a 23 per cent advantage by 1975–9. This became a 10 per cent disadvantage by 1980–4, widening to a 20 per cent disadvantage by 1985–8. It appears that a similar trend applies irrespective of occupation and education (SBA, 1986). Adjusting for the longer average work hours of the self-employed reduces further their relative income position – by around 70 per cent according to Aronson. The UK also witnessed a downward trend in relative average self-employment incomes since the 1970s as their numbers grew (Robson, 1997; Clark and Drinkwater, 1998). Overall, UK evidence mostly points to a relative income advantage to selfemployment. Disagreement centres on how large this advantage is. GHS (General Household Survey) micro-data suggest a small premium to self-employment of 7 per cent over 1983–95, according to Clark and Drinkwater (1998) (see also Meager et al., 1996). In contrast, aggregate UK National Accounts data indicate a larger difference, of 35 per cent in terms of pre-tax gross income in 1993 according to Robson (1997). It may be relevant that the latter, unlike the former estimate, includes an adjustment for income under-reporting by the self-employed. Recent evidence suggests that the self-employed possess an income advantage over employees in some European countries, too. According to Blumberg and de Graaf (2004), the Dutch self-employed in the 1980s and 1990s enjoyed an income advantage over their employee counterparts of around 40 per cent. Both OECD (1986) and Constant and Shachmurove (2006) have estimated that self-employed West Germans earn more than their salaried counterparts, with the Constant and Shachmurove study finding that immigrants to Germany fare the best in this respect. Further probing into the data reveals some interesting dimensions of the recent German experience compared with the USA. McManus (2000) used US PSID and German SOEP (Socio-Economic Panel) data to compare outcomes of the self-employed compared with employees. She reported that the self-employed in these countries are no worse off than employees in terms of earnings and job stability. Remunerative self-employment opportunities are especially plentiful in the USA but so too are low-paying and unstable self-employment jobs compared with Germany. McManus went on to establish that the self-employment average earnings advantage is largely explained by greater average stocks of human capital and longer work hours. In the transition economies of Russia, Slovakia, Bulgaria, Hungary and the Czech Republic between 1988 and 1993, the mean income of employees was less than that of own-account self-employed individuals, which in turn was less than the mean income of self-employed employers (Earle and Sakova, 2000). Only Poland bucked the trend. Nevertheless, the self-employed are not well remunerated relative to employees in some other developed countries, with Finland, Sweden and Japan exhibiting selfemployed/employee income ratios below that of the USA (OECD, 1986). Similar evidence was also chronicled by Covick (1983) and Kidd (1993) for Australia; and

372

Performance

Covick (1983) noted the same downward trend in relative self-employment incomes as observed in the USA and the UK. In summary, the evidence does not point to a clear-cut income advantage to entrepreneurship in all countries. But nor does the available evidence suggest that entrepreneurs clearly earn less than employees, as is sometimes claimed. There is a pressing need for further research to enhance the current state of knowledge by using high-quality data while controlling for the most serious measurement problems described above. 13.4 The inequality and volatility of entrepreneurs’ incomes

In most countries the incomes of entrepreneurs are more unequal than those of employees. This fact usually becomes immediately obvious when histograms of incomes are graphed separately for employees and entrepreneurs.5 Higher proportions of the self-employed earn both very low and very high incomes compared with employees. Consequently, when data from the two occupations are combined, the self-employed are found to be disproportionately concentrated in both the upper and lower tails of the overall income distribution. In their British analysis based on 1991 BHPS data, Meager et al. (1996) showed that this result is not an artefact of sampling error, and remains after controlling for observable characteristics such as gender, work status, work hours, age, education, industry and occupation. It might be instructive to consider each of the two tails of the income distribution separately. The concentration of self-employed people in the lower tail of the income distribution is a widespread phenomenon. For example, SBA (1997) reported that onesixth of self-employed people earn less than the minimum wage. The fraction is even higher in developing countries, being two-thirds for micro- and small enterprises in Kenya for example (Daniels and Mead, 1998). Exacerbating the problem, the selfemployed often lack access to social security benefits (Hakim, 1989a). Although one cannot rule out the possibility that low incomes reflect widespread part-time work arrangements, under-reporting and the voluntary reinvestment of profits, it is probable that most low incomes derive primarily from low levels of skill. For instance, according to Amit et al.’s (1995) study of 55,434 respondents from the 1988–90 Canadian Labour Market Activity Survey, employees who switched into self-employment earned $2,340 less, on average, than employees who did not switch.6 There is also a link with unemployment. Long unemployment durations have been found to make workers accept less desirable jobs. So studies finding that unemployment duration significantly increases the hazard of entering self-employment imply that those self-employed jobs may be less desirable (Alba-Ramirez, 1994). Further evidence that few unemployed people become successful entrepreneurs appears in the Swedish study of Andersson and Wadensjoe (2007). It is less obvious why so many entrepreneurs are observed in the upper tail of the income distribution. Roy (1951) argued that if income-earning opportunities are more dispersed in one occupation (entrepreneurship, say) and if ability generates income,

Entrepreneurs’ incomes and returns to human capital

373

then the ablest will rationally select into the occupation with more dispersed returns. Therefore the highest abilities and incomes will be observed in entrepreneurship. However, this begs the question of why incomes are more dispersed in entrepreneurship in the first place. An important theoretical contribution in this regard is Rosen’s (1981) theory of ‘superstars’. This theory explains why the most talented individuals command salaries that are disproportionately higher than those of their nearest rivals. Consider by way of example top concert pianists. Consumers’scarcity of time makes them willing to pay a premium to see the world’s best rather than the world’s second-best pianist. This imparts convexity into the return function and translates small differences in underlying ability into large differences in payoffs. This is also relevant in entrepreneurship where returns to individual ability are often taken to be greater than in paid employment (where in contrast productivity often reflects team performance: see chapter 2). For these reasons, the upper tail of the income distribution becomes disproportionately filled with entrepreneurs. Another theory, which applies to the entire distribution of entrepreneurs’ incomes, is outlined in Parker (1997b). Parker proposes a stochastic parametric model of the entrepreneurial income distribution, which combines several stylised facts about income dynamics and firm size growth processes. This model ultimately derives the Pearson Type VI distribution as a parametric form for the income frequency distribution.7 The Pearson Type VI is a unimodal and positively skewed distribution, which seems to provide a good representation of self-employment income inequality. It also has the advantage of being able to handle zero and negative incomes, an important property given that over 6 per cent of UK self-employees had zero or negative incomes in 1991 (Parker, 1997b). Pronounced inequality of entrepreneurs’ incomes is not only observed by looking at its frequency distribution, but also by mapping sample data into scalar inequality measures, such as the Gini coefficient or the mean log deviation.8 Parker (1999b) computed a range of inequality measures using UK FES data over 1979–94/5, and found that inequality indices for the self-employed were between two and five times greater than those for employees, depending on the year and the particular inequality index chosen. This study corrected for self-employment income under-reporting and deployed inequality measures which are robust to mean-preserving measurement error. The UK context is particularly interesting because self-employment income inequality grew especially rapidly in that country in the 1980s, the same decade when self-employment itself expanded substantially (see chapter 1). The increased self-employment income inequality in the 1980s was large enough for Jenkins (1995) to cite it as the primary reason for the growth of overall UK income inequality in the 1980s, exacerbated by a greater self-employment population share. The opposite story seems to apply in the cases of the transition economies, however, where mass layoffs caused by the appearance of hard budget constraints, privatisation and restructuring of state-owned firms ensured that new and small firms became the primary sources of job creation. Together with the greater productivity of new small firms in these economies and the greater competition they stimulated, expanded

374

Performance

entrepreneurship helped to reduce inequality. Thus Berkowitz and Jackson (2006) estimated a significant and sizeable positive impact of new enterprise formation on the relative and absolute incomes of the poorest 40 per cent of earners in Poland and Russia. Yet serious gaps in our understanding about the underlying causes of entrepreneurial income inequality remain. For example, Parker (1999b) found that none of age, gender, marital status, region, job type, work status, or educational qualification explains more than a small fraction of UK self-employment income inequality levels or trends. In contrast, these variables can explain a sizeable share of employee income inequality. Parker concluded that these results reflect the marked heterogeneity of the self-employed, heterogeneity that seems to have increased in Britain and other countries in the 1980s along some unobserved dimensions (Arum and Müller, 2004). Two issues remain to be discussed before moving on to the individual-specific determinants of entrepreneurs’ incomes. The first is the distinction between income inequality and volatility. The second is social mobility, which is related to income mobility. This is an especially important issue as it has been argued, for example, that ‘business is in America the most effective form of social mobility for those who meet prejudice’ (Glazer and Moynihan, 1970, p. 36). It is also obviously relevant to people born into any kind of disadvantage. Volatility and inequality are terms which carry distinct meanings.9 Income inequality is a cross-section population concept, whereas income volatility relates to individualspecific temporal variations. US evidence clearly points to greater income volatility among entrepreneurs than among non-entrepreneurs. Carrington et al. (1996) report that median hourly wages and annual incomes in self-employment are substantially and significantly more volatile and pro-cyclical than those of employees; and that while economic downturns are associated with lower wages for all workers, declines for the self-employed are some three or four times greater on average than for employees. According to Heaton and Lucas (2000), business income is twice as variable as wage income, and is much more highly correlated with stock market returns than wages are. Evidence of greater income volatility despite equivocal evidence about a mean income advantage begs the question of whether risk premiums in entrepreneurship are adequate or are in some sense ‘too low’. Over-optimism is one reason why there may be excess entry into entrepreneurship, bidding down the risk premium (see chapter 4). Another explanation is imperfect competition, which biases occupational choices towards firm ownership, resulting in too many firms remaining in the market and causing the profits of all firms to decline (Clemens, 2006). A further possibility might be that entrepreneurs disregard risk or do not frame contingent outcomes in terms of risk (Palich and Bagby, 1995). For example, survey evidence suggests that self-employed Americans do not perceive greater levels of job insecurity than employees (Blanchflower, 2004). They may even perceive lower levels of insecurity than employees (Manski and Straub, 2000). Finally, I will briefly consider income mobility. In view of the subtleties surrounding entrepreneurial payoffs discussed so far, it is probably unsurprising that

Entrepreneurs’ incomes and returns to human capital

375

researchers have obtained conflicting findings about the influence of entrepreneurship on income mobility. For example, while Gentry and Hubbard (2000) found that entrepreneurs enjoy greater upward mobility in the income and wealth distributions than non-entrepreneurs do, with those leaving entrepreneurship experiencing downward mobility, a more nuanced picture comes from other research. According to Holtz-Eakin et al. (2000), becoming self-employed increases upward earnings mobility for low-income Americans (keeping characteristics constant), relative to remaining as employees. But Holtz-Eakin et al. (2000) found the opposite for high-income Americans, for whom self-employment tends to entail downward mobility.10 Case study evidence also casts doubt on the notion that entrepreneurship generates rapid social mobility (Bhide, 2000). According to Bhide, most of the successful founders of Inc.500 firms have college degrees and come from middle-class backgrounds; very few are born into poverty. Meager and Bates (2004) observe that most transitions from paid employment to self-employment in Britain take place within the same broad occupational group, with professional employees becoming professional self-employees, skilled employees becoming skilled self-employees, etc. Because productivity is linked with skill, the outcome is limited social (occupational) mobility. For their part, sociologists sometimes argue that entry into entrepreneurship is easier if one is born into a high social class, with favourable social networks being conducive to resource acquisition (Müller and Arum, 2004). Immigrant entrepreneurship may be related to income mobility, especially intergenerational income mobility. It is commonly claimed that immigrant entrepreneurs work extra hard to enable their children to become financially successful, often as professionals in paid employment. Yet D. Y. Kim (2006) provides evidence from a small sample of Korean immigrants which shows that the children of entrepreneurs fare no better in life outcomes than children of professionals. The offspring of immigrant entrepreneurs can, however, get stuck in a low-income ‘mobility trap’ if the demands of working in the family business ultimately hold back their life and career prospects. 13.5

Determinants of entrepreneurs’ incomes: theory and methods

This section asks whether it is possible to explain individual entrepreneurs’ incomes in terms of personal and economic variables. The most popular method for investigating this issue involves estimating ‘earnings functions’. Earnings functions were originally developed by labour economists to explain the determinants of earnings in paid employment. An earnings function typically regresses log earnings, ln y, on a set of explanatory variables that includes age or experience, a, years of education, s, and a vector of other personal and family characteristics, X . Let u be a stochastic disturbance term, and index individuals by i in a sample of size n. Employee earnings functions typically take the form ln yi = β0 + β1 ai + β2 ai2 + β3 si + γ
Xi + ui

i = 1, . . . , n

(13.1)

376

Performance

where the βs and γ are coefficients. The β3 coefficient measures the rate of return to an extra year of education (RTE), and for this reason is of particular interest to human capital theorists. In principle, it is a straightforward matter to estimate (13.1) using data for a sample of entrepreneurs. However, there are several reasons why one would expect the coefficients of (13.1), and their interpretation, to differ from those obtained using employee samples. The following discussion will focus on two major inputs in (13.1): schooling and age. It is commonly observed that some of the most successful entrepreneurs have relatively low levels of education. For example, Bill Gates and Michael Dell dropped out of Harvard and the University of Texas, respectively, while Sir Richard Branson dropped out of high school. These (rather exceptional) cases have been used to justify a popular view – sometimes encouraged and promoted by entrepreneurs (especially, one suspects, by those with modest schooling!) – that formal education is unrelated to entrepreneurial success and may even impede it. According to this view, while education is likely to have a directly productive role in paid employment, entrepreneurs’ success depends on numerous factors other than formal education, including salesmanship, informal skills (especially an ability to satisfy customers’fuzzy wants) and confidence (Brown and Sessions, 1998, 1999; Bhide, 2000). Formal education might even inculcate attitudes and behaviours, such as excessive pontificating, which are antithetical to entrepreneurship (Casson, 2003). Human capital theory offers another reason why entrepreneurs might receive a lower RTE than employees. This is based on the notion that employers use educational qualifications to screen workers who differ in terms of hidden ability. Consequently, the RTE for employees will include a screening component, unlike entrepreneurs, who do not have a prospective employer (Wolpin, 1977; Riley, 1979). As a result, entrepreneurs have fewer incentives to obtain formal education than employees. The screening idea has been taken even further by Orzach and Tauman (2005), who study optimal human capital investment in a model with two dimensions of ability: skill derived from schooling and innate entrepreneurial talent. Suppose the latter is only privately observable and has a greater marginal value when deployed in entrepreneurship. The former, which is publicly observable, is assumed to be more valuable in paid employment. Then individuals intending to become entrepreneurs can convey an informative signal to lenders about their innate ability by acquiring less formal education. They thereby obtain more favourable loan terms from lenders, since this strategy is only profitable for talented entrepreneurs. Individuals possessing only moderate entrepreneurial abilities will not find it in their interests to emulate the gifted individuals, because with their abilities they benefit most by taking more formal education and becoming employees. On the other hand, there are counter-arguments suggesting that entrepreneurs may enjoy a higher RTE than employees. First, formal education can help individuals to integrate and accumulate new knowledge and adapt to new situations (van Praag, 2005, p. 116). This is likely to be valuable in entrepreneurship as well as paid employment (Marshall, 1930). Second, it is not necessarily true that entrepreneurs are unscreened, since customers, suppliers of credit and government agencies may also

Entrepreneurs’ incomes and returns to human capital

377

screen entrepreneurs (Fredland and Little, 1981; Parker and van Praag, 2006a). The value of signalling, and hence the RTE, could therefore be similar for workers and entrepreneurs. Furthermore, when acquiring his or her education, a future entrepreneur might anticipate the possibility of working in paid employment, which makes human capital investment a valuable hedge (Parker, 2005a). Third, entrepreneurs might possess greater flexibility to employ their human capital in productive ways, unlike employees who are constrained by more inflexible formal workplace structures (van der Sluis et al., 2004). Like the RTE, the size of entrepreneurs’ rate of return to age relative to employees is also theoretically ambiguous. In terms of (13.1), a steeper profile implies a larger β1 and/or a smaller β2 coefficient. There are at least three reasons why the earnings–age profile of entrepreneurs could be steeper than that of employees. First, entrepreneurs do not share the returns of their costly human capital investments with employers, who smooth out these costs and returns over employees’ lifetimes. Second, if entrepreneurs learn about their abilities over time with the ablest surviving (Jovanovic, 1982), then the average returns of a cohort of entrepreneurs will increase over time. This mechanism does not apply to employees. Third, investment in physical capital reduces the earnings of young entrepreneurs, while the returns of that investment accrue to older entrepreneurs – again implying a steep earnings–age profile. On the other hand, if employees (agents) can shirk on the job, employers (principals) may respond by steepening employees’ earnings profiles in order to elicit appropriate worker effort (Lazear and Moore, 1984). Naturally, no such agency problem arises in entrepreneurship where the principal is the agent. And flat earnings–age profiles can emerge in entrepreneurship if entrepreneurs optimally make few on-the-job human capital investments (Kawaguchi, 2003).11 When estimating earnings functions for entrepreneurs, it is necessary to avoid two sources of possible bias. One is selection bias. As noted in chapter 3, individuals who become entrepreneurs are not randomly drawn from the population. For example, Roy (1951) recognised the possibility that individuals select occupations on the basis of their ability, the returns to which vary across occupations. If ability is partly unobserved, then failing to control for occupational selection can bias the coefficients of (13.1). Entrepreneurship researchers are increasingly recognising this problem and are starting to utilise sample selection corrections in their empirical work (see chapter 3 for details). Researchers estimating earnings functions also need to avoid endogeneity bias. As noted in chapter 3, if OLS is used to estimate (13.1) when s is endogenous, the estimate βˆ3 of the RTE may be biased. There are at least two possible reasons why s might be endogenous. First, individuals are likely to base their schooling investment decision, at least in part, on their perceptions of the expected payoffs to their investment (ln y) (Parker, 2005b). Second, there might be unobserved individual characteristics, such as ability and motivation, that affect both schooling levels attained and subsequent business performance.12 Endogeneity problems can be tackled in various ways. First, researchers can try to incorporate controls for unmeasured ability into (13.1). Second, akin to a ‘natural

378

Performance

experiment’, they can try to exploit unexpected changes in education policy as exogenous determinants of schooling choices. A third strategy is to study samples of monozygotic twins, to control for genetic and environmental factors. The fourth, and most commonly used approach is to deploy Instrumental Variables (IV) techniques (see van Praag, 2005, chap. 8, for details). To the best of my knowledge, only IV has yet been used to robustly estimate entrepreneurs’ RTE, in cross-section (Williams, 2003; van Praag, 2005, chap. 8; Parker and van Praag, 2006a) and panel settings (GarciaMainar and Montuenga-Gomez, 2005). The next section discusses results derived from IV and conventional (but potentially biased) least squares estimators.

13.6

Determinants of entrepreneurs’ incomes: findings

This section deals with the following issues. First, I review results pertaining to entrepreneurs’ RTE. I start by reviewing the findings of meta-analyses based on potentially biased OLS regressions and then go on to interpret results which deal with endogeneity and sample selection problems. Second, I consider the role of other explanatory variables, chiefly age and experience; describe some extensions of entrepreneur earnings functions; and conclude with a brief summary about what we have learned in this chapter. 13.6.1

Entrepreneurs’ rate of return to education Numerous entrepreneurial earnings functions, both with and without corrections for sample selection and endogeneity biases, have been estimated, using a wide variety of data sources and explanatory variables. Rather than present a detailed and exhaustive classification of results, it will be more economical to summarise the results of two meta-analyses performed by van der Sluis et al. (2003, 2005). The first of these articles (2003) focuses on industrialised countries, while the second (2005) treats developing countries. The studies included in these meta-analyses generally utilise simple OLS estimation, so their findings are prone to bias. For industrialised countries, van der Sluis et al. (2003) report that 67 per cent of the studies in their survey established a significant positive estimate of β3 in (13.1), implying that entrepreneurial performance is significantly enhanced by schooling. They report an average RTE of 6.1 per cent for a small set of US observations which use similar measures of education and earnings. Thus an extra year of schooling increases enterprise income on average by 6.1 per cent. On the basis of twenty studies in the literature which compare entrepreneurs and employees using a single data-set and a common set of definitions, time period, country, and so on, they went on to establish that entrepreneurs’ average RTE is very similar to the average RTE for employees, being slightly lower than that of employees in Europe and slightly higher in the USA. Turning to developing countries, a meta-analysis comprising over eighty studies by van der Sluis et al. (2005) estimates the RTE for entrepreneurs to be 5.5 per cent, slightly

Entrepreneurs’ incomes and returns to human capital

379

lower than the 6.1 per cent estimated in developed countries. These authors also report that entrepreneurs’ RTEs in developing countries are higher for women, urban workers and in economies with larger agricultural sectors. A striking finding from Instrumental Variables (IV) estimates of (13.1) is that OLS dramatically underestimates entrepreneurs’ RTE. Consider two studies, by van der Sluis et al. (2004) and Parker and van Praag (2006a), which utilise data on American and Dutch entrepreneurs, respectively. Following the procedure set out in chapter 3, both studies performed Hausman tests and discovered that schooling appears to be an endogenous variable. Both studies then assembled a set of instruments drawn from respondents’ family backgrounds which significantly affect the number of years of education they choose but not their entrepreneurial performance. These instruments include father’s education and the number of siblings. These authors then went on to obtain IV estimates of entrepreneurs’RTE which were nearly double the OLS estimates, of up to 14 per cent. In the case of van der Sluis et al. (2004), this estimate compares with an IV RTE of 10 per cent for employees, implying that an extra year of education is more valuable for entrepreneurs than for wage workers. Van der Sluis et al.’s (2004) analysis was based on large samples drawn from the NLSY79, which contains annual data on earnings and other characteristics spanning nineteen years.13 The sample sizes were large too: n = 2, 952 for the entrepreneur (self-employed) sample and n = 47, 152 for the employee sample. Robustness checks revealed that this rather large occupational difference in the RTE could not be attributed to sample selection bias, mismeasurement of incomes, or several other factors. In summary, these results show that estimation methods which take account of the endogeneity of education imply large gains from formal education for entrepreneurs. However, other IV analyses using data from other countries – chiefly Germany and Spain – provide weaker findings (Williams, 2003; Garcia-Mainar and MontuengaGomez, 2005). For example, Garcia-Mainar and Montuenga-Gomez (2005) did not estimate a higher RTE for entrepreneurs than employees in their Spanish survey data. However, compared with the American and Dutch studies reported above, neither Garcia-Mainar and Montuenga-Gomez (2005) nor Williams (2003) performed thorough statistical testing to establish the validity, quality or relevance of their instrumental variables. According to most studies, on average entrepreneurs acquire similar levels of formal education and vocational training to employees.14 This outcome is not what one would expect from the classical employee screening hypothesis. Indeed, some direct evidence from IV estimates based on Dutch data suggests that entrepreneurs are also screened by lenders, with education reducing the incidence of Type I credit rationing (Parker and van Praag, 2006a). More generally, considerable care has to be taken when trying to attach screening interpretations to observed phenomena. For example, a lower R2 for entrepreneurs than employees in (13.1) does not necessarily imply that employees are screened whereas entrepreneurs are not – as Fredland and Little (1981) and Tucker (1987) have claimed. Low R2 s for entrepreneurs could be caused by a wide variety of factors relating to unobserved heterogeneity.

380

Performance

Turning next to sample selection bias, there is no general agreement about the sign and significance of sample selection effects. Thus there is no clear-cut evidence that entrepreneurs select into entrepreneurship because they enjoy a comparative earnings advantage there. For example, while Constant and Zimmerman (2006) detected significant positive self-selection into self-employment in Germany, van der Sluis et al. (2004) reported insignificant effects for the USA, with Wong (1986) detecting negative earnings selectivity among Hong Kong males. (See also sample selection evidence presented in chapter 4.) In contrast, many studies find positive selection effects for wage employment, implying that abler individuals would earn more if they became employees.15 It could be that the pronounced heterogeneity of entrepreneurs partly accounts for the lack of robust selection effects observed by previous researchers. However, to date relatively few entrepreneurship researchers have used the more sophisticated empirical sample selection methods which have recently become available. Co et al. (2005) illustrate the possibilities that these methods might be able to offer, by considering a second form of selection alongside that of occupational choice. This concerns whether individuals choose to participate in any occupation (i.e. being in the workforce) in the first place. They argue that this consideration is likely to be especially relevant in transition economies, where workforce participation rates tend to be lower than in developed economies. Note that most of the studies discussed above measure entrepreneurs as the selfemployed. But the self-employed are an unusually heterogeneous occupational group, and estimates of their RTE can be sensitive to the inclusion or exclusion of professionals in the sample (Fredland and Little, 1981; Tucker, 1987). This point should be borne in mind when estimating earnings functions using data on self-employed workers. 13.6.2

Other explanatory variables and extensions Several researchers have estimated a positive effect of age on entrepreneurial incomes.16 Most of the evidence points, however, to a flatter earnings–age profile for entrepreneurs than for employees. This finding is in line with the agency-cost model of Lazear and Moore (1984) mentioned above. Work experience in entrepreneurship has also been linked with higher returns in entrepreneurship, although there is some evidence that it is less valuable in paid employment than previous experience as an employee (Evans and Leighton, 1989b; Williams, 2000, 2003). Overall, the literature suggests that human capital is the major determinant of entrepreneurs’earnings (van Praag, 2005, p. 9). Few other explanatory variables, including ethnicity, family background, social capital, business strategy, or organisational structure of the venture, possess much explanatory power. In fact, most entrepreneurial earnings functions exhibit relatively poor goodness-of-fit: it is common to find R2 values of 10 per cent and less for estimates of (13.1). This is consistent with the poor performance of univariate decompositions of self-employed income inequality measures described earlier. In contrast, employee earnings functions tend to provide much better fits. Subsequently, researchers have tried including a range of additional variables into

Entrepreneurs’ incomes and returns to human capital

381

(13.1) in an effort to improve explanatory power. These include dummies for industries, disabilities, ill health and immigration status. But with a handful of exceptions (e.g. Honig, 1996, 1998), this approach has largely failed to improve matters much. The earnings function has been extended in several specific ways to address various practical issues in entrepreneurship. I consider four below. The first extension analyses the financial implications of the choice between becoming a franchisee or an independent business owner. Building on Lazear’s (2005) ‘entrepreneur as jack of all trades’ hypothesis, the second extension explores the impact of diversity of human capital on performance. The third extension raises the prospect that the education of one’s spouse might affect one’s payoff in entrepreneurship. The fourth asks how earnings functions can be extended to deal with individuals who mix work hours between entrepreneurship and paid employment (see chapter 12). The findings relating to these four extensions are tentative and more research is needed to build on them. First, Williams (1999) extended the analysis of earnings comparisons to ask whether a self-employed worker can expect to earn more by becoming a franchisee (F) or an independent business owner (I). Williams estimated selectivity-corrected profit functions in both F and I using data on 14,550 firms taken from the 1987 CBO data set. Selectivity effects were found to be important, with individuals in F earning less than those in I, after controlling for personal characteristics. Yet franchisees were also estimated to be substantially worse off if they were to become independent owners instead. These findings suggest that franchisees possess relatively low average ability, compared with independent business owners. A policy implication is that legal restrictions on franchising activities may therefore have deleterious effects on the material well-being of disadvantaged individuals. Second, Astebro and Thompson (2007) analysed a sample of 830 independent inventors and 300 individuals from the general population, and observed that entrepreneurs have a more varied educational and professional background than employees. Recall Lazear’s (2005) hypothesis that greater diversity of human capital is associated with a greater likelihood of becoming an entrepreneur. Astebro and Thompson’s (2007) question is somewhat different: does this diversity translate into superior performance in entrepreneurship? Astebro and Thompson concluded that diverse human capital has a negative effect on entrepreneurial incomes. This suggests that it may be possible to complete the JAT aphorism: entrepreneurs are ‘jacks of all trades but masters of none’. Third, Wong (1986) points out that a spouse’s education increases the earnings of workers and entrepreneurs in Hong Kong, especially entrepreneurs. The mechanism may be knowledge spillovers transferred between spouses (Parker, 2008a). Fourth, some individuals are observed to mix their work hours between entrepreneurship and paid employment. This practice appears to be especially widespread in developing countries, with Vijverberg (1986) reporting that some 20 per cent of respondents in a 1976 Malaysian survey data set performed work mixing; the percentage in developed countries is much lower (see chapter 12). If work mixing occurs, selectivity-corrected earnings functions become inappropriate, and one should instead estimate a composite earnings function of combined individual wages wi .17

382

Performance

To conclude, this chapter has reviewed the available evidence relating to entrepreneurial performance, measured in terms of income. Contrary to an apparently widespread belief in some quarters, we cannot conclude from the evidence that entrepreneurs earn lower incomes on average than employees do. It is true that a greater proportion of entrepreneurs earn very low incomes compared with employees; but there is also a higher proportion of entrepreneurs in the top tail of the income distribution. In terms of mean or median incomes, the bulk of previous work (including the influential article of Hamilton, 2000) has not paid sufficient attention to problems of income underreporting and other sources of mismeasurement, so it is far from clear what the relative average income position of entrepreneurs really is. In fact, those studies which have paid attention to measurement problems suggest that if anything entrepreneurs enjoy a modest income premium over employees. Whether this is sufficient to compensate them for bearing greater risk is not yet clear, although some attention has been paid to this issue with respect to their wealth – an issue discussed in chapter 9. Perhaps the most important finding to emerge from this chapter is growing evidence that entrepreneurs enjoy a high rate of return to their education. This finding is drawn from recent research using IV estimation methods which enable more reliable estimates of this rate of return to be obtained. But much more careful econometric modelling of entrepreneurs’ payoffs is needed to establish the robustness and generality of this finding. Notes 1. Since the self-employed account for around 10 per cent of UK GDP, this implies that the size of this part of the ‘black economy’ in the UK was about 5.5 per cent of GDP in 1982. 2. Besim and Jenkins (2005) applied the Pissarides–Weber approach to data from North Cyprus, and estimated that the self-employed under-report their incomes to the household survey by 11–14 per cent on average. 3. Let hj , yj and wj denote an individual’s work hours, total ‘labour’ income and wage rate respectively in sector j ∈ {E, S}, where E is paid employment and S is self-employment. Also, let yk denote returns to capital in self-employment. While hj and yj values are observed in most data sets, wj s are not and must be calculated (in E) or estimated (in S). We have yE = wE hE and yS = wS hS + yk . To estimate yk , first calculate wE = yE /hE and estimate an ‘earnings function’ (discussed later in the chapter) like ln wE = β
X + u, where X is a vector of personal characteristics, β is a vector of coefficients, and u is a random disturbance. This yields parameter estimates
ˆ Second, predict wˆ S for the self-employed from ln β. wS = βˆ
X . This can be taken as the return to self-employed labour assuming (i) that employee incomes are purely returns to labour, and (ii) that the self-employed have the same rates of return to personal characteristics as employees do (see below for a critical assessment of this assumption). Finally, denoting sample mean values by overbars, calculate yk /hS = yS /hS − wˆ S . Using this approach Headen (1990) estimated the returns to labour and capital as, respectively, 84 and 16 per cent for a sample of US self-employed physicians in 1984. 4. Studies finding higher average incomes in paid employment include Fain (1980), Becker (1984), SBA (1986), Haber et al. (1987), Carrington et al. (1996) and Hamilton (2000). In contrast, Form (1985), Borjas (1986), Evans and Jovanovic (1989), Ferber and Waldfogel (1998), Quadrini

Entrepreneurs’ incomes and returns to human capital

383

(1999), Rosen and Willen (2002), Fairlie (2005a) and Prottas and Thompson (2006) report an income advantage to self-employment on average, while Borjas and Bronars (1989, Table 1) report similar average incomes. 5. For examples based on UK data, see Curran et al. (1987), Hakim (1989a), Rubery et al. (1993), Goodman and Webb (1994), Meager et al. (1994, 1996) and Storey (1994a). Fairlie (2004) compares histograms of the earnings distribution for American self-employees and employees, and also provides an ethnic breakdown. 6. Other studies showing how higher wages deter moves into self-employment include Evans and Leighton (1989b), Dolton and Makepeace (1990), Boden (1999b) and Johansson (2000). 7. The Pearson Type VI density function is

f (y) =

(a + b) (y − γ )a−1 , (a)(b) (y + 1 − γ )a+b

y ≥ γ,

where (·) is the gamma function, a and b are shape parameters, and γ < 0 is the lower bound on entrepreneurs’ incomes (e.g. the maximum loss that entrepreneurs can carry before foreclosure). 8. For British evidence, see Rees and Shah (1986), Pissarides and Weber (1989), Dolton and Makepeace (1990), Jenkins (1995), Meager et al. (1994, 1996) and Parker (1997b, 1999b). For a Dutch example, see Nisjen (1988). 9. This can be demonstrated by considering the following simple model of log income among entrepreneurs i over time t: ln yit = β
Xit + µi + uit .

10.

11.

12. 13. 14. 15. 16.

(13.2)

Here Xit is a set of observable covariates specific to (i, t); µi is an individual-specific ‘fixed effect’ (see chapter 3); and uit is an idiosyncratic disturbance term. Variation of uit over time for a given i in the full conditional model (13.2) captures the notion of income volatility, or risk. This is distinct from inequality, which can be measured as the variation in the disturbance term ui of a special case of (13.2), namely where β = 0, µi = µ (constant) and t is suppressed: ui = ln yi − µ. These results might be prone to ‘regression to the mean’ bias, as noted by Holtz-Eakin et al. (2000). Nonetheless, van Praag and Versloot (2007) concluded their overview of the evidence by declaring that ‘average incomes seem to be of similar levels for entrepreneurs and employees in regression frameworks that allow the returns to a broad set of indicators of human capital to differ across entrepreneurs and employees. Entrepreneurship might be good for social mobility and becoming a “super income earner” (p. 375). These considerations suggest that earnings functions should be estimated separately for entrepreneurs and employees. It is not advisable to estimate an earnings function which pools data on the two occupations, such as ln yi = β
Xi + δzi + ui (as in, e.g., Amit et al., 1995), where zi is an indicator (dummy) variable for entrepreneur/paid employment status. While this specification looks attractive by providing a direct estimate of relative occupational earnings advantage via the parameter δ, it imposes the strong restrictions of identical rates of return to all of the explanatory variables for both occupations, which for the reasons given above are unlikely to hold. See Parker (2005a), van Praag (2005) and Parker and van Praag (2006a). See Fairlie (2005a) for a detailed description of self-employment data in the NLSY. See e.g. Wolpin (1977), Fredland and Little (1981), Robinson and Sexton (1994), Parker (1999b) and van Praag (2005, p. 125). See Evans and Leighton (1989b), Headen (1990), Maxim (1992) and Hamilton (2000). See e.g. Schiller and Crewson (1997), Burke et al. (2000) and Taylor (2001).

384

Performance

17. For example, estimate

ln wi = (1 − ξi )βE
XEi + ξi βS
XSi + ui ,

(13.3)

where ξi is the proportion of work hours i devotes to entrepreneurship; the βs are vectors of coefficients; and X is a matrix of observations on personal and occupation-relevant characteristics. Because ξi is likely to be endogenously chosen, (13.3) can only be estimated after an equation for ξi is specified in terms of a set of exogenous variables. This permits direct comparison of the two occupations’ βˆ coefficients.

14 Survival

This chapter discusses the survival of entrepreneurial ventures. This is a topic of key economic and social importance. At the outset, it is helpful to make some observations about some important definitional issues which are sometimes overlooked in both scholarly and popular discourse. First, the continuation of an individual in entrepreneurship is not necessarily equivalent to the survival of their business. For instance, serial entrepreneurs ‘survive’ in entrepreneurship even when their successive businesses do not. This can partly explain why individual exit rates from entrepreneurship tend to be lower than venture closure rates (Ronstadt, 1986). Second, entrepreneurs close businesses for a variety of reasons. Sometimes they do so in order to start new, more promising, ventures. In other cases, closure is a reluctant choice prompted by persistent financial under-performance. It is helpful to bear in mind that the determinants of venture prospects sometimes differ from the factors affecting entrepreneurs’ willingness to continue in business. Third, survival is sometimes treated as a measure of entrepreneurial performance. Unfortunately, this interpretation is problematic for several reasons. Some survivors scrape by on a pittance, while others deliberately exit a small niche profitably once more promising options appear. Indeed, some entrepreneurs deliberately plan to exit within a predetermined time to harvest their investment. As will be discussed further below, calling these closures ‘failures’ seems harsh, if not actually misguided. Survival is also closely linked to the issue of venture growth (see chapter 11). Most ventures have to grow to survive, especially early on when they are trying to establish themselves in the market. The evidence duly shows that growth enhances survival prospects (see below). But in turn, only firms which survive can grow, so researchers interested in identifying determinants of venture growth need to recognise that their data samples are conditioned on self-selected survivors. This provides another reason to study survival, since without correcting for sample selection (survivorship) bias in growth studies there is a risk that researchers might end up explaining the success of only successful firms, thereby overstating the positive impact of explanatory variables. The sample selection methods outlined in chapter 3 can help deal with this problem.1 385

386

Performance

The next section clarifies distinctions between different notions of closure, failure and performance. The key message emerging from this discussion is that it is essential not to conflate these notions. The second section presents some stylised facts about: the survival rates of new ventures; the related issue of quits from entrepreneurship; and the temporal distribution of venture closures. The third section outlines some theoretical perspectives on the determinants of survival. The fourth section summarises the body of empirical results on firm-specific, owner-specific, industry-specific and economy-wide factors associated with survival. The final section concludes. 14.1

Closure is not necessarily failure

Many researchers take it for granted that business closure can be associated, or even equated, with failure. The implicit assumption here is that entrepreneurs always want to keep their venture going and only exit reluctantly by some combination of bad luck, poor trading conditions, lack of resources or inadequate entrepreneurial capabilities. In fact, the raw data on venture closures and individual exits show that this assumption and interpretation is often inaccurate. For example, CBO data reveals that nearly 40 per cent of the American entrepreneurs whose ventures closed between 1992 and 1996 described their firm as having been successful at the time of discontinuation (Headd, 2003; Bates, 2005). According to Baird and Morrison (2005), only 10–15 per cent of US businesses which close file for bankruptcy (a popular definition of business failure: see below), while only 62.5 per cent of those that do file for bankruptcy are shut down and liquidated. Data on individual transitions tells a similar story. Less than 20 per cent of male entrepreneurs leaving self-employment in Britain do so because they face bankruptcy; about one-half quit in order to take another job (Taylor, 1999). Self-assessments of ‘successful closures’ may be upward-biased if people want to avoid admitting personal failure to survey interviewers, or if they seek other employment prior to an impending bankruptcy. Nevertheless, these figures are striking enough to debunk the popular belief that entrepreneurs only, or usually, exit entrepreneurship involuntarily. An important implication is that ‘the common practice of applying negative connotations uncritically to small firm closure – failure, death and the like – should be abandoned’ (Bates, 2003, p. 345). As intimated earlier, some closures can be positively beneficial. Whether an entrepreneur continues in business or quits does not simply depend on the financial performance of the venture, but also on whether its performance exceeds some personal threshold which the entrepreneur deems necessary to make continuation attractive (Gimeno et al., 1997; DeTienne et al., 2008). In terms of the simple occupational choice model of chapter 2, for example, an entrepreneur with some characteristic x remains in entrepreneurship if his or her expected profit, Eπ(x), exceeds a threshold value, w(x). Otherwise, they rationally (and optimally) close their firm and exit. Hence individuals with high thresholds may rationally quit entrepreneurship and close their businesses even though their expected performance in entrepreneurship may be above average. For instance, entrepreneurs who close a venture might move into even more profitable

Survival

387

fields, or simply retire (Ronstadt, 1986). Moreover, a voluntary exit does not necessarily imply that a venture failed to meet its owner’s objectives, since during its lifetime it may have yielded a handsome income to its owners. At the same time, entrepreneurs with below-average financial performance but even lower thresholds may rationally choose to continue and are categorised as ‘survivors’. US evidence based on NFIB (National Federation of Independent Business) data on 887 firms indicate that entrepreneurs of this type tend to be older; value independence; run firms with plenty of fixed capital; and have parents who owned their own businesses (Cooper et al., 1992). By not distinguishing between successful and unsuccessful exits, empirical researchers conflate both. At one level this is understandable: it is difficult, if not impossible, to classify successful as opposed to unsuccessful closures in practice (see below). But conflation of the two types of exit risks generating ambiguous empirical findings about the determinants of survival, because some explanatory variables (such as human capital) might both ward off involuntary failure by increasing E(π ) and promote voluntary closure by increasing w(x). These offsetting effects might cancel each other out, suggesting (wrongly) that human capital is unimportant to survival. Although most of the findings from the literature reviewed below are subject to this fallacy, some researchers have tried to address this problem directly. They have done so either by estimating a joint structural performance and threshold model, or by trying to classify some closures as voluntary and the rest as involuntary, before analysing the determinants of each separately. There are several structural models of performance and thresholds. Two mentioned in chapter 3 are Cooper et al. (1992) and van Praag and van Ophem (1995). The latter distinguishes between a willingness and opportunity to continue. Still others could no doubt be proposed. In general, the determinants of financial performance and survival differ. Cooper et al. (1992) claim that thresholds are more important for explaining survival than financial performance is. It has already been noted that efforts to classify closures into ‘successful’ and ‘unsuccessful’ cases can be hard to implement in practice. Watson and Everett (1993) discuss five different definitions of ‘failure’ and point out that none is unambiguously superior to any other, though there can be large empirical differences between them. The narrowest measure is ‘bankruptcy with loss to creditors’, which is rarely used for small companies and omits break-even closures. It is arguably too narrow to be useful. The broadest measures are ‘discontinuance for any reason’. Intermediate measures include ‘termination to prevent future losses’ and ‘failing to make a go of it’; but these can be difficult to operationalise in an objective way. Researchers who classify whether closures are successful or not can at least avoid self-serving biases inherent in selfclassification by the entrepreneurs themselves. On the other hand, researchers can be uninformed, making their constructs prone to measurement and interpretation errors. There have been some attempts in the empirical literature to distinguish between different types of closure.2 For example, Cueto and Mato (2006) estimate that about one-half of subsidised Spanish start-ups exit voluntarily, while the other half exit via bankruptcy. Everett and Watson (1998) observe that higher interest rates are associated

388

Performance

with a greater incidence of bankruptcies, while high unemployment rates are associated with more ‘discontinuations of ownership’. Bates (2005) estimates that entrepreneurs who lack industry experience and do not have a high-school degree are more likely to have unsuccessful closures. Individuals with limited human capital face both below-average performance and modest outside options w(x) if they exit. A summary of findings from the literature suggests that entrepreneurs from ethnic minorities are also more likely to have unsuccessful closures, although women are significantly more likely to have a successful rather than an unsuccessful exit. Perhaps surprisingly, one study found that heavy capitalisation at start-up is associated with unsuccessful closures. This is distinct from the impact of initial capital on closure for any reason (Cooper et al., 1994). The following conclusion emerges from this discussion. When analysing entrepreneurial survival, researchers should refer to ‘closure’ rather than ‘failure’, and should eschew pejorative connotations when discussing exit. Likewise, they should be careful when treating survival as a performance measure, since survival can be associated with low profitability among entrepreneurs with few alternative options. It is also important to recognise that entrepreneurs can learn from the mistakes of others, implying that some closures can convey social benefits (McGrath, 1999). For instance, evidence from US banking industry suggests that experience of local (in-state) bank failures or near-failures significantly reduces the failure rates of young banks, whereas non-local (out-of-state) failures or near-failures have insignificant effects (Kim and Miner, 2007). And exits of old Canadian manufacturing firms are associated with higher entry rates of new, more productive firms (Pe’er and Vertinsky, 2008). Market entrants can reallocate displaced labour, capital, knowledge and other resources (which are released when firms close) to more productive uses in new ventures. These considerations suggest that a positive exit externality may be at work.Although entrepreneurs who close their businesses following poor performance stand to lose money and self-esteem (Shepherd, 2003), other entrepreneurs can learn about what is possible from these outcomes and exploit the underlying opportunities more effectively. An interesting implication is that there may be less private investment in entrepreneurial firms than would be socially optimal, since entrepreneurs fearing involuntary future exit do not internalise the positive externality. 14.2

Survival rates and their distribution

It is well known that new firms have low survival rates, and that many people who enter entrepreneurship do not remain entrepreneurs for long. This fact holds regardless of whether entrepreneurship is measured in terms of self-employment or new venture creation, although there are some differences in the magnitude of survival rates depending on which of these measures is used. In terms of self-employment, approximately one-third of entrants exit within the first two or three years, in both the USA and the UK.3 There is an intriguing possibility that self-employment exit rates have exhibited pronounced temporal stability, with Nenadic (1990) reporting a three-year drop-out rate from self-employment in nineteenth-century

Survival

389

Edinburgh of around 40 per cent. Higher exit rates are observed during recessions, with 40 per cent of British self-employees failing to survive their first year in the early 1990s (Taylor, 1999). Survival rates tend to be higher when entrepreneurship is measured in terms of business creation. Phillips and Kirchhoff (1989) challenged in what they described as the ‘prevailing wisdom’ the claim that ‘4 out of 5 new [American] firms fail in the first 5 years’. Phillips and Kirchhoff (1989) reported more favourable survival rates, with 40 per cent of new American firms surviving for at least six years (see also www.sba.gov). Survival and employment growth rates are below average among new construction firms, and above average among ventures in service industries. New firm survival rates in some other countries are higher than US rates. For example, Pfeiffer and Reize (2000) reported one-year survival rates in Germany of around 90 per cent in the mid-1990s, while Brüderl et al. (1992) reported over 60 per cent survival rates in Germany after five years of entry. There are certainly marked cross-country variations in entrants’ business survival rates. According to OECD data, between 30 and 50 per cent of new entrants survive for seven years or longer, with the highest survival rates being found in France and Portugal, and the lowest in Finland and the UK (Fuest et al., 2002; Bartelsman et al., 2005). The distribution of closure times of new start-ups is positively skewed and inverse U-shaped. It takes time for entrepreneurs to learn that they lack managerial ability or face persistently adverse market conditions (Jovanovic, 1982; Frank, 1988). It also takes time for debt to accumulate to the point where creditors perceive that a firm is financially troubled and seek to foreclose. Hence there is an initial ‘honeymoon’ period of a year or two in which business closures are relatively infrequent (Fichman and Levinthal, 1991). When reality finally dawns, at between two and four years, closures reach a peak, after which their frequency decreases steadily with the length of time the firm has been trading.4 Mature businesses have low closure rates. US data on labour transitions show that after eleven years the probability individuals exit entrepreneurship falls to approximately zero.5 According to Cressy (1999), the distribution of failure times is temporally quite stable. 14.3 Theoretical determinants of survival

There are several major strands of theoretical analysis about the determinants of survival. One relates to venture age and size. Young and small firms are especially prone to closure, giving rise to the claim that entrepreneurial ventures suffer from a ‘liability of newness’ (or ‘adolescence’) and a ‘liability of smallness’. Other prominent determinants of survival include sunk costs, the intensity of competition and start-up team size. 14.3.1

The liabilities of newness and smallness Liabilities of newness and smallness are both consistent with the stylised fact of a positively skewed distribution of venture closure rates by venture age. After an initial

390

Performance

‘honeymoon period’, numerous young and small firms succumb to these liabilities and exit. The remaining firms grow to a scale which enables them to survive, resulting in progressively lower closure rates. What explains the liabilities of newness and smallness? There are several stochastic models of survival in which random shocks to successful innovation or profits separate losers (who exit) from winners (who remain and grow: Jovanovic and MacDonald, 1994; Cressy, 2006a). Two other classes of economic model, described in chapter 11, also prove useful for answering this question. The first set of models, associated with the work of Steven Klepper, proposes that firms which enter at mature stages of the industry life cycle tend to be small-scale imitators compared with large early innovators. Later entrants have higher costs than continuously innovating older incumbents. Later entrants become victims of industry shakeouts prompted by ongoing cost-cutting engineered by ongoing process innovations. Thus it is not venture age per se, but rather a characteristic correlated with it (namely innovation) which explains longevity.6 Only entrants that can find an innovative strategy that helps them exploit profitable niches in mature industries are able to compete with the large early innovators (Agarwal, 1997). The second set of survival models stresses the role of learning. According to Jovanovic (1982), firms run by entrepreneurs who learn that their abilities do not justify costly expansion of production remain small and vulnerable to exit. Older firms are larger and are run by entrepreneurs who have established that their ability is high: they are the self-selected survivors (Ericson and Pakes, 1995; Pakes and Ericson, 1998; Li and Weinberg, 2003). Banks also learn about the abilities of their entrepreneur clients, diminishing the amount of asymmetric information to which they are initially subject. As a result, able entrepreneurs running older ventures receive lower interest rates. With cheaper capital, these entrepreneurs have incentives to grow their ventures more to attain a larger scale, which further enhances their survival chances (Brito and Mello, 1995). Hence age and size promote survival for this reason too. Asymmetric information in credit markets provides a compelling explanation for the liability of smallness, if abler entrepreneurs starting larger ventures can signal their ability by committing high levels of sunk cost investments at the time of start-up. Turning to non-economic models of survival, organisational ecologists argue that firms need legitimacy (i.e. credibility and acceptance by customers and suppliers) to survive. Lacking legitimacy, new and small firms are especially vulnerable to failure (Freeman et al., 1983). Certification might be one constructive strategy which new firms can use to gain legitimacy and enhance their survival prospects (Shane and Foo, 1999). On a different tack, theorists in this tradition also claim that in industries characterised by rapid technological change, firm age can have a negative effect on survival, giving rise to a ‘liability of obsolescence’ (or ‘senescence’). Reasons include capital vintage effects and structural inertia of successful early entrants in the face of changing market conditions.7 Some sociologists also explain the phenomenon of declining entry rates as industries mature (as in, e.g., brewing and telephony in the USA: Barnett, 1997) in terms of the emergence of a ‘dominant design’, which chokes off new experimental entry (Baum et al., 1995).

Survival

391

14.3.2

Other factors Most industries are characterised by high rates of entry and exit. As Geroski puts it, ‘entry appears to be relatively easy, but survival is not’ (1995, p. 23). Hence barriers to exit, such as the loss of sunk investments, might offer a more convincing explanation of industry variations in survival rates than barriers to entry. The theory of investment with sunk costs under uncertainty suggests that the probability of early exit should be lower in industries with higher sunk costs, since only capable entrepreneurs would be willing to sink costs irreversibly into a venture (Dixit, 1989; Hopenhayn, 1992). The same logic suggests that industries where sunk costs are relatively low, such as services, will be associated with high rates of exit. A different theory relates to the intensity of competition. Economists argue that more intense competition leads to lower survival prospects for new entrants. Sociologists argue, however, that the relationship between the number of competitors and survival of new starts can be more complex. The ‘organisational ecology’ literature hypothesises that the number of firms already in the industry has an inverse U-shaped effect on entry and survival of new firms (Barnett, 1997; Carroll and Hannan, 2000). When there are few firms initially, a greater number of firms signals opportunities and confers information that reduces uncertainty and confers legitimacy in the eyes of outside stakeholders; social ties between the firms increase. But greater competition for resources eventually turns the impact of greater numbers of firms on probabilities of entry and survival negative. Another issue is the size of entrepreneurial teams. Shane (2003, p. 237) cites several advantages to starting with a large venture team: it increases the depth and breadth of the information pool available to a venture; combines talent, resources and professional contacts; and leverages trust and emotional support (Barkman, 1994). Hence one might expect larger team starts to be associated with superior survival outcomes, all else equal. Similar arguments based on superior access to finance, management expertise and other resources lead one to expect that de alio entrants (i.e. entry by an arm of an established firm) will survive longer on average than de novo (i.e. new) entrants (Mata and Portugal, 2002). 14.4

Empirical determinants of survival

There is now an extensive body of empirical research on the individual-, firm- and industry-specific determinants of entrepreneurial survival and exit. Rather than attempt to provide an exhaustive survey of every study, I summarise below the key determinants of survival for entrepreneurs and their ventures in developed countries. As noted in chapter 3, hazard models are statistically superior to binary models such as logit or probit, since they deal explicitly with censored observations and can adopt flexible survival functions consistent with the inverse U-shape survival–age relationship noted above (Mahmood, 2000). In practice, however, the basic conclusions arising from estimates tend to be fairly similar by econometric method, so I will not distinguish explicitly between them in what follows.

392

Performance

Table 14.1. Summary of economic determinants of venture survival Explanatory variable INDIVIDUAL LEVEL Age of entrepreneur Education of entrepreneur Self-employed parent(s) Unemployed/unemployed spell Time in business/venture age Industry experience Work experience Management experience Entrepreneurial experience Manual occupation Married; spouse an entrepreneur Female Minority/immigrant FIRM LEVEL Initial size/capital Business partners Multiple establishments Previous growth/regional or macro growth rate INDUSTRY & MACRO LEVEL Industry MES Innovator/innovative sector Intensity of competitiona Urban environment Unemployment rate Interest rate b

No. +

No. 0

No. −

23 24 4 0 26 5 3 2 6 0 4 0 0

2 8 7 7 3 0 2 3 1 0 4 1 2

5 3 1 4 2 0 2 0 2 2 0 4 3

30 4 6 16

6 1 1 0

3 0 2 1

1 6 1 0 1 1

2 0 1 1 0 2

7 1 14 3 4 7

Notes: +, − and 0 denote significantly positive, significantly negative and zero (insignificant) coefficients, respectively. Only multivariate studies (i.e. those including controls for other explanatory variables) are included; descriptive studies are excluded. MES is ‘minimum efficient scale’.8 a Measured in various ways, including low industry concentration, high industry entry rates, low industry price–cost or profit margins, numerous competitors in the industry, etc. b Sources are: Shepherd (1984), Wadhwani (1986), Reid and Jacobson (1988), Storey et al. (1987), Simmons (1989), Hudson (1989), Hudson and Cuthbertson (1993), Keeble et al. (1993), Audretsch and Mahmood (1995) and Kim and Miner (2007).

Table 14.1 summarises estimates of the determinants of survival, based on statistically significant (at 5 per cent) coefficients appearing in studies utilising multivariate analysis (since these best control for omitted variable bias). Given the substantial differences in numbers of control variables used, a simple tabulation provides an adequate, if ‘rough and ready’, flavour of the general findings.

Survival

393

Taking the individual level first, Table 14.1 shows that the entrepreneur’s education, age, and duration in entrepreneurship are the major determinants of survival at the individual level. The last two of these variables are strongly associated with venture age, which is therefore combined with entrepreneur’s age in Table 14.1. With the exception of prior industry experience,9 most types of experience outside entrepreneurship do not enhance survival chances. Experience of unemployment does not promote survival, with a handful of studies finding significant negative impacts. For example, there is evidence that starting a business from a state of unemployment triples exit hazard rates compared with starting from paid employment (Carrasco, 1999; Pfeiffer and Reize, 2000). So despite the high entry rates from unemployment (see chapter 4), many of these entrants are likely to have rusty human capital, lowerquality information about business opportunities, and possibly also lower motivation – culminating in significantly lower survival rates. Referring back to chapter 4, parental self-employment experience is another factor associated with high rates of entry to entrepreneurship. However, Table 14.1 shows that this variable does not clearly assist survival. Female entrepreneurs do tend to have higher exit rates than men, consistent with the discussion in chapter 6, although this claim is derived from a relatively small number of studies. The most important determinants of survival at the firm level are the venture’s age and size. Both have generally positive impacts. This is consistent with the theoretical arguments about the liabilities of newness and smallness discussed earlier. Larger firms are also more likely to be incorporated and have multiple establishments, which might be why these variables are also associated with survival. Founders with industry-specific knowledgeandexperiencetendtostartlargerbusinesses.Thisshouldremindusthatmany firm attributes are actually endogenous, being functions of the entrepreneurs’ attributes (Mata and Machado, 1996; Colombo et al., 2004).With this caveat, there is also some limited evidence that larger founding teams are associated with superior survival outcomes, a finding which is consistent with the theoretical arguments outlined earlier. Whether or not venture size is measured at current or initial (start-up) values, the finding that larger firms have superior survival prospects must be placed in context. Venture size is most pertinent when it is considered in relation to the minimum efficient scale (MES) in the industry of which the firm is a member. As Table 14.1 shows, new firm survival prospects tend to be less favourable in industries where scale economies are important (Audretsch, 1991; Audretsch and Mahmood, 1995). One of the notable findings in Table 14.1 is that growth trajectories promote survival (Phillips and Kirchoff, 1989; Bednarzik, 2000). Rapid early growth is often needed to eliminate cost disadvantages of starting with sub-optimal stocks of capital and know-how. Firms which grow rapidly can more rapidly obtain the MES and become competitive, reducing their probability of closure (Audretsch, 1995b). In contrast, firms which grow slowly remain uncompetitive and are vulnerable to withdrawal of resources by lenders and suppliers. Evidence that growth rates of survivors are higher, and survival rates are lower, in industries with substantial economies of scale supports these arguments (Audretsch and Mahmood, 1994; Audretsch, 1995b).

394

Performance

In empirical work, firm size is often (but not always) measured in terms of amounts of capital. That is why size and capital have a combined entry in Table 14.1. Capital promotes profitability and rapid growth; and most studies report a positive association between survival in business and access to capital.10 Self-finance and debt finance are both conducive to survival (Reid, 1991; Bates, 1994b). Self-finance is a positive indicator of venture quality (Repullo and Suarez, 2000) and carries less risk exposure and lower debt servicing costs. However, as a group, exclusively self-financed entrepreneurs have relatively low probabilities of survival. This group includes entrepreneurs who operate projects which are too risky to attract external finance (de Meza and Webb, 1988), as well as unrealistically over-optimistic entrepreneurs who naturally prefer to use their own funds rather than those of banks (de Meza and Southey, 1996). Household income also matters: according to Lin et al. (2000), having a spouse who is an entrepreneur reduces the likelihood of exits from self-employment in Canada through the provision of a steady stream of family income. Numerous studies in the business and management literature have identified strategies which promote survival. These include innovation, which according to Table 14.1 has broadly positive effects on survival prospects. For instance, it has been found that young firms which innovate in science-based and specialised-supplier sectors survive on average for 11 per cent longer than rivals that do not innovate (Cefis and Marsili, 2006). It is also informative to disaggregate innovations by their type. Cosh et al. (1999) report that product innovations in Britain significantly increase the probability that a small firm is acquired, while process innovations significantly decrease the probability of small firm closure. Other venture strategies which affect entrepreneurial venture survival (not tabulated in Table 14.1 because there are too few studies) include alliances, product focus, marketing, R&D and franchising. For example, Brüderl et al. (1992) report that failure rates are lower for firms which form alliances with existing businesses. Ventures with a tighter product focus tend to enjoy higher survival rates.11 The evidence also broadly supports the notion that market size enhances survival prospects.12 Even though local markets can enable cash-strapped new ventures to occupy niches which shield them from competition from larger and more powerful incumbents, serving as a springboard for more ambitious subsequent initiatives (Bhide, 2000), the fact remains that local markets are smaller and offer fewer growth opportunities, hampering overall survival prospects. The evidence on franchising is more mixed. While Shane and Foo (1999) claim that 70 per cent of US franchise arrangements survive for twelve years and more, other evidence suggests that franchisees have similar rates of survival to other ventures, unless they work in tandem with a large corporate franchisor.13 A business plan has positive effects on survival according to Tamásy (2006). Some industry sectors have higher new venture survival rates than others. Especially noteworthy is that new service sector ventures have higher average odds of survival than those founded in manufacturing (Bednarzik, 2000). In the service sector, ventures providing education and health services are the most likely to survive, while those in transportation, retail and restaurants are the least likely to.14 Young ventures in the food,

Survival

395

beverage and accommodation sectors are more likely to go bankrupt than otherwise similar ventures located in different sectors (Thornhill and Amit, 2003). Industry factors complicate the relationships between survival, venture size and venture age. Simple positive relationships between these variables do not necessarily hold at mature stages of the product life cycle, or for technologically intensive products (Agarwal and Gort, 1996; Agarwal and Audretsch, 2001). Most industry life cycles are characterised by periods of rapid growth and new firm entry, followed by a slowdown as the industry matures. Entrants to new or technologically fast-changing industries can obtain a survival advantage by exploiting radical new ideas,15 whereas in mature technologically intensive industries, survival tends to entail occupying profitable niches rather than performing radical innovation. Several other industry life-cycle considerations are relevant, too. First, evidence from a range of industries shows that firms entering during the early or growth stages of a product or industry suffer less from the liability of newness than firms which enter in the mature phase.16 Second, a ‘liability of smallness’ affects entrants during the growth phase of an industry, but disappears in mature or technologically intensive industries. This is consistent with an outcome in which the number of profitable niches expands in mature markets.17 Third, growth has a greater impact on survival later in the industry life cycle. Fourth, the survival-enhancing properties of entrants’ technology strategies depend on the stage of the industry life cycle when entry occurs. Agarwal (1997) argues that a diversifying strategy enhances survival chances during the early stages of the product life cycle, as diversified firms have other options if they back a technology which does not become the ‘industry standard’ later on. But at later stages, a focused technology strategy pays off for agile new entrants which can exploit the latest cuttingedge technology. Bayus and Agarwal (2007) find some support for this hypothesis from the US personal computer industry. The appropriateness of specific firm-level strategies is likely to depend on the industry sector. Evidence from nearly 2,000 north-east US ventures suggests that a ‘product quality’ strategy enhances survival of manufacturing ventures; a ‘technology’ strategy works best in retail; and a ‘niche purveyor’strategy works best in services (Stearns et al., 1995). In general, though, a price-competition strategy does not enhance survival in the long run for most new ventures. The last rows of Table 14.1 show that industries and economic conditions which entail more intense competition, lower demand or higher unemployment, and higher interest rates tend to be associated with lower survival rates. None of these findings should be very surprising. At a more aggregated level, it is of interest to ask whether entry and exit rates are related dynamically. I conclude by taking a brief look at this question. Table 14.2 sets out several possible linkages between the aggregate birth rate of firms in a region at time t, denoted FBt , and the corresponding death rate, FDt . Depending on the nature of the firms that enter the market, and the market itself, births at t could either promote more or fewer births and deaths at a future time t + t, where t > 0. The same ambiguities also apply to the effects of firms dying at t on births and deaths at t + t.

396

Performance

Table 14.2. Firm birth and death interactions Expected sign of each effect Multiplier

Competition

Marshall

∂FBt+t ∂FBt

+

−

n.a.

∂FDt+t ∂FDt

+

−

n.a.

∂FBt+t ∂FDt

−

+

n.a.

∂FDt+t ∂FBt

−

+

+

Notes: Derivatives predict the effect of a marginal change in firm births (FB) or firm deaths (FD) on future births and deaths. n.a. = not applicable. Source: Johnson and Parker (1994, 1996).

For example, births could promote future births (and retard future deaths), in the case of a new innovative product that spawns imitators and spin-offs, and so increases the survival prospects of existing firms. This is called a ‘multiplier’ effect in Table 14.2. Alternatively, births could decrease the opportunity for future births and cause the eventual demise of rivals because of increased competition. This is called a ‘competition’ effect. Likewise, deaths can promote future births by releasing valuable economic resources (Pe’er and Vertinsky, 2008). In addition, as Marshall pointed out in his graphic ‘trees of the forest’analogy of small firm longevity, ‘sooner or later age tells on them all’ (1930, p. 263). Since (barring takeover) death inevitably follows birth, this is termed the ‘Marshall’ effect. Dejardin (2004) argues that competition effects are likely to prevail within industries, whereas multiplier effects function between different industries. It is possible to estimate the structure of birth–death interdependence using panel data.18 Estimates from a variety of countries suggest that lagged births and deaths have long-lasting impacts on future birth and death rates.19 In Holtz-Eakin and Kao’s (2003) US analysis, for example, past regional birth and death rates not only had significant positive impacts on current birth and death rates, but also had positive and persistent effects on subsequent state-wide levels of productivity. These findings suggest that researchers should be wary of treating firm entry and exit issues separately when using aggregate data. 14.5

Conclusion

Most entrepreneurial ventures start small. The entrepreneurs who found these ventures learn rapidly about their abilities to succeed in business and the majority soon become

Survival

397

aware that they are unable to compete as successfully as they had anticipated. Many of these entrepreneurs leave the market between two and four years after entry. Generally, firms which exit are young and small. Most surviving firms settle down into maturity, but only a small proportion of them experience protracted rapid growth. As noted in chapter 11, only a small fraction of these firms are the gazelles which ultimately become the large enterprises of the future. Low new-venture survival rates are consistent with Geroski’s (1995, p. 23) assertion that ‘entry appears to be relatively easy, but survival is not’. Much of the policy debate about encouraging enterprise and small businesses focuses on encouraging entry (see chapters 16 and 17). But in view of high new-venture closure rates, this emphasis may be misdirected. It might be more productive instead to investigate whether governments could and should intervene to help improve venture survival rates, or whether resources could be allocated more efficiently if some start-ups were discouraged altogether. For example, if failing companies free up scarce resources which can be used more productively by new entrepreneurial ventures, subsidies, tax concessions or other policies designed to shore up failing firms could be counter-productive (Pe’er and Vertinsky, 2008). Currently, our knowledge about appropriate public policy interventions in this domain remains extremely limited. As Holtz-Eakin (2000) points out, the literature currently lacks a notion of what the optimal rate of business failure is, making the role of optimal government policy towards survival rates moot. The problem is compounded by a lack of clarity in much empirical research between closure and failure, which as noted at the start of the chapter are two very different concepts. And as a careful reading of the literature makes clear, academic researchers have achieved only partial success in identifying the factors that affect business survival. Promoting high-tech enterprise might increase overall survival rates, since high-tech firms have superior survival prospects than ‘conventional’ firms. This may suggest a possible role for tax breaks for R&D and subsidies for innovative start-ups; government funding of university innovation-based research; and investment in basic education and skills. Even if these policies do not succeed in generating higher levels of entrepreneurial survival and activity, they might still be justified in terms of their positive impact on the economy’s stock of human capital. In fact, many countries have policies designed to foster innovation and entrepreneurship. Some of these policies will be discussed in chapter 16.

Notes 1. Mengistae (2006) provides a recent application. 2. See Everett and Watson (1998), Honjo (2000a, 2000b), Headd (2003), van Praag (2003), Bates (2005) and Cueto and Mato (2006). 3. Evans and Leighton (1989a) and Knaup (2005) provide US evidence; Daly (1991) offers UK evidence. 4. See Cressy (1993), Mahmood (2000) and van Praag (2003).

398

Performance

5. See Altman (1983), Hudson (1989) and Evans and Leighton (1989b) for US evidence, and Hudson (1987) and Cressy (1999, 2006a) for UK evidence. 6. Klepper and Simons (2000a) also argue that a central geographical location in an industry is not the principal cause of longevity but simply reflects the tendency for older innovating firms with a survival advantage to be located at the industry centre. 7. See Jovanovic and Nyarko (1996), Carroll and Hannan (2000), Thornhill and Amit (2003) and Esteve-Pérez and Manez-Castillejo (2008). Esteve-Pérez and Manez-Castillejo (2008) provide supportive evidence of a liability of obsolescence using a survival model containing high-order terms in venture age. 8. References for all apart from a include: Brock and Evans (1986), Evans (1987a, 1987b), Reynolds (1987), Cooper et al. (1988, 1989, 1992, 1994), Dunne et al. (1988, 1989a), Evans and Leighton (1989b), Hudson (1989), Phillips and Kirchhoff (1989), Preisendörfer and Voss (1990), Bates (1990, 1995, 1997), Audretsch (1991), Reid (1991), Baum and Mezias (1992), Brüderl et al. (1992), Mahmood (1992), Barron et al. (1994), Dunne and Hughes (1994), Holtz-Eakin et al. (1994b), Mata and Portugal (1994, 2002), Storey (1994a), Wagner (1994), Audretsch (1995a, 1995b),Audretsch and Mahmood (1994, 1995), Geroski (1995), McPherson (1995, 1996), Stearns et al. (1995), Westhead (1995), Westhead and Cowling (1995), Cressy (1996, 1999), Hart and Oulton (1996), Holmes and Schmitz (1996), Mata et al. (1995), Mata (1996), Nafziger and Terrell (1996), Storey and Wynarcyzk (1996), Gimeno et al. (1997), Sutton (1997), Everett and Watson (1998), Pennings et al. (1998), Santarelli (1998, 2000), Carrasco (1999), Audretsch, Santarelli and Vivarelli (1999), Brüderl and Preisendörfer (1998), Fairlie (1999), Hinz and JungbauerGans (1999), Nucci (1999), Quadrini (1999), Reid (1999), Shane and Foo (1999), Taylor (1999, 2001, 2004), Bates and Servon (2000), Bednarzik (2000), Boden and Nucci (2000), Lin et al. (2000), Honjo (2000a, 2000b), Mahmood (2000), Pfeiffer and Reize (2000), Klepper (2001), Kangasharju and Pekkala (2002), Segarra and Callejon (2002), Shane and Stuart (2002), Astebro and Bernhardt (2003), Doms et al. (1995), Headd (2003), Lotti et al. (2003), van Praag (2003), Arum and Müller (eds.) (2004, all chapters), Bosma et al. (2004), Giannetti and Simonov (2004), Meager and Bates (2004), Georgellis et al. (2005), Cueto and Mato (2006), Taymaz (2005), Cefis and Marsili (2006), Mengistae (2006), Sarkar et al. (2006), Tamásy (2006), Batjargal (2007), Bayus and Agarwal (2007), Fairlie and Robb (2007b), Kim and Miner (2007), Strotmann (2007), Esteve-Pérez and Manez-Castillejo (2008), Lin and Huang (2008) and Michael and Combs (2008). 9. See Carroll et al. (1996), Klepper and Simons (2000b) and Bayus and Agarwal (2007). 10. See in particular Bates (1990, 1997), Cooper et al. (1994), Westhead (1995), Gimeno et al. (1997), Reynolds and White (1997), Brüderl and Preisendörfer (1998) and Taylor (1999). 11. See Romanelli (1989) and Gimeno et al. (1997). Srinivasan et al. (2008) report that the effects of product market diversity depend on the number of patents and trademarks owned by ventures. 12. See Reid (1991), Brüderl et al. (1992) and Strotmann (2007). 13. Bates (1998), Lafontaine and Shaw (1998) and Stanworth et al. (1998). See also Michael and Combs (2008), who observe that features of franchise contracts such as royalty rates also bear on franchisee survival rates. 14. Brüderl et al. (1992), Cooper et al. (1988), Shane (2003, pp. 118–20), Knaup (2005) and Cueto and Mato (2006). Long spells in continuous entrepreneurship are found disproportionately in agriculture, business services and repair businesses (van Praag, 2003). 15. See Agarwal (1996), Agarwal and Gort (1996) and Lin and Huang (2008). 16. Agarwal (1997), Agarwal et al. (2002), Bayus and Agarwal (2007). 17. Agarwal (1998), Agarwal and Audretsch (2001), Agarwal et al. (2002), Sarkar et al. (2006).

Survival

399

18. Johnson and Parker (1994, 1996) proposed the following reduced-form panel-data vector autoregression model for this purpose: FBit =

m 

α1t FBi,t−t +

t=1

+

m 

β1t FDi,t−t

t=1

m 

γ1t Xi,t−t + uit

(14.1)

t=1

FDit =

m 

α2t FBi,t−t +

t=1

+

m 

m 

β2t FDi,t−t

t=1

γ2t Xi,t−t + vit ,

(14.2)

t=1

where i denotes a region, t is a time period, m is the number of lags through which birth–death interdependence manifests itself, the αs, βs and γ s are parameters to be estimated, uit and vit are disturbances and X is a vector of macro-economic variables that potentially affect births and deaths (see e.g. Highfield and Smiley, 1987; Yamawaki, 1990). Ideally, data on FB and FD should exclude sales of firms, which are sometimes recorded as both entries and exits, even though only relabelling has occurred. Another problem when estimating (14.1) and (14.2) is collinearity between explanatory variables, including past entries and exits (Carree, 2006, p. 178). Johnson and Parker estimated (14.1) and (14.2) using OLS, though IV is more efficient (Kangasharju and Moisio, 1998). 19. See Johnson and Parker (1994, 1996), Kangasharju and Moisio (1998), Holtz-Eakin and Kao (2003), Dejardin (2004) and Bartelsman et al. (2005).

Part IV Public Policy

15 Principles of entrepreneurship policy

Government interventions in market economies can impact heavily on both the welfare of entrepreneurs and the size of the entrepreneurial sector. Government intervention is often predicated on a belief that entrepreneurs generate valuable positive social and economic externalities, emanating from new ideas, new products, new employment and enhanced competitiveness. If entrepreneurs are unable to appropriate all of these benefits themselves (i.e. unable to ‘internalise’ the benefits), there might be too little investment in entrepreneurship for the social good, justifying government intervention. Governments often also see entrepreneurship as a route out of poverty and dependence on state benefits, and as a means of achieving self-reliance. At the same time, though, governments often evince concern about the living and working conditions of the poorest and most vulnerable workers. These workers are often found in entrepreneurship, either as struggling small business owners or own-account self-employees. At the outset it is helpful to explain what is meant by the term ‘public policy towards entrepreneurship’. Lundström and Stevenson (2005, 2007) distinguish between entrepreneurship policies, which focus on start-ups of new businesses, and SME policies, which focus on the stock of existing small businesses.1 More precisely, Lundström and Stevenson define entrepreneurship policy as ‘policy aimed at the prestart, the start-up and early post-start-up phases of the entrepreneurial process, designed and delivered to address the areas of motivation, opportunity and skills, with the primary objective of encouraging more people in the population to consider entrepreneurship as an option, move into the nascent stage of taking actions to start a business and proceed into the entry and early stages of the business’ (Lundström and Stevenson, 2007, p. 105). It is sometimes argued, for example, that government should dispense with ‘laissez faire’ and actively encourage market entry in fields like telecommunications and e-commerce (e.g. Mayer-Schönberger, 2003). In contrast, ‘the primary aim of small business policy is to level the playing field for small firms through measures to overcome their disadvantages in the marketplace resulting from their “smallness” and “resource poverty”, and to improve their competitiveness’ (Lundström and Stevenson, 2007, p. 105). Storey (2003) provides a commanding overview of public policies towards SMEs. He identifies several 403

404

Public policy

underlying aims of SME policies, which include helping young people and promoting job creation.2 It is possible to take different views about the stance of specifically US public policies towards entrepreneurship and SMEs. On the one hand, Dennis (1999) contends that ‘the United States has no small business or entrepreneurship policy’, being more concerned with fostering competition and defending the interests of the consumer. On the other hand, the USA undoubtedly has the longest-running suite of policies which bear on entrepreneurship and small business in the developed world. These cover access to finance (e.g. the SBA’s 7(a) Loan Program); SME access to federal procurement (e.g. the Subcontracting Assistance Program); and counselling, training and education of SME owners and entrepreneurs (e.g. the Small Business Development Centres). This list does not include advocacy programmes (e.g. Regulatory Enforcement Assistance) or innovation assistance programmes (e.g. the Small Business Innovation Research programme). In a historical perspective, Audretsch (2001) argues that US public policy towards business in the post-war period has shifted from constraining the power of large firms towards encouraging the creation and commercialisation of knowledge, especially by small businesses. Overall, entrepreneurship policies tend to focus more at the individual level than SME policies, which usually operate at the firm level. But because many entrepreneurs run SMEs, the distinction between the two types of policy is blurred. Both of them will be discussed below, subsumed (for expositional convenience only) under the joint heading of ‘entrepreneurship policies’. I will not attempt to describe how public policies can be co-ordinated across diverse areas and shaped into a coherent whole in an attempt to forge an ‘entrepreneurial society’ (Audretsch, 2007a). The reason is that, when evaluating policies of any kind, economists usually assume that the end goal is economic efficiency or social welfare. That is, policies are only useful to the extent that they promote one or other of these objectives. It cannot simply be assumed that an ‘entrepreneurial society’ is intrinsically desirable (i.e. that it always generates the highest social welfare). A practical difficulty is that it is rarely clear how long one must wait before improvements in efficiency or social welfare are felt. This is a practical question which policy evaluations must deal with, as explained below. This chapter is the first of three which analyse public policy towards entrepreneurship. The present chapter deals with some basic principles of intervention which underpin the two chapters that follow it. The first section outlines cases for and against public policies towards entrepreneurship. At first sight, it might seem strange that one might want to make a case against entrepreneurship policies. But in the absence of market failures, public policies could be unnecessary and even counter-productive. Furthermore, as we saw earlier in this book, the free market can sometimes generate over-investment in entrepreneurial projects. The second and third sections of the chapter deal with two important principles of policy-making in relation to entrepreneurship, namely their design and evaluation. In practice, all entrepreneurship policies have to be targeted on specific groups, otherwise finite public resources would be spread too thinly to be effective. The challenge is to

Principles of entrepreneurship policy

405

target policies in a way which does not just throw money at people who would have taken the desired actions anyway (Lerner, 1999). But the effective use of resources can only be gauged by performing robust policy evaluations. Some basic principles of programme design and policy evaluation are discussed at the end. The two chapters which follow will build on these principles and cite evidence relating to specific entrepreneurship policies which have been implemented in various countries around the world. 15.1 The case for and against pro-entrepreneurship public policies

Storey (2003) identifies two commonly articulated rationales for pro-entrepreneurship policies. The first draws on findings that entrepreneurs create jobs and innovations (see chapter 10), implying that more entrepreneurship is better than less. The second rationale is based on the notion of ‘market failures’which hinder the operation of the free market and can lead to insufficient entrepreneurship. The conventional wisdom among economists is that, in the absence of market failures, free and competitive markets populated with rational agents will generate efficient amounts of entrepreneurship, innovation, employment and output. Hence economists usually ask where the market failure is first, and only propose corrective public policies once convincing market failures have been identified. A point that is sometimes overlooked is that market economies can produce too much rather than too little entrepreneurship. Accordingly this section is divided into two parts, the first making the case for pro-entrepreneurship policies, and the second making the contrary case. No definitive conclusion will (or arguably can) can be reached at the end of this discussion; but the exposition hopefully clarifies the issues at stake, many of which are implicit rather than explicit in most real-world policy discussions. 15.1.1 The case for pro-entrepreneurship policies There are several market failures which might justify pro-entrepreneurship policies. One is information imperfections. At its simplest, such imperfections can include basic ignorance among the populace about the private benefits of starting a business, which can cause insufficient numbers of people to enter entrepreneurship. It can also entail ignorance among established entrepreneurs about ways they can improve their businesses. Asymmetric information is another manifestation of imperfect information. Chapters 7 and 8 analysed this issue and discussed the under-investment problems that can ensue in the context of access to entrepreneurial finance. New firms can also be at an informational disadvantage relative to incumbents with respect to labour markets. Hiring workers from scratch, new firms are especially susceptible to asymmetric information about hidden worker types, making it costly for them to hire workers. This consideration might justify subsidising hiring by new firms (Boadway and Tremblay, 2005). A second and related rationale for pro-entrepreneurship policies arises when entrepreneurial actions signal the existence of valuable opportunities to others, who

406

Public policy

can enter and compete away the returns. The danger is that too few entrepreneurs have incentives to enter a new industry when as first entrants they inadvertently signal to other, stronger, firms the profit opportunities that are available. The ‘first mover’ bears all of the costs of discovery; but absent perfect IPR (e.g. patent) protection has to share the benefits with free-riders who can copy innovations without sharing the costs (Alexander-Cook et al., 1998). Anticipating lower returns owing to free-riding behaviour by rivals, entrepreneurs might no longer find it worthwhile to develop new innovations. The outcome is under-investment in innovation (Klette et al., 2000). A third rationale for pro-entrepreneurship policies is that entrepreneurs can generate positive externalities. For example, entrepreneurs who introduce process innovations reduce production costs and therefore product prices, but do not take into account the resulting increase in consumer surplus. As competition diffuses innovations within and across industries, consumers can make enormous gains, while the inventors and entrepreneurs responsible for creating and bringing new innovations to market might reap only meagre financial returns by comparison. Estimates made for the US non-farm business sector indicate that ‘innovators are able to capture [only] about 2.2 per cent of the total surplus from innovation. . . the rate of profit on the replacement cost of capital over the 1948–2001 period is estimated to be 0.19 per cent per annum’(Nordhaus, 2004, p. 34). The remainder of the surplus is captured by consumers. It is striking that while some entrepreneurs can become fabulously rich from introducing radical innovations, many do not, while society gains in all cases. This suggests the existence of a positive externality not corrected by the market, and hence under-provision of innovative activity in general and entrepreneurship in particular. Amazon, which started in 1995, affords a good example of this phenomenon. Despite the substantial gains that millions of consumers have made from Amazon’s success in reducing the prices of books and other products, together with increased convenience and enhanced levels of service, Amazon became profitable only in 2002, and even by full-year 2005 its net income was only $78 million. Another source of positive entrepreneurial externalities can be business clusters, which disseminate knowledge spillovers at relatively low cost. Left to themselves, no single entrepreneur would find it worthwhile to kick-start a cluster because they cannot capture all of the benefits, while they would incur the set-up costs. This might justify a role for public policies aimed at developing attractive places for entrepreneurs to cluster in. Specific examples include science, technology and research parks; dedicated city spaces to attract talented and creative people; and specialists who can help bridge the invention-innovation gap, such as VCs and technology-oriented law firms.3 Successful examples of clusters in the USA which are located close to universities include the Research Triangle Park in North Carolina; the Thomas Edison centres in Ohio; the Advanced Research Program in Texas; and, of course, Silicon Valley in California. Knowledge itself can be a prime source of externalities, which drives a wedge between private and social returns. Sequential and cumulative innovations spill over to future researchers, benefits that the original innovators cannot anticipate or capture (Aghion and Howitt, 1992). Spillovers can be sizeable in magnitude too, amounting

Principles of entrepreneurship policy

407

to 50–100 per cent of private returns (Griliches, 1992). Nor is it uncommon for ideas created by failed firms and projects to become integral parts of successful projects in other firms (McGrath, 1999; Holbrook et al., 2000). These externalities are not internalised by entrepreneurs or investors of failed companies. This provides another reason why market economies might generate too little entrepreneurship. More generally, one might ask whether market economies attain the correct balance between investment in innovation-generating research and entrepreneurs who bring innovations to market. Michelacchi (2003) studied this issue using a model of free occupational choice where rents to research and entrepreneurship are exogenously fixed by Nash bargaining. Michelacchi showed that, if rents to entrepreneurship are too low, an economy can end up wasting research, with insufficient entrepreneurial skills to exploit the inventions produced by researchers. This concern is often articulated by European policy-makers, who point to a strong record in successful invention in Europe but a mediocre record in commercial innovation. Once again, this might justify government intervention to promote entrepreneurship.4 Entrepreneurial firms might also be more efficient organisations than large incumbents, especially if the corporate market for control is imperfect. Corporations with separated ownership and control are vulnerable to agency costs, as well as social costs incurred by self-aggrandising managers who pursue non-value-adding mergers with other firms. Arguably, entrepreneurial firms in which ownership and control are combined in the entrepreneur are not as prone to these problems, suggesting that anti-trust concerns might be mitigated when entrepreneurial firms are the dominant organisational form (Straume, 2006). This buttresses long-standing arguments that entrepreneurship stimulates competition, and not only through the conduit of market entry. As usual, if these benefits are not internalised by the entrepreneurs themselves, the free market might generate too few entrepreneurial firms.

15.1.2

The case against pro-entrepreneurship policies Converse arguments can be proposed, claiming that public policy should either forbear from intervening in markets to promote entrepreneurship, or should even actively discourage entrepreneurship. Some of these arguments simply reverse the logic of their counterparts above. For instance, entrepreneurs might be ignorant of the full costs of starting a business, including the stress on personal relationships it entails (q.v. chapter 4). In which case, social welfare could be increased by discouraging entry. And asymmetric information in credit markets can lead to over-investment, implying an efficiency gain from discouraging entrepreneurship (q.v. chapter 7). There might also be too much innovation by entrants. The ‘business stealing effect’ is the name given to the phenomenon whereby entrants do not internalise the destruction of rents caused by displacing established firms, and so innovate too much for the social good (Aghion and Howitt, 1992). The problem of excessive entry is especially pronounced when entrepreneurs engage in a contest to be the first to make a new discovery (Futia, 1980), or when product markets are imperfectly competitive. In the

408

Public policy

latter case, innovation by entrants generates too much product diversity and too little informative advertising (Grossman and Shapiro, 1984). In which case, pro-innovation policies for new entrants could be misplaced and the appropriate policy response is instead to reduce their incentives to innovate. In the words of Boadway and Tremblay (2005, p. 188): ‘On balance, it is hard to make a case for policy incentives to new firms based on the externalities of their activities. At best, a general incentive for innovation by all firms might be appropriate.’ Entrepreneurship can also have adverse effects on workers. Earlier chapters have observed that start-ups destroy as well as create most new jobs, and business failures can harm customers and suppliers. These adverse impacts are often ignored in discussions of entrepreneurship policy. Furthermore, there is evidence that entrepreneurs create more ‘low quality’ jobs than large incumbents do (see chapter 10). They could even compete workplace standards downwards, driving out good labour practices. Finally, it is important to remember that not every problem is worth fixing. Enterprise policies can have the unintended side effect of diverting entrepreneurial effort into bidding for subsidies (Sobel, 2008) and resources into job creation for bureaucrats. More generally, intervention carries the risk of ‘government failure’, whereby civil servants divert resources from the intended recipients of the policy to satisfy their own private interests. While it is true that government is not the only source of ‘rent seeking’ of this kind, it can be an important one, especially in developing and transition economies, as discussed in chapter 12. 15.2

Principles of entrepreneurship policy design

Three basic principles underpin successful public policy design. One is ‘additionality’. This requires that a policy increases the net number of people performing the favoured activity. The converse of low additionality is high ‘deadweight’ cost. A related issue is ‘displacement’, whereby a treated group is advantaged relative to incumbents, leading to the replacement of the latter in the market by the former. For example, a policy which subsidises entry by new firms in a highly competitive market is likely to displace unsubsidised incumbents. The result is little or no net impact from the policy on efficiency, output or employment. Displacement is similar to the ‘business stealing effect’ described earlier, and generally reduces programme impacts. A second important principle of entrepreneurship policy design is that programmes successfully reach the individuals they are intended to reach. That is, policies should be well targeted. A third principle is that programmes should neither be too complicated for their target participants to understand nor too costly to administer. It is worth pointing out that these principles can be difficult to reconcile in practice. Measuring accurately what would have happened in the absence of a programme is problematic unless the researcher constructs a genuine ‘natural experiment’ – which is rarely possible. Instead, researchers usually have to estimate what would have happened in the absence of a programme, on the basis of several assumptions about preferences and behaviour. These estimates might not be accurate if the assumptions are

Principles of entrepreneurship policy

409

incorrect. It is common for evaluators to ask programme participants whether they think they would have engaged in the activity if the programme had not existed. An obvious problem with this strategy is that respondents might not give accurate responses. Bates’ (2003) overview of US support programmes for minority business enterprises illustrates some of these problems. Bates points out that these programmes have targeted disadvantaged entrepreneurs disproportionately engaged in overcrowded and low-profitability lines of business, who lack the skills, experience and education to operate viable businesses. Although this seems like an eminently deserving target group, these programmes have stimulated low levels of actual business development and are associated with high loan default rates (Bates and Bradford, 1979). On a more positive note, Bates (2003) cites as an example of good practice Maryland’s Small Business Development FinancingAuthority, which lends to minority enterprises that successfully competed for public procurement contracts but which were denied loans by commercial banks on the grounds of limited experience or track records. This programme has enjoyed superior repayment rates which have allowed further disbursements of funds. Bates argues that public support should be targeted only to viable businesses that offer reasonable prospects of economic development, investment and repayment; and he points out these programmes will probably have to move in this direction anyway if they are to remain solvent. Likewise, Santarelli and Vivarelli (2007) recommend that subsidies should be targeted towards post-entry enterprises to avoid wasting resources on ‘hopeless’ entrants. The central challenge for policy-makers is to identify those ‘infra-marginal’ firms which are strong enough to survive, but not strong enough to grow without external assistance. This is unlikely to be an easy task, to say the least. The inescapable dilemma is that the people whom it is easiest and most profitable for entrepreneurship programmes to help are precisely those who are least in need of help. This of course comes back to the additionality issue. Parker (2007a) identifies two general problems with targeted entrepreneurship policies. One problem is caused by targeting entrepreneurs on the basis of past average rather than future marginal performance. The other comes from ignoring adverse reactions of entrepreneurs and other private sector agents to new policy initiatives. Regarding the first point, efficient resource allocation mandates directing scarce funding to places where it generates the highest marginal return. It should not be assumed that entrepreneurs who have generated high average returns in the past will also generate high marginal returns in the future, especially if diminishing marginal returns to extra funds set in. Regarding the second point, Parker (2007a) cites several examples where private sector agents frustrate government efforts to induce desired outcomes. Some of these examples will be discussed in the chapters which follow. 15.3

Entrepreneurship policy evaluation

It is essential to monitor and evaluate publicly funded entrepreneurship policies to ensure that taxpayers receive value for money. Unfortunately, evaluations of

410

Public policy

public entrepreneurship policies are not widespread in practice. One reason is that entrepreneurship programmes often have unclear aims, or multiple objectives. It is therefore difficult to know what benchmarks to use for evaluation purposes. Storey (2000) has proposed a taxonomy comprising six different ways of evaluating entrepreneurship programmes in increasing order of sophistication. Steps I to III simply monitor programme take-up rates and the average venture performance of programme participants. These are relatively unsophisticated evaluation methods. Step IV compares the performance of assisted ventures with ‘typical’ non-assisted firms. Step V goes further and compares assisted ventures with a matched sample of comparable firms. Step VI evaluations are like Step V but also take account of self-selection of firms into the programme, including possibly ‘committee selection’of participants from the set of loan applicants.5 Storey (2003) emphasises the importance of controlling for self-selection, which if neglected can yield biased estimates of programme impacts. Storey (2006) points out that Step VI evaluations can be expected to find the smallest programme impacts because they hold the most factors constant which cannot be wrongly attributed to the programme itself. Storey (2006) observes that some programmes are never evaluated, whereas others are evaluated repeatedly. For example, £2.5 billion of the £7.9 billion spent on small businesses and entrepreneurs in the UK is taken up by favourable tax treatment of small firms.6 However, the effectiveness of this spending has never been evaluated, while the much less costly Small Firms Loan Guarantee Scheme (see chapter 16) has been evaluated no fewer than five times since it was first piloted in 1981. Storey adds the criticism that governments rarely evaluate the combined impact of regulations. On a related point, Dennis (1998) notes that we lack hard data about the regulatory impacts of all units of government on rates of new venture creation and the growth rates of new businesses. Meager et al. (2003) make four points about government pro-entrepreneurship policies. First, they typically suffer from low impact and high deadweight costs which can reach 60–70 per cent and more. Second, they are often characterised by high displacement rates. Third, one-off upfront grants encourage more sustainable start-ups than ongoing regular allowances. Fourth, careful initial screening of recipients generally improves survival rates of supported start-ups. Finally, it is important to remember that the outcomes of most evaluations are sensitive to the length of time which elapses before policies play out. Hart (2003) distinguishes between policies generating ‘intermediate’ outcomes over a period of several years, and those affecting ‘background’ conditions which take a decade or more to emerge. This point is relevant when trying to evaluate the effects of policies on business survival rates, which can take several years to emerge. Mainly reflecting data limitations, most of the specific policies and evaluations discussed in the next two chapters measure shorter-term impacts which can be reasonably assessed over the span of a few years.

Principles of entrepreneurship policy

411

Notes 1. To implement SME policies, it is necessary to define an SME. In the USA, it is a firm with fewer than 500 employees; in Europe, it is a firm with fewer than 250 employees. Since most ventures in any country are SMEs, further decompositions are helpful, including ‘micro’ (0–9), ‘small’ (10–49), and ‘medium’ (50–250) in the EU. There are other criteria for measuring size too, of course, including turnover and balance sheet measures. 2. For a detailed review of policies designed to promote small business, see the OECD Best Practice Policies for SMEs publications. Storey (2003) surveys the policy landscape across the world while Lundström and Stevenson (2005) document policies relating to advice and support for entrepreneurs in detail across thirteen countries. 3. See Audretsch (2003), Florida (2003) and Auerswald and Branscomb (2003). 4. On the other hand, entrepreneurial over-investment is also possible in this model. And it is unclear whether Michelacchi’s (2003) model is robust to endogenous rents which reflect the productivity of effort in each sector. 5. Almus (2004) provides an example of a Step VI evaluation, based on a non-parametric matching approach. This avoids problems of mistakenly attributing programme success to factors that actually affect programme participation. To illustrate, let Y 1 and Y 0 be some outcomes for firms which received assistance and which did not receive any, respectively. Index a firm which participates in the programme by j = 1, else set j = 0. The true effect of assistance (as opposed to participation) is given by E(Y 1 | j = 1) − E(Y 0 | j = 1) . But E(Y 0 | j = 1) is unobserved because by definition everyone who participates receives assistance. Nor can one use instead the average performance of non-participants E(Y 0 | j = 0) because participation is non-random, so E(Y 0 | j = 0)  = E(Y 0 | j = 1). However, now suppose that the propensity to participate in a programme is the same for firms with the same characteristics X = x. Then one can measure the impact of assistance as E(Y 1 | j = 1, X = x) − E(Y 0 | j = 0, X = x). All the researcher has to do is to identify pairs of subsidised and non-subsidised firms with similar characteristics. 6. These figures come from a 2002 UK DTI report. £7.9 billion per annum amounted to 0.8 per cent of GDP at the time. Storey points out that this is more than the annual cost of running the UK police force.

16 Finance and innovation policies

Governments try to promote entrepreneurship in numerous ways. These include: improving access to start-up finance; providing grants and subsidies for innovation; structuring the tax system to help entrepreneurs; and offering entrepreneurs subsidised assistance and advice. In addition, regulation can affect the performance of entrepreneurial ventures and hence the attractiveness of entrepreneurship. This chapter and the next one outline theory and evidence relating to government intervention in these areas. The present chapter focuses on public policies designed to enhance access to finance for entrepreneurs and to promote innovation. Several programmes belong to this policy set. The first section introduces one of the best-known and longest-established finance policies, loan guarantee schemes (LGSs). I present some stylised facts, theoretical insights and empirical evaluations of LGSs. The second section considers other types of credit policies including direct government interest subsidies, loans and grants. The third section deals with policies designed to increase the provision of equity finance, especially venture capital. Several policies come under this heading and overlap somewhat with policies towards innovation. This is only natural, as so much venture capital is directed to innovative new companies (see chapter 8). Other aspects of innovation policy are briefly discussed in the final section. 16.1

Loan guarantee schemes

Chapter 7 reviewed the theoretical case for credit rationing and under-investment in entrepreneurial ventures. This issue has received substantial attention in policy circles and has motivated various kinds of government intervention, including loan guarantee schemes (LGSs). In this section, the main features of LGSs are first described, before theoretical issues are considered. Then I review evidence about their impact in the UK and the USA. 16.1.1 Organisation of LGSs A LGS provides a government-backed guarantee to encourage banks and other financial institutions to lend to small firms which are unable to raise conventional finance 412

Finance and innovation policies

413

Table 16.1. Features of LGSs in the UK, the USA, France, Germany and Canada Terms

UK

USA

France

Germany

Canada

Maximum guarantee Arrangement fee

70–85%

75–80% a

50–65% b

80%

85%

1% of loan

None

1.5% of loan

Loan term Number of guarantees Net cost of scheme

2–10 years 5,957 (2005/06) £46m (1997/98)

0.75% of guarantee 0.8–1% of credit 15–23 years 6,850 (1996) n/a

2% of loan

Loan premium

2–3.875% of guarantee 0.5% of credit 7–25 years 45,300 (1997) £95m (1998)

0.6% of credit 2–15 years 5,000 (1996) £36m (1996)

1.25% of credit 0–10 years 30,765 (1997) £42.5m (1997)

Notes: a 70 per cent for start-ups; 85 per cent for firms with at least two years’ trading history. b 65 per cent for start-ups. n/a. = Not available. Source: KPMG (1999, table 4.1), updated.

because of a lack of collateral or an established track record. Some schemes, such as the Canadian one, do not specify lack of collateral as a criterion, and can be regarded as more of a publicly funded credit scheme. LGSs are widespread, with versions operating in the USA, the UK, France and Germany as well as Canada and several other countries. These schemes are predicated on a belief, recently articulated by the European Commission, that ‘access to finance remains a major barrier to entrepreneurs’ (EC, 2003, p. 11). Although the details vary between countries – and within given countries depending on the type of loan1 – the basic principles of an LGS are the same. An LGS typically guarantees finance for both new start-ups and going concerns. Only borrowers unable to secure a conventional loan from a bank, and in non-proscribed industrial sectors,2 may apply for a loan guarantee. A bank nominates ventures for approval by the government and administers the loans. Conditional on approval by the government, the bank takes the usual repayment if a venture succeeds. If it fails, the bank is liable only for a fraction of the loss, with the rest borne by the government. In the USA, the Small Business Administration (SBA) underwrites between 75 and 80 per cent of loans. In the UK, the British government underwrites between 70 and 85 per cent. Borrowers are charged an arrangement fee, and pay the usual market rate plus a small premium. Operational costs can be substantial. Table 16.1 furnishes details about these and other aspects of LGSs operating in several countries. 16.1.2

Theoretical perspectives on LGSs If financial markets are efficient, one would expect all loans involving reasonable degrees of risk and promising reasonable rates of return on capital to be made by private sector lenders. Any rejected loan applications would reflect unacceptable levels of risk

414

Public policy

for a given rate of return; and any government-backed lending to facilitate such ventures would therefore be expected to result in high default rates. However, if the financial market for small firm lending is inefficient, the case for government intervention in credit markets may be stronger. The basic question is whether governments without privileged access to inside information about entrepreneurs’ ventures can use an LGS to improve on the market’s allocation of credit under asymmetric information. There are several possible sources of inefficiency in entrepreneurial credit markets, including asymmetric information (chapter 7) and information externalities. The latter can arise when banks have to acquire costly information about the risk profile of their borrowers, which is then priced into their interest rate and thereby becomes public information which other banks can exploit when competing for these customers. This can give rise to the classic ‘public good’ free-rider problem (Boadway and Sato, 1999). Collectively, banks under-invest in acquiring information about borrowers. In principle, governments can improve on this situation by funding an LGS to supply credit which is not forthcoming from the private sector. Other researchers have focused on dangers which can arise from using LGSs. Several of these findings illustrate the ‘Law of Unintended Effects’. For example, using a simple model with two types of entrepreneurial borrower, one high-risk ‘bad’ type (b), and one low-risk ‘good’ type (g), Gale (1990a) and Smith and Stutzer (1989) propose a separating equilibrium in which the g borrower types accept the risk of credit rationing in return for a lower interest rate, while the b types face no rationing and a higher interest rate. But to preserve separation, banks respond to an LGS which increases the probability that a rationed g borrower obtains a loan by rationing credit to g types even further. Thus the government’s policy is frustrated by the reactions of private sector agents, and overall welfare decreases. However, there is little evidence that banks use credit rationing to separate hidden types, so this model is probably only of theoretical interest. And Gale (1990a) demonstrates that an LGS can improve efficiency when only low-ability entrepreneur types are targeted. Gale (1990b) points to two other dangers of government intervention in credit markets. One is that if the government provides credit on commercial terms it is likely to simply crowd out private sector bank credit, yielding no net benefit to the economy. Hence a subsidy element is needed to improve on the competitive equilibrium, which might explain why government credit programmes tend to lose money in practice – as shown in the final row of Table 16.1. The second problem is that with multiple target groups subject to rationing or redlining, a subsidy to one group may increase banks’ expected returns by so much that other, initially non-rationed, groups may become rationed or redlined. This can be seen in Figure 7.3, where an LGS targeted on group b increases b’s expected rate of return schedule above ρ ∗ in that diagram if the scheme is sufficiently generous. But ρ ∗ also increases, reflecting the greater overall profitability of private loans – possibly leading to redlining of the non-targeted group o, who were not initially redlined. These perverse crowding-out effects may serve to generate pressure for further subsidies. In Gale’s words: ‘credit subsidies create demand for more credit subsidies’ (1990b, p. 190).

Finance and innovation policies

415

Innes (1991) analysed an asymmetric information model in which under- or over-investment (though not credit rationing) can occur. Heterogeneous entrepreneurs can choose loan sizes, and there is scope for high-ability entrepreneurs to signal their types by asking for loans that are larger than the socially efficient level. Innes showed that an LGS is an inefficient and ineffective means of intervention in this scenario. The reason is instructive. The government’s absorption of some of the default costs gives entrepreneurs an incentive to make inefficient investment choices. Large maximum guarantee levels simply exacerbate over-investment by entrepreneurs, with the complicity of banks whose downside risks are covered by the government.

16.1.3

Evaluation of LGSs In practice, as Table 16.1 shows, the charge to the government of failed guaranteed ventures can be high, so in line with one of the principles of entrepreneurship programme design mentioned in the previous chapter, it is desirable to evaluate the effectiveness of these schemes. This is all the more important when an LGS plays a prominent role in the entrepreneurial credit market. In the USA, the SBA’s portfolio of nearly one-quarter of a million loans was worth about $60 billion by 2004, comprising some 10 per cent of commercial bank small firm financing (Craig et al., 2007). This makes the SBA the single largest backer of small business in America. The success of an LGS is limited by several factors. As noted in the previous chapter, one possible problem is that some borrowers would have received a loan in the absence of the LGS. That is, the ventures financed under the scheme are not truly ‘additional’. Estimating ‘additionality’ of LGS-backed loans is actually quite difficult. Despite obvious drawbacks with this method, both KPMG (1999) and Riding and Haines (2001) estimate additionality for the UK and Canadian schemes respectively by asking borrowers whether they would have started without the LGS-backed loan. KPMG (1999) estimated that 70 per cent of UK Small Firms LGS (SFLGS)-supported firms were ‘finance additional’ (i.e. would not have been financed without the involvement of the LGS), leveraging £161 million of additional private finance between 1993 and 1998. Using self-reported estimates of job creation, Riding and Haines (2001) estimated the cost of the Canadian Small Business Financing (CSBF) LGS programme to be only C$2,000 per job. However, these estimates of additionality might be upward-biased if borrowers are inclined to over-attribute favourable outcomes to the scheme. More recently, Riding et al. (2007) (who also survey the literature about the measurement of LGS additionality) estimated the additionality of the CSBF LGS programme to be 74.8 ± 9.0 per cent. However, this estimate is also based on questionable assumptions.3 The only truly satisfactory way of measuring additionality is to identify those loans which banks would genuinely only back with the guarantee but would not back without it. Although this might be possible in principle, it would require researchers to work closely with, and obtain accurate data from, bank loan officers handling actual loan applications. To the best of my knowledge, this kind of research design has not yet been conducted.

416

Public policy

A second potential limitation of an LGS is that it might encourage self-selection of the riskiest ventures, while reducing banks’ incentives to evaluate thoroughly the viability of investment proposals (Rhyne, 1988; Innes, 1991). The severity of this problem can be expected to increase in line with the loan guarantee fraction. Early evidence pointed to high failure rates of LGS-backed businesses (Bosworth et al., 1987; Rhyne, 1988). For example, Rhyne (1988) calculated that ‘defaults require a 9 per cent subsidy from SBA, or 11 per cent if SBA’s administrative expenses are included’ (1988, p. 96). Rhyne (1988) claimed that ‘SBA loans are ten times more likely to default than private ones’ (p. 63). But as Mandel (1992) pointed out, Rhyne focused on short-maturity SBA loans, which are the most risky, and overestimated losses from defaults. Mandel estimated that SBA loans were only three times more likely to result in loss than private ones. Subsequent to this research being conducted, the UK and US LGSs have been made more stringent with lower guarantee fractions. More recent UK evidence suggests that LGS-funded businesses no longer have higher failure rates on average than non-LGS-funded businesses (KPMG, 1999). Third, although LGSs promise supply-side benefits such as greater competitiveness and help for fledgling enterprises to survive their early growing pains, there is a risk that successful start-ups financed by an LGS could displace viable existing enterprises. This was thought to be a particular problem in retailing, catering and motor vehicle maintenance in the UK, leading to the UK SFLGS being restricted in 1989 and 1996 for precisely this reason. Nevertheless, KPMG (1999) still estimated high displacement rates from the UK SFLGS of between 76 and 86 per cent. This limited the employment creation contribution of guaranteed ventures to between 0.3 and 0.6 jobs per firm in the eighteen months following the loan, amounting to under 10,000 net additional jobs in total. The net additional cost per job charged to the government was estimated at between £9,500 and £16,600. Indirect evidence points to a more favourable regional effect from LGS-backed lending, however. Some US evidence associates it with expansions in average employment rates of as much as 200 per cent in local markets with a high proportion of minority-owned small businesses (Craig et al., 2007). Fourth, LGSs can be expensive to operate, leading one to wonder whether they are worth the money. It should be pointed out immediately that scheme losses cannot of themselves be construed as prima facie evidence of risky lending, since high administration costs could also be responsible (Rhyne, 1988). With this in mind, Bosworth et al. (1987) conducted a cost-benefit appraisal of the SBA’s LGS. Bosworth et al. (1987) started with an ‘optimistic’ estimate that only one-fifth of SBA-backed loans were non-additional. Applying an empirically based 23 per cent default rate for the whole portfolio, this means that only around 60 per cent of SBA loans could generate benefits to offset the costs of the scheme, requiring a net subsidy of 8.7 per cent. Thus the economic benefits required for the scheme to break even were 14.5 (= 8.7/0.6) per cent of the loan amount – over and above the mandated loan repayments. Furthermore, Rhyne (1988) estimated that in order to cover programme costs, each successful SBAbacked loan would have had to generate additional benefits over the lifetime of the loan

Finance and innovation policies

417

of between 17 and 29 per cent above the return necessary to stay in business. Bosworth et al. (1987) and Rhyne (1988) doubted whether the economic benefits would be as great as this. Furthermore, it is possible that these calculations under-estimate scheme costs, since many business failures and losses emerge several years after the loans are made (and after evaluations are performed). Evidence shows that three aspects of scheme design bear especially closely on default rates and operational costs. One is the guarantee rate (row 1 of Table 16.1). Riding and Haines (2001) show that a small change in the guarantee rate can have a major impact on scheme costs.4 A second instrument which governments can use to control LGS programme costs is loan sizes. Riding and Haines (2001) provide Canadian evidence that default rates are highest for projects financed with the largest loans. A third sensitive programme parameter turns out to be the firm’s age as an eligibility criterion for access to LGS-backed loans. For example, in 2005 the British government withdrew access to the SFLGS for firms which had traded for more than five consecutive years. This had an immediate and dramatic effect on take-up, reducing the number of guaranteed loans from 5,957 in 2005–06 to just 2,844 during 2006 (See the Financial Times, 18 January 2007). In contrast to the SBA’s LGS, though, the UK SFLGS accounted for only 1–2 per cent of all SME on-lending by value before the policy change, making it an already marginal source of lending to the UK SME sector. Overall, the empirical evidence is only mildly supportive of the effectiveness of LGS programmes. Despite the valuable work which has been performed to date, more robust measurement of their additionality, long-run operating costs and optimal design is needed. 16.2

Other credit market interventions

As well as funding the SBA LGS, the US Federal Government also issues direct loans to small businesses, mainly in rural sectors (Gale, 1990b). The scale of direct US federal lending has declined over time, with a move towards guarantees rather than loans. The rationale has been twofold. First, loan guarantees facilitate bank–borrower relationships, unlike direct federal loans; and they also reduce undesirable competition between the SBA and banks (Rhyne, 1988). Between 1991 and 2000, a total of $95 billion in loans to small businesses was made (SBA, 2001). The US federal government also provides interest subsidies to new start-ups. Using a general equilibrium model calibrated to the US economy, Li (2002) predicts that this policy increases the number of entrepreneurs in the targeted category, but reduces the number of entrepreneurs in the non-targeted category. The upshot is a net reduction in entrepreneurial activity. Interest subsidies crowd out capital by decreasing the incentive for would-be entrepreneurs to save, while the taxes needed to finance them blunt incentives. Ultimately, Li (2002) concluded that the policy yields a net output loss. Government would do better by removing the policy, although it could do better still by targeting poor and capable entrepreneurs directly. This once again illustrates the importance of the targeting principle set out in the previous chapter.

418

Public policy

Li’s (2002) results are consistent with other findings too. For example, despite shortterm benefits Bergström (2000) estimates that the long-term effects of capital subsidies on the productivity levels of recipients are negative. Gale (1991) argues that when credit rationing exists, a subsidised loan guarantee is more effective than an interest subsidy, at least on theoretical grounds (Gale, 1990b). Some US state governments provide loans directly to entrepreneurs or in conjunction with private lenders. Some also provide venture capital in the form of equity investments, either directly or indirectly via state-subsidised venture capital companies (Smith and Stutzer, 1989; and see the next section). In the UK, direct government lending is less commonly observed than grants from local or regional agencies, although there are centrally funded grants and equity investment schemes including, most recently, the Enterprise Investment Scheme (EIS). The EIS is a government scheme which provides a range of tax reliefs for investors who subscribe for qualifying shares in small unquoted enterprises trading in the UK. Government grants and loans might be expected to have a positive impact on employment in targeted firms, unless there is crowding out via displacement. Wren (1998) estimated the effects on employment among English manufacturing firms attributable to government capital grants and rent assistance. Wren (1998) reported that grants had larger long-term employment effects on new small firms than on their older and larger firms – despite the lower survival rates of the former relative to the latter. However, the absolute numbers of jobs created were modest. Even before accounting for displacement and opportunity costs of funds, Wren (1998) estimated that capital grants generated just 0.082 job per £1,000 after five years, with an implied gross direct cost per job to the exchequer of £12,195 (= 1,000/0.082). The corresponding job creation rate for large firms was effectively zero. Wren (1998) conjectured that the larger effect for small firms reflected their higher marginal cost of external funds, which made them more responsive to assistance. These findings once again bear out the benefits of policy targeting. Almus (2004) applied the non-parametric matching technique described in note 5 of the previous chapter to evaluate the effects of a German start-up assistance loan programme in the 1990s. Almus (2004) estimated that for a hypothetical firm starting with five employees, loan assistance would help create three extra jobs after six years relative to what would have happened if no assistance had been given. Part of this net job creation reflects improved survival prospects, and part of it reflects superior job creation performance by survivors. However, these estimates do not take account of possible self-selection by programme participants.

16.3

Policies to promote equity finance

This section considers a range of policies designed to promote equity finance in general and venture capital in particular. Part of the rationale is that public policies which stimulate venture capital could enhance innovation owing to the close links between

Finance and innovation policies

419

the two (Kortum and Lerner, 2000). The first part of this section discusses several policies of a regulatory kind, while the second focuses on tax incentives for venture capital.

16.3.1

Regulatory policies One obvious policy to promote equity finance is to reduce the costs of new issues and secondary market transactions (Stoll, 1984). This might be accomplished by improving the efficiency of securities markets and by eliminating unnecessary regulations. Likewise, Initial Public Offerings (IPOs) are a key exit route for VC-backed entrepreneurship, which can be hampered by the non-existence or poor development of public stock markets. It is worth pointing out that the earlier young firms go public, the quicker VCs’ ‘informed capital’ can be redirected towards different new firms (Michelacci and Suarez, 2004). So when informed capital is in limited supply, factors which reduce the costs of start-ups going public encourage overall business creation. This helps makes the case for promoting special stock markets for new start-ups, such as Nasdaq, for example. On the other hand, entrepreneurs who take their firms public have incentives to conceal information from investors, which can deter investment ex ante. This suggests there is a role for regulators to impose mandatory disclosure rules (Östberg, 2006). Another way of increasing the supply of venture capital funds might be by removing legal restrictions on the sources of investor finance, as happened in the case of pension funds in the USA at the end of the 1970s. Governments might also be able to increase financial market efficiency by subsidising agencies dedicated to improving information flows and providing matching services between entrepreneurs and VCs. In early work, Wetzel (1987) emphasised a general problem of poor information about investment opportunities which can cause poor matches between VCs and entrepreneurs and lead to inefficient investment outcomes. However, it is unclear why private sector firms cannot perform this function. Mason (2006) contends that publicly sponsored business angel networks (BANs) serve different clients from commercial BANs; but this might be a direct product of the intervention, and does not necessarily mean that entrepreneurs who currently use publicly subsidised services (such as the SBA’s ACE-Net internet forum which connects business angels and entrepreneurs) would not be served by the private sector if they were withdrawn. Another policy is the creation of public-sector venture capital funds. Although VCs are sophisticated private financial intermediaries, public venture capital programmes might still be able to perform a valuable function by certifying firms to make them more attractive to outside investors, especially in technology-intensive industries where traditional financial measures are of little use. And they can in principle price in the social value of R&D spillovers for small firms which might otherwise choose not to enter the market (Lerner, 2002). In fact, the evidence relating to public-sector venture capital funds is rather discouraging. Countries with a higher proportion of publicly funded venture capital tend to

420

Public policy

have smaller venture capital markets overall (Leleux and Surlemont, 2003; Armour and Cumming, 2006). It is not clear where the causality lies here, but crowding out of private by public funds is one possibility. A striking example of crowding out is observed in the ‘Labour Sponsored Venture Capital Corporations’ (LSVCCs) introduced by the federal government of Canada in the late twentieth century in an attempt to expand the pool of Canadian venture capital. The generous tax subsidies which underpin these schemes can explain their rapid expansion relative to private funds, despite their lower average performance (Cumming and MacIntosh, 2006). On the other hand, Cumming (2007) argues that some public schemes can add value, such as the Australian Innovation Investment Fund programme introduced in 1997 which focuses on government–private sector partnerships. Cumming (2007) reports that this programme facilitated investment in start-up, early-stage and high-tech firms and stimulated the provision of investor mentoring and advice. These studies illustrate how careful design of public policies, and careful consideration of their incentive effects, are crucial for determining their ultimate effectiveness. In the light of these arguments and evidence, what advice can scholars give to governments seeking to design effective public venture capital programmes? Lerner (2002) highlights a common fault of government efforts in this regard: the presumption that technological criteria can be divorced from business considerations when evaluating venture investments. Gompers and Lerner (1999) and Lerner (2002) warn governments against trying to emulate VC-like decision-making on individual projects, a practice akin to ‘picking winners’. Decision-making of this kind requires specialised expertise and profit-seeking motives which government agencies generally lack. In addition, Lerner (2002) chronicles problems of ‘regulatory capture’, whereby powerful entrepreneurs can gain disproportionate benefits from public-sector venture capital schemes. This does not mean that there is no constructive role for public policy. What is necessary is that government officials seek to understand the business environment in which young, high-tech firms operate. Lerner (2002) argues that officials can add value as long as they are willing to learn from the venture capital industry; concentrate on unfashionable but promising ventures rather than on ‘hot’ sectors where there is already plentiful private funding; provide follow-on finance when private flows of capital are drying up; appreciate the need for flexibility in decision-making; and evaluate both high performers and under-achievers. Institutions can also impact the effectiveness of VC-backed entrepreneurship. According to La Porta et al. (1999), for example, countries offering strong protection to investors tend to have broader and deeper capital markets. Leleux and Surlemont (2003) provide specific evidence of this in the context of venture capital: countries with English Common Law traditions possess larger venture capital markets than those with Civil Law traditions. Common Law countries tend to have lower levels of labour market regulation and labour rigidities on average too. This is relevant in view of crosscountry evidence that VCs’ investments in early-stage ventures are negatively affected by labour market rigidities, while later-stage investments are unaffected by them (Jeng

Finance and innovation policies

421

and Wells, 2000; da Rin et al., 2006). For their part, Armour and Cumming (2006) propose a supply-and-demand model in which more liberal bankruptcy provisions stimulate the demand for venture capital, while an index of ‘investor-friendliness’ of a country’s legal and fiscal environment is associated with a greater supply of venture capital investment in entrepreneurial firms. Armour and Cumming (2006) conclude that the legal environment matters as much as the scale and efficiency of stock markets. Da Rin et al. (2006) criticise supply-and-demand models of venture capital as a basis for policy advice. They advocate instead the use of innovation ratios as a dependent variable. Innovation ratios are the ratios of early-stage or high-tech VC to total VC investment. They are a relatively direct measure of the depth of specifically entrepreneurial equity finance. Da Rin et al.’s (2006) fixed-effects regression estimates show that innovation ratios are positively associated with light regulation, the creation of stock markets targeted at entrepreneurs, and low rates of capital gains tax (see below). Interestingly, innovation ratios are invariant to public-sector R&D spending. This seems plausible, as few early-stage entrepreneurial start-ups requiring venture capital can presumably benefit directly from such expenditure.

16.3.2 Taxation policies Tax policies can also influence the impact of venture capital on entrepreneurship. Reflecting the fact that VCs’ rewards primarily take the form of capital gains, rather than dividends or interest income, one potentially important policy instrument is the rate of capital gains tax (CGT). The evidence about CGT effects is mixed, however. On the one hand, there is some econometric evidence of a positive impact of cuts in CGT rates on VCs’ funding commitments and the use of convertible preferred equity in the USA and Canada.5 On the other hand, Poterba (1989a, 1989b) argues that because individuals who are liable to CGT provide relatively little risk capital, it is unlikely that changes in CGT will have much effect on the supply side of the market. Also, the impact of CGT on entrepreneurs’ demand for VC is likely to be limited because entrepreneurs can defer their gains and so reduce their effective CGT rate below the statutory rate. Furthermore, CGT is a blunt instrument, because the VC industry is just a small part of the CGT tax base. For all these reasons, Poterba (1989a, 1989b) argues that CGT is probably unlikely to offer much potential for stimulating the use of venture capital. In a series of important articles, Christian Keuschnigg and Søren Bo Nielsen analyse moral hazard problems that connect venture capital, entrepreneurship and public policy. Keuschnigg and Nielsen (2001, 2003) analyse a ‘single moral hazard’ problem in which VCs and entrepreneurs must both devote effort to the investment project, but where VCs provide inefficiently low effort because they bear all the cost of managerial advice, but obtain only a share of the benefits. Any taxes on project returns further discourage VC advice, though a tax on start-up investment reduces market entry and so increases the equilibrium output price, strengthening VCs’ incentives to provide advice.6

422

Public policy

Subsequently, Keuschnigg (2004a) and Keuschnigg and Nielsen (2004a, 2006) went on to consider the case of ‘double moral hazard’where entrepreneurs as well as VCs have insufficient incentives to supply costly effort. This model is sketched out in the chapter appendix, where their key proposition is derived. This states that, in the absence of corrective taxes, entrepreneurs and VCs both generally devote too little effort because they each appropriate only a part (i.e. the equity shares s and 1 − s, respectively) of the social return while bearing all of their respective private marginal costs of effort. In other words, the profit-sharing inherent in equity contracts impairs incentives. Corrective taxation, in the form of a reduction in the rate of CGT, can strengthen the marginal private returns to effort. In contrast, an increase in the CGT rate reduces the returns to both parties, and so decreases both the number of entrepreneurs and the supply of entrepreneurial goods.7 A cut in the rate of CGT is not the only way of solving the double moral hazard problem. Market-based solutions might also be possible too, such as the use of convertible securities (see Dewatripont et al., 2003; Schmidt, 2003). Keuschnigg and Nielsen point out however that market solutions that depend on effort pre-commitment may be unavailable if (as seems plausible) entrepreneurs lack funds to pre-commit to high effort.Another general solution for the under-provision of effort, proposed byArnott and Stiglitz (1986), is to subsidise outcomes which are complementary with entrepreneurial effort (or to tax those that are substitutes). But an adjustment to CGT looks like a more direct policy in the present context. It is worth pointing out, by way of emphasising the practicality of Keuschnigg and Nielsen’s policy recommendation, that reductions in the CGT rates on entrepreneurs and VCs (τ E and τ F in the chapter appendix) can be self-financing if they are funded by an entry tax or a reduction in the start-up subsidy z. Subsidies to start-up investment costs are not performance-related and do not directly affect effort incentives, because start-up costs are already sunk by the time effort choices are made. They may have an indirect effect, however, by encouraging entry, which reduces the equilibrium output price and hence the VC’s venture returns, so reducing VC managerial advice a further. This consideration could justify imposing a tax on start-up investment, to help finance the recommended cut in CGT rates. Note that this policy set promotes the quality rather than the quantity of entrepreneurship, a theme which is becoming increasingly prominent in policy circles. Keuschnigg and Nielsen’s framework has been extended to deal with related issues. Building on Kanniainen and Keuschnigg’s (2003, 2004) observation that there is a fundamental trade-off between portfolio size and the degree of VC involvement, Keuschnigg (2004b) applies the double moral hazard framework to a problem where VCs also choose their portfolio size. Keuschnigg shows that VC managerial advice (effort) and portfolio size are both generally too small in the private equilibrium, because VCs have to compensate entrepreneurs with a higher equity stake for lower support effort per entrepreneur. Keuschnigg and Nielsen (2005) also analyse the problem of search costs incurred by matching VCs with entrepreneurs. If entrepreneurs and VCs bargain over the surplus from projects, then VCs will provide insufficient (and

Finance and innovation policies

423

inefficiently low) search effort when entrepreneurs’ bargaining power is high – leading to too few deals. Excessive search effort can also be expended when entrepreneurs’ bargaining power is low, however. Other tax policies can also be used to promote VC-backed entrepreneurship. Fuest et al. (2002) studied the relative roles of corporation tax (CT) and income tax (IT). They argued that reducing CT below IT stimulates equity at the expense of debt finance, which is desirable if there happens to be over-investment in debt finance (see chapter 7). Business angels are a relatively overlooked target of government policy in this area, despite their prominent role in financing entrepreneurial companies (chapter 8). It is instructive that over one-half of technology firms in Ottawa, Canada, which had received funding from business angels went on to raise institutional finance from VCs, compared with only one-tenth which had raised no angel finance (Madill et al., 2005). Mason and Harrison (2000a) review programmes in several countries designed to address ‘equity gaps’ relating to business angels. These include tax breaks to encourage individuals to become business angels; initiatives designed to increase entrepreneurs’ awareness of angels; and the allocation of public money to private VC funds to induce them to make investments that would not otherwise be made.8 Mason (2006) claims that tax incentives have been useful for angels, with some estimates putting additionality from this policy at over 50 per cent.9 It is appropriate to close this discussion with a caveat about the use of taxation as a policy instrument to stimulate VC-backed entrepreneurship. Gompers et al. (2005) argue that policies designed to foster entrepreneurial and VC activity by providing investment incentives might be less important than access to networks of other entrepreneurs, suppliers and customers. In their view, ‘regions may need to attract firms with existing pools of workers who have the training and conditioning to become entrepreneurs’ (Gompers et al., 2005, p. 612). The basic idea is that if the talent is there, the funds will follow, regardless of whether tax incentives are available. 16.4

Innovation policy and entrepreneurship

The previous chapter mentioned several reasons why innovation policies might be able to improve economic efficiency and social welfare. While a plethora of different innovation policies exist, which I lack the space to review, it is worthwhile to mention two in particular. These are the US Small Business Innovation Research (SBIR) programme and the Small Business Technology Transfer (STTR) programme.10 These are perhaps the best-known and most widely admired entrepreneurial innovation policies in the world. The SBIR was established in 1982 to address concerns about the declining competitiveness of US industry. The SBIR legislation aimed to increase the share of procurement contracts going to small firms from the largest federal R&D agencies and to increase the amount of commercialisation of federally funded research. Under SBIR, all federal agencies with extramural R&D budgets exceeding $100 million annually are required by law to set aside 2.5 per cent of those budgets with small US companies via a

424

Public policy

competitive bidding process. The expenditure is on contracts for the development of new technology needed by the agencies. SBIR is a seed fund which significantly reduces the risk faced by private equity providers. The SBIR is the federal government’s most important R&D funding programme for small business. In December 2000, the programme was re-authorised by Congress for another seven years with a budget of $1.5 billion in annual funding. It came up for renewal again in 2008. The STTR programme requires federal agencies with extramural R&D budgets exceeding $1 billion annually to set aside 0.3 per cent of those budgets for co-operative research between small businesses and a federal laboratory or non-profit research institution. Eleven federal departments participate in the SBIR programme and five participate in the STTR programme; together they award $2 billion annually to small high-tech businesses. Lerner (1999) analysed a matched sample of firms, some of which received public funds and some of which did not, to evaluate the SBIR. Lerner (1999) concluded that the SBIR enhances small firm performance. The SBIR programme is often used as a commercialisation channel by biomedical academic scientists, promoting successful academic entrepreneurship by providing start-up finance and stipulating full-time work commitments from the entrepreneurs (Toole and Czarnitzki, 2007). However, other studies have been unable to link the SBIR to additional employment or private sector R&D expenditures, which it may crowd out by as much as one-for-one (Wallstein, 2000). External observers who attribute American innovative entrepreneurial dynamism to free and competitive markets sometimes overlook the major role that the US federal government plays in funding innovation. According to Auerswald (2006), even excluding the impact of federal procurement, roughly 20–25 per cent of early-stage technical development funds come from federal sources. Arguably, though, greater attention has been paid to the Bayh–Dole Act of 1980, which made it easier and more profitable for US university researchers to exploit commercial opportunities arising from their research. After the Bayh–Dole Act was passed, notably higher rates of patenting and licensing by academics were observed. This has generated enormous interest in policy circles and may have prompted other countries to pass similar legislation. However, Mowery et al. (2004) assert that there is little or no evidence the Bayh–Dole Act had a causal effect on entrepreneurial innovation. Mowery et al. (2004) speculate instead that the growth in US academic entrepreneurship would have taken place without it. They further point to a long history of collaboration and interaction between university researchers and industry, which is more or less unique to the USA, and which primarily contributed to the formation of the political coalitions that led to the passage of the Bayh–Dole Act – and which may have contributed to the successes that followed. According to Mowery et al. (2004), Bayh–Dole should therefore be regarded as an endogenous rather than an exogenous policy change. If so, it seems unlikely that it can be replicated with the same efficacy elsewhere. A consistent message to emerge from this review of the literature is that effective entrepreneurship policies should target genuinely innovative sectors, rather than broad

Finance and innovation policies

425

sectors where business entry entails mundane replication by ‘revolving door’ firms. It seems likely that market failures are more prevalent in innovation than in most other areas of economic activity, for the reasons (and with the caveats) outlined in the previous chapter. It has certainly not been established that markets are failing to facilitate entry by ‘replicative’ firms. A harder challenge is to frame policy in the context of innovative spin-offs. As noted earlier in this book, large successful firms are often started by exceptionally talented people who garner experience from market-leading incumbents before spinning off an innovative new firm in a closely related industry. This has been observed, for example, among the early founders of the US TV industry (who started in radio); the US drugs industry (spun off from penicillin); and the US automobile tyre industry (spun off from rubber) – among many others. Governments looking to promote high-growth entrepreneurship might therefore be advised to foster a few great firms as generators of tomorrow’s stock of innovative entrepreneurship capital, rather than trying to encourage lots of start-ups today – many of which are merely replicative. Picking up on an earlier theme, policy-makers need to focus more on the quality than the quantity of entrepreneurship. Perhaps even more contentiously, governments should not expect entrants to provide competitive discipline to incumbents in capital-intensive industries. The evidence about industry shakeouts shows that very few entrants can come to challenge the dominant market leaders in such industries. Governments might therefore need to be less concerned about anti-trust issues in the short term, in order to foster environments favourable to entrepreneurial spinoffs and successful domestic entrepreneurship in the long term. Governments can further encourage spinoffs by removing employee non-compete covenants and weakening trade-secret laws. I will say more about the impact of legal and regulatory interventions in the next chapter.

16.5 Appendix: the Keuschnigg–Nielsen (2004a, 2006) double moral hazard model

Keuschnigg and Nielsen consider a static one-shot setting in which new VC-backed firms supply goods alongside established firms. If individuals become entrepreneurs, their projects succeed and yield payoffs of V with probability p = p(e, a) ∝ e aα ,

 + α < 1,

(16.1)

which depends on the entrepreneur’s work effort e and the VC’s support effort a.11 Importantly, p(·, ·) is concave and increasing in e and a: pee , paa < 0 < pe , pa . Hence the two kinds of effort are strategic complements. If projects fail they yield zero. VCs offer start-up finance of I (which costs them zI less if a proportional subsidy z is available from the government) and an upfront salary of y ≥ 0 to entrepreneurs in return for taking an equity stake of 1 − s (0 < s < 1) in their enterprise. To be attractive, entrepreneurship must offer an expected income (inclusive of an income subsidy ς E but net of an effort cost βe) that is higher than that in paid

426

Public policy

employment, where a wage of unity plus an income subsidy ς L is available: PC E :

(1 − τ E )[p(e, a)sV + y] + ς E − βe ≥ 1 + ς L ,

(16.2)

where PC E denotes that this is the entrepreneur’s participation constraint, and τ E denotes the rate of capital gains tax (CGT) levied on entrepreneurs’ payoffs. VCs and entrepreneurs then simultaneously supply effort subject to double moral hazard. VCs have the objective max{(1 − τ F )[(1 − s)p(e, a)V − y − (1 − z)I ] − γ a}, s,a

(16.3)

where the VCs’ marginal cost of effort is γ , and where τ F is their CGT rate. Differentiating (16.2) and (16.3) with respect to e and a respectively yields the following incentive compatibility constraints: pe (e, a)(1 − τ E )sV − β = 0

(16.4)

pa (e, a)(1 − τ F )(1 − s)V − γ = 0

(16.5)

Competition among VCs squeezes the optimised value of (16.3) to zero. This can be solved along with (16.1), (16.2), (16.4) and (16.5) to yield the optimal contract  = (s∗ , y∗ ). It turns out in this model that s∗ depends only on the parameters α and , and so can be treated as a constant. Keuschnigg and Nielsen’s key result is contained in the following proposition: Proposition 1 (Keuschnigg and Nielsen double moral hazard) (a) In the untaxed equilibrium, both entrepreneurial and VC effort levels are too low: e < e∗ and a < a∗ . (b) Optimal CGT subsidies that restore the first best levels of effort exist, and are given by (16.6) (1 − τ E∗ )s∗ = (1 − τ F∗ )(1 − s∗ ) = 1 . Proof (a) The joint surplus of each start-up project is pV − βe − γ a − I − 1 (recall that the foregone wage is unity). The first-best effort levels maximise this surplus and are given by pa (e∗ , a∗ )V = γ . (16.7) pe (e∗ , a∗ )V = β, Comparing (16.7) with the untaxed versions of (16.4) and (16.5) yield the result. (b) Conditions (16.7) and (16.4), (16.5) yield e = e∗ and a = a∗ only if (16.6) holds.
Notes 1. For example, US variants of the basic scheme include the Certified Lender Program (which promises banks a three-day approval time) and the Preferred Lender Program (under which loans can be made without prior approval by the Small Business Administration (SBA), at the price of a smaller guarantee).According to Smith and Stutzer (1989), about one-third of state governments in the USA also guarantee loans, with guarantee percentages varying within and between states. The

Finance and innovation policies

427

US SBA LGS dates back to 1953; Rhyne (1988) provides a review of the historical and political context. The UK Small Firms LGS (SFLGS) dates back to 1981: see Cowling and Clay (1995) for a full background to this scheme, operational details and an econometric model of take-up rates by entrepreneurs. Bates (1984) describes the Economic Opportunity Loan programme. 2. Proscribed sectors include banking, real estate development and publishing in the USA; and retailing, catering and motor vehicle maintenance in the UK. One reason for excluding support in certain sectors is to reduce business displacement rates (see below). 3. The Riding et al. (2007) estimate was generated from a binary credit scoring model Di = Xi β +ui where Di = 1 if SME i received a non-guaranteed loan and 0 if it was turned down for a nonguaranteed loan. Riding et al. claimed a loan was incremental if, for a sample of SMEs funded ˆ i = 0 (i.e. one would conditionally predict the loan would not be made by a private under CSBF, D bank) but Di = 1 because of the programme. However, the problem with this measurement approach is that it ignores samples selection issues and assumes that the specification of the model for Di is correct. While Riding et al.’s model exhibited high goodness-of-fit, this problem remains non-trivial. 4. To see how, let δg and δng be the proportion of guaranteed and non-guaranteed loans respectively which default. Then by equating bank profits both with and without the scheme Riding and Haines (2001) show that 1 d ln(δg /δng ) ≈ dβ , 1−β

5. 6. 7.

8.

9. 10.

11.

where β is the guarantee rate (e.g. β = 0.90 if 90 per cent of a loan is underwritten). So if β is reduced from 0.90 to 0.85 (i.e. d β = −0.05), the reduction in guaranteed loan defaults is (approximately) as large as 50 per cent. The reason for this sensitivity is that banks can make the same profits when the government increases β by making riskier loans and so tolerating higher default rates. Gompers and Lerner (1998), Cumming (2005), da Rin et al. (2006). See also Secrieru and Vigneault (2004), who analyse a public-sector rather than a private-sector VC under asymmetric information about agents’ types. See (16.4) and (16.5) in the appendix. Because the function p(e, a) described there is concave, an increase in CGT worsens rather than improves incentives by reducing privately chosen e and a further and hence [by (16.1)] the proportion of start-ups that succeed. In general equilibrium, where a given aggregate demand for goods has to be supplied, the output price of entrepreneurial goods must rise to clear the goods market. So inefficient effort, which is exacerbated by an increase in the rate of CGT, also reduces consumer surplus and thereby social welfare. For example, ‘I-tec’ is a public policy launched in the UK in 1997 to provide grants to cover VCs’ due diligence costs, in order to stimulate funding of more seed and early-stage ventures. But this policy has apparently suffered from heavy administration costs. The principal UK tax incentive scheme is the Enterprise Investment Scheme (see above); some US states also offer tax incentives (Lipper and Sommer, 2002). See Acs and Armington (2006, chap. 7) for a description of several micro US policies also aimed at promoting firm formation and growth of small firms involved in the commercialisation of new knowledge. Keuschnigg (2004a) studied the case where the coefficient of proportionality in (16.1) increases in VCs’ experience and industry knowledge.

17 Taxation, regulation and other policies

This final chapter discusses several other public policies which affect entrepreneurship. The first section discusses the taxation of entrepreneurs. It starts by reviewing theory and evidence about the effects of income and payroll taxes, before turning to corporation, wealth and inheritance taxes. The content of the first section complements the analysis of capital gains tax provided in the previous chapter. The second section discusses evidence relating to three types of public policy which affect entrepreneurs’ labour and product market activities. The first type of policy is employment assistance schemes, designed to move unemployed workers off the welfare rolls and into independent business ownership. The second type is targeted government procurement of entrepreneurs’ products, including affirmative action policies for small business owners. The third type of policy is the provision of assistance and advice to entrepreneurs. The third section of this chapter deals with an increasingly important topic: government regulation. Regulations affect entrepreneurial entry, exit, production, hiring and corporate governance decisions, as well as the availability of bank credit. These topics are all analysed in turn. The emphasis in this section is mainly on presenting empirical findings, although some theoretical arguments are also discussed. The fourth section discusses some remaining, mainly macro, policy issues such as the welfare state and minimum wages – and how they affect entrepreneurship. It also discusses the effects of several institutions which, while not directly under government control, are shaped by government policies, including the roles of ‘enterprise culture’ and trade unions in the economy. The fifth section concludes. 17.1 Taxation

This section deals with the taxation of entrepreneurs. Capital gains tax was already discussed in chapter 16, in the context of venture capital; the present section deals with the other major taxes that affect entrepreneurs. They include income and payroll taxes (denoted by IT and PT hereafter); corporation tax (CT); and wealth and inheritance taxes (WT and IHT, respectively). I commence with a brief theoretical overview of how IT and PT affect entrepreneurial choices. The simplest case of linear IT and PT will be 428

Taxation, regulation and other policies

429

analysed before turning to more complicated, non-linear tax systems. Then empirical evidence on the effects of IT and PT on entrepreneurs is discussed. The remaining subsections deal with CT, WT and finally IHT. 17.1.1 Income and payroll taxes: theoretical issues The material in this section builds on that of chapter 2, which analysed models of occupational choice and participation in entrepreneurship. Those models did not incorporate government activities, an omission that is rectified below. For notational brevity, I will refer below to IT rather than to ‘IT and PT’ because the theoretical effects of the two taxes are so similar. More of a distinction will be made between them when the empirical evidence is assessed. One reason why governments might seek to influence occupational choices emerges from the model of Kihlstrom and Laffont (1979). As discussed in chapter 2, Kihlstrom and Laffont (1979) show that there might be too little or too much entrepreneurship when there are missing or incomplete markets for risk-sharing. That might justify government intervention. Taxation seems an attractive instrument in this regard compared with government efforts to eliminate risk directly, for example by insuring entrepreneurs’ incomes. Such efforts can be costly, and governments might not have a cost advantage in providing insurance over the market (Black and de Meza, 1997). Governments often presume that higher taxes impede entrepreneurship. For example, when reviewing the impact of UK IT rate cuts in the 1980s, the UK Employment Department concluded that ‘a more beneficial tax regime has been an important factor in stimulating enterprise’ (1989, p. 16). Indeed, the simplest theoretical models which abstract from risk do straightforwardly suggest that higher IT rates on entrepreneurial incomes decrease the number of firms, and disadvantage the smallest at the expense of the largest ventures.1 But when one takes account of risk, more subtle results emerge. The model of Kanbur (1981) provides a good entry point to the theoretical literature on risk-taking, taxation and entrepreneurship. Assuming that entrepreneurs earn risky profits while employees earn a safe wage, and that both types of agent are riskaverse, Kanbur asks a simple question. If social welfare is defined as the sum of entrepreneurs’ and employees’ utility, would a tax on entrepreneurs whose proceeds are recycled to employees via a wage subsidy increase aggregate social welfare? Kanbur (1981) shows that it would.2 The reason is that although the tax reduces net profits for entrepreneurs, it also reduces the variability of those profits, which has an offsetting positive ‘insurance’ effect for the entrepreneurs. In contrast, employees gain unambiguously from the subsidy financed by the entrepreneurship tax, and become more numerous as individuals switch out of entrepreneurship and into wage work.3 The gain made by each employee more than offsets the loss incurred by each entrepreneur, so aggregate social welfare increases. Kanbur (1981) interpreted this to mean that there are too many entrepreneurs in the no-tax equilibrium relative to the higher welfare equilibrium with an entrepreneurship tax. This result is not as clear-cut as it seems, however. Simple supply and demand logic suggests that the price of the risky-sector good consumed by employees will increase

430

Public policy

as the number of entrepreneurs (aggregate supply) declines. Black and de Meza (1997) show that this ‘general equilibrium’ effect can make employees and hence society as a whole net losers from this policy, in which case entrepreneurs should be subsidised rather than taxed. The analysis becomes more complicated still if both occupations are subjected to a common proportional tax (with tax rate 0 ≤ τ ≤ 1) together with a common lump-sum subsidy ς > 0 financed from the tax proceeds.4 Consider a policy of taxing away all income in both occupations (τ = 1) and replacing it with lump-sum subsidies. This policy provides complete insurance to risk-prone entrepreneurs and so maximises social welfare as defined above. But in contrast to the earlier results, the effect on the equilibrium number of entrepreneurs is now ambiguous. And when entrepreneurs gain a non-pecuniary benefit from autonomy, as the evidence suggests (see chapter 4), the optimal tax rate is less than one, reflecting a trade-off between equality, efficiency and insurance considerations (Boadway et al., 1991).5 The models discussed so far study linear taxation. In contrast, the tax schedules in most countries are piecewise linear and progressive. It turns out that the impact of progressive taxation (e.g. relative to linear taxation) on incentives to become an entrepreneur depends on a feature of the tax system known as ‘loss offsetting’. A ‘loss offset’ allows entrepreneurs who make operating losses to apply these losses against income subject to taxation from other sources. Domar and Musgrave (1944) first showed that total investment in risky activities (such as entrepreneurship) can increase under a proportional IT with full loss offsets. Because progressive taxation with complete loss offsets not only confers this benefit but also smoothes entrepreneurs’ risky incomes, it follows that progressive taxation with full loss offsets will encourage entrepreneurship even more – and can in some models imply increased economic growth (Garcia-Penalosa and Wen, 2008). More generally, public insurance can be explicit (e.g. via public ownership of entrepreneurs’ productive inputs) as well as implicit (e.g. via progressive taxation). An interesting historical example of public insurance is Renaissance Venice, whose maritime merchants reaped ‘astronomical gains from trade’ (Sinn, 1995, p. 264) which helped make Venice the richest city in the world: ‘while private insurance covered the loss of cargo, the galleys were owned by the state, which completely absorbed their risk of loss and destruction’ (Sinn, 1995, p. 265). If loss offsets are restricted or absent, progressive taxation can be expected to discourage entrepreneurship, by taxing profits of successful projects at a higher rate than the rate at which losses can be deducted. This serves as a disincentive to risk-taking which can overwhelm the countervailing ‘insurance’ effect from smoother post-tax incomes under progressivity. Gentry and Hubbard (2000) claimed to find evidence that more progressive taxes are indeed associated with lower rates of entrepreneurship in the USA, although the evidence on this issue is in fact mixed (see also Kanbur, 1979; Yuengert, 1995). At this juncture it is necessary to register a caveat to the argument that entrepreneurs’ average utility and social welfare increase when IT confers insurance. This argument

Taxation, regulation and other policies

431

does not necessarily follow if risk-bearing fulfils a socially useful purpose, as it can do in a ‘second best’ environment with moral hazard (Banerji and van Long, 2007). A standard result from agency theory is that principals generally need to subject agents to some income variation, in order to incentivise them to provide discretionary effort. One can observe this principle at work for example when VCs incentivise entrepreneurs by paying them a combination of guaranteed base salary and risky profit-share (see the appendix to chapter 16). If the government imposes an income-smoothing tax with social insurance characteristics, VCs optimally respond by making pre-tax incomes even more variable in order to restore incentives for entrepreneurs to supply discretionary effort (Keuschnigg and Nielsen, 2004a). So this particular policy is neutralised by the responses of the private sector: the public sector cannot in this case improve on the market allocation of risk. This caveat should not be taken too far, though, because IT neutrality is not a general property of models of entrepreneurship under moral hazard, even specifically models of VC-backed entrepreneurship. VCs do not fully internalise the cost of moral hazard arising from the increased insurance created by a profit tax. The government also bears some of this cost, because lower entrepreneurial effort increases the average rate of project failures and hence losses of government tax revenue. In cases like this, higher taxes can once again have real effects, reducing entrepreneurial effort and increasing business failure rates (Hagen and Sannarnes, 2007). The effects of progressive taxation on entrepreneurship also depend on risk preferences.Any insurance effect of taxation which increases the number of entrepreneurs also increases the aggregate demand for labour by the entrepreneurial sector. This increases the wage in paid employment, which increases the attractiveness of that occupation too. Entrepreneurs have to be ‘sufficiently’ risk-averse to value the insurance from progressive taxation in order for entrepreneurship to increase on net.6 Using a framework of this kind, Meh (2005) studied the effects of a switch from progressive to flat-rate taxation, and predicted few long-run effects on overall wealth inequality in a calibration exercise for the US economy. So far, we have chiefly examined IT systems which treat entrepreneurs and employees the same. In practice, although entrepreneurs and employees usually face similar IT schedules, there can be some differences in the tax treatment of these occupations. On the one hand, entrepreneurs are disadvantaged by some aspects of the IT system, including incomplete loss-offsetting for new entrants compared with incumbents; profit-insensitive taxes which hit small firms harder than large ones; and the fact that new ventures raising equity finance are liable to dividend taxation whereas incumbents can often finance investment out of (untaxed) retained profits (Boadway and Tremblay, 2005). More generally, institutional ownership of firms is frequently tax-favoured relative to direct ownership – the most common organisational form in entrepreneurship (Henrekson, 2005). An additional problem is that governments have incentives to tax fixed factors of production, such as capital already invested. On grounds of horizontal equity, new capital has to be treated the same way as capital in existing firms: this also deters entry (Boadway and Tremblay, 2005). Unlike incumbents, start-ups are often in

432

Public policy

a weak position to lobby for tax exemptions, putting them at a further disadvantage. And in some countries, entrepreneurs face additional taxes which other taxpayers do not, e.g. the self-employment SECA (Self-Employment Contributions Act) tax in the USA (Joulfaian and Rider, 1998). Entrepreneurs do receive advantageous IT treatment in other respects, however. First, they can legally deduct tax on various business expenses, including private pension contributions (e.g. Keogh plans in the USA); and they can sometimes enjoy capital gains exemptions (see Holtz-Eakin, 2000). Second, in some tax jurisdictions entrepreneurs (unlike employees) can benefit from various forms of tax planning. One example is ‘income splitting’, whereby a high-earning entrepreneur can transfer profit to the lower-earning spouse through their business in order to minimise their tax liabilities. Schuetze (2006) estimates that income splitting costs the Canadian exchequer about C$500 million per annum. Another example is the ability of self-employed people to draw up accounts which enable them to practise intertemporal tax shifting (Sillamaa and Veall, 2001). Third, in many countries entrepreneurs pay lower social security contributions, albeit in return for lower social security entitlements (OECD, 1994b). For example, although the self-employed are eligible for social security pensions in almost all countries, they are usually ineligible for state-provided accident, sickness and unemployment insurance. In contrast, employees are often eligible for all of these benefits.7 It should also be remembered that entrepreneurs enjoy greater opportunities to practise illegal tax evasion than employees do, implying that entrepreneurs face lower effective tax burdens than employees who face the same tax system and who earn the same incomes (see chapter 13). Theory suggests, not surprisingly, that incentives to evade and to be an entrepreneur both decrease when the government audits entrepreneurial incomes more intensively (Watson, 1985; Jung et al., 1994). But changes in income tax rates have ambiguous effects on the number of entrepreneurs when there are evasion opportunities. An increase in the rate of IT increases the expected benefit of evasion. But the resulting reduction in post-tax incomes makes entrepreneurs with DARA preferences less willing to bear the risk that tax evasion entails (Jung et al., 1994). A different question is what income taxes and enforcement policies the government should choose in the presence of tax evasion and free occupational choice. Apart from a paper by Pestieau and Possen (1991), relatively little attention has been paid to optimal auditing intensities. Only slightly greater research effort has been devoted to the question of whether entrepreneurs ought to face the same IT schedule as employees (Pestieau and Possen, 1992; Parker, 1999a). Parker (1999a) calibrates a simple model of occupational choice using data from the UK economy and solves for the optimal linear differential IT rates that maximise an inequality-averse social welfare function. He finds that UK governments should tax entrepreneurs at a lower marginal rate than employees, to encourage entry into entrepreneurship. The reason is that efficiency effects of IT dominate inequality and insurance considerations in his calibration exercise (see also Parker, 2001).

Taxation, regulation and other policies

433

Parker (1999a) studied a linear differential taxation model because non-linear models are complex and yield few concrete predictions (see Moresi, 1998, for an example). Parker (1999a) also allowed for tax evasion opportunities in entrepreneurship. Evasion opportunities are not directly productive but draw too many people into entrepreneurship for the social good. However, Parker (1999a) did not consider the possibility that greater numbers of entrepreneurs depress average returns in entrepreneurship, which intensifies competition and so hits honest entrepreneurs the hardest, possibly forcing them to become dishonest as well in order to survive. That can be expected to exacerbate tax evasion and resource allocation problems and might even help to explain the high tax evasion rates typically practised by the self-employed (see chapter 13). This provides a counter-argument to Parker’s (1999a) policy recommendation of a differential tax which favours entrepreneurship. But as noted above, the effects of IT rates on the number of entrepreneurs are ambiguous when tax evasion opportunities exist. To conclude, the theoretical literature demonstrates that the effects of income taxes on the equilibrium number of entrepreneurs and social welfare depend on several key assumptions about the structure of the economy. When entrepreneurs are heterogeneous and face risky returns, income taxation affects those returns and aggregate social welfare in diverse ways. One effect is on efficiency, especially if taxes are distortionary. A second effect is on inequality, because tax systems redistribute incomes. Another effect is on risk-sharing, because income taxes can effectively insure entrepreneurs. These effects may offset each other; and even strong restrictions on preferences and technology cannot be guaranteed to simplify matters enough to deliver unambiguous predictions. It is sometimes easier to predict the effects of a tax policy on incentives to become an entrepreneur than on social welfare. This is another reason why one must exercise caution when making policy recommendations in this area. For example, Keuschnigg and Nielsen (2001) show that while government spending on entrepreneurial training and subsidies to equipment investment and output all stimulate entrepreneurship, they do not enhance social welfare; only cost-effective government services can do that. Presumably, social welfare rather than the number of entrepreneurs per se is the appropriate objective of policy-makers. The inescapable conclusion is that theory alone cannot resolve the question of appropriate taxation of entrepreneurs’ incomes. Data are needed to identify the salient effects. I now turn to the evidence on the empirical relationship between entrepreneurship and income and payroll taxes.

17.1.2

Income and payroll taxes: empirical evidence Several researchers have tested the hypothesis that income and payroll taxes affect participation in entrepreneurship. Most of these tests involve estimating a binary choice model (see chapter 3) with one or more IT or PT variables added as explanatory variables. An important question concerns what tax variables to use. Researchers have used

434

Public policy

several, including (i) tax liabilities faced by employees; (ii) average tax rates (ATR) and/or marginal tax rates (MTR) on all incomes liable to tax; and (iii) tax rates on entrepreneurial incomes relative to employee incomes. The rationale for these variables is usually that tax evasion opportunities in entrepreneurship become more attractive the higher tax rates are. Robson and Wren (1999) suggest that MTRs reflect incentives to supply effort, whereas ATRs capture tax evasion incentives. Hence both variables should arguably be included in binary choice models. The MTR can be thought of as capturing a ‘substitution’ effect and the ATR an ‘income’ effect (see chapter 12). In fact, empirical estimates of aggregate marginal and average tax rate effects invariably compound income and substitution effects (Georgellis and Wall, 2006). Approaches (i), (ii) and (iii) above have been implemented using the cross-section, panel and time-series econometric models discussed in chapter 3 (see Schuetze and Bruce, 2004, for a taxonomy). Overall, the various methods have generated mixed results about the effects of IT rates on rates of participation in entrepreneurship (usually measured as self-employment).8 Researchers need to take various methodological issues into account when they estimate empirical models of entrepreneurship and taxation. These include attention to endogeneity; the possibility of model non-linearity; and awareness of variations in tax compliance enforcement. I will discuss these issues in turn with respect to IT before discussing PT. The practice of using individual incomes to measure tax liabilities is prone to endogeneity bias, because the tax rates individuals end up facing are not random but (to the extent that discretionary effort determines incomes) are under the control of the individuals themselves. One way around this problem is to exploit regional and temporal variations in the tax code. For example, Schuetze (2000) computed median and 90th percentile incomes in Canada and the USA averaged over several time periods, and then used these to calculate tax liabilities which vary with exogenous changes in the tax code (not average incomes) across countries, years, states and provinces. Using this method, Schuetze (2000) estimated that an increase of 10 per cent in the ATR in one year increases the probability of becoming self-employed in the next year of between 2.1 and 3.7 per cent in the USA, and 1.6–3.0 per cent in Canada.9 These are fairly sizeable effects. Schuetze (2000) (and Blau, 1987) found that changes to the IT system explain a greater proportion of changes in North American self-employment rates than any other macroeconomic or demographic explanatory variable. Most empirical studies enter tax variables in a linear regression framework. Yet the relationship between entrepreneurship participation and MTR might be non-linear (Parker, 2003b; Georgellis and Wall, 2006). Researchers should therefore consider estimating non-linear or polynomial models when including income tax variables. In his empirical analysis of a non-linear model estimated using British micro-data, Parker (2003b) found no evidence that entrepreneurship is affected by pecuniary factors in general, let alone IT in particular. IT rates might have mixed and weak effects on entrepreneurship if their effects are obscured by the degree of tax compliance enforcement (TCE) (Bruce and Mohsin,

Taxation, regulation and other policies

435

2006). For example, if TCE is lax, higher IT rates can encourage entrepreneurship as it is easier to evade taxes there than in paid employment. Whereas if TCE is strict and entrepreneurial incomes are more sensitive to effort than incomes in paid employment are, higher IT rates can discourage entrepreneurship. The enforcement of tax law can be an especially salient consideration in transition economies (Estrin et al., 2006). The importance of tax enforcement also becomes obvious when one inspects the time-series of self-employment rates in the British construction industry (Briscoe et al., 2000). Regimes in which self-employed tax compliance are policed more carefully are associated with noticeably lower self-employment rates. Torrini (2005) regressed self-employment rates on IT rates, where IT rates were interacted with an inverse proxy for TCE, namely a national index of perceived corruption. Torrini (2005) found the interaction term to be positive and significant, consistent with the above discussion. Most of the empirical results discussed so far refer only to IT. A few separate studies analyse PT. These studies tend to argue that PT encourages entrepreneurship in general and self-employment in particular. If employers can pass some of the payroll tax on to workers, this decreases wages in paid employment and so makes entrepreneurship more attractive (Stabile, 2004). To the extent that they cannot pass these taxes on, employers have incentives to substitute self-employed contract labour for regular employees (see chapter 2). On the other hand, payroll taxes might discourage employees from becoming entrepreneurs if they expect to become employers later on (Parker and Robson, 2004). In fact, the evidence generally finds a significant positive relationship between PT and entrepreneurship.10 Even less attention has been paid to the effects of taxation on entrepreneurial decisionmaking and venture performance. There are some exceptions, though. Carroll et al. (2000b, 2001) analysed effects of changes in IT rates on entrepreneurs’ business receipts, investment outlays and survival prospects. These authors estimated that if the MTR of a sole proprietor drops from 50 per cent to 33 per cent, business receipts increase on average by 28 per cent, while an increase in the MTR by 5 percentage points reduces the mean probability of making a business investment by 10.4 per cent, with mean investment expenditure dropping by 9.9 per cent. This amounts to a sizeable effect in aggregate, because entrepreneurial ventures account for at least 10 per cent of the US economy’s total amount of non-residential fixed investment. This evidence has been buttressed by an independent calibration study which estimates that a tax reform that cuts taxes on business incomes stimulates entrepreneurial investments, especially in economies prone to financial market imperfections (Kitao, 2008). And although Carroll et al. (2000b, 2001) detected no evidence that IT rates affect entrepreneurs’ survival prospects, a subsequent analysis estimates that an across-the-board income tax cut would significantly increase entrepreneurial survival prospects (Gurley-Calvez and Bruce, 2008). To conclude, the evidence obtained to date has not clearly established that individuals’ decisions to become entrepreneurs are shaped by the IT system. Nor is it clear why the evidence is so inconclusive. It could be that the gulf between theory and empirical

436

Public policy

work in this area is simply too wide. It is notable, for example, that empirical models are quite ad hoc in structure and rarely isolate the efficiency, equality and insurance effects of taxation highlighted in theoretical work. Of course, the conflicting nature of these three effects might itself explain the conflicting nature of the evidence. In practice, much of the motivation for the empirical work on IT is provided by tax evasion (see chapter 13). Stronger positive effects on entrepreneurial participation rates have been detected for PT, while IT seems to have greater effects on the investment decisions and tax evasion choices of existing entrepreneurs. But the research here is sparse and further analysis is needed before more definitive conclusions can be reached.

17.1.3

Corporation tax

Entrepreneurs who do not incorporate their businesses pay IT. Entrepreneurs who do incorporate pay CT on their business profits and can choose to either pay themselves an employee salary as a director of their company, or dividends (to avoid payroll taxes). In most countries, CT is a flat-rate tax, unlike IT which tends to be progressive with marginal tax rates that rise with income. This raises the issue of optimal incorporation choices by entrepreneurs, since there is usually a tax incentive to reclassify earnings as corporate rather than personal. By incorporating, entrepreneurs can convert highly taxed labour income into more lightly taxed dividends, and can also enjoy greater than full loss-offsets (Fuest et al., 2002; Cullen and Gordon, 2007). This creates a tax incentive to become an entrepreneur and suggests that ventures are likely to start out under IT but increasingly move to CT as they grow (Gordon, 1998; Cullen and Gordon, 2007). Some evidence supports these arguments. An Israeli study finds that significant numbers of entrepreneurs in the top 1 per cent of the income distribution responded to increases in IT rates by incorporating (Romanov, 2006). And cross-country evidence shows that business entry rates are significantly higher in countries where CT rates are lower than IT rates (Klapper et al., 2006). Incorporation of a business renders it susceptible to company law, which requires the owner to publicise stipulated data about the business. In the UK and USA, for example, publication requirements include audited accounts if business turnover exceeds a specified level. Despite the costs and inconvenience involved, there are several advantages to incorporation, over and above the tax savings noted above. Incorporation furnishes protection from creditors via limited liability; entails favourable pension contribution rules; and can foster greater credibility with customers (Storey, 1994b). Fuest et al. (2002) ask whether it is socially optimal for entrepreneurs to face lower CT than IT rates. They argue that it is if debt finance is costlier than equity finance, as lower CT rates benefit shareholders and so promote cheap equity in place of costly debt. In practice, however, very few entrepreneurs use equity finance (see chapter 8). Hence some other reason probably explains why CT rates are lower than IT rates. Perhaps one reason is that this is the easiest way to tax-favour successful entrepreneurs without

Taxation, regulation and other policies

437

placing horizontal inequity into the IT system. Concerns about international corporate tax competition might furnish another reason. Finally, some Nordic countries possess ‘dual’ IT systems which split an entrepreneur’s income into capital and labour components and tax each differently. As in the above discussion, at high incomes (flat) capital tax rates are lower than (progressive) income tax rates. As long as double taxation of dividends is not too severe, a dual IT system encourages the formation of new enterprises by entrepreneurs who anticipate high profits (Kanniainen et al., 2007). This policy can be efficient provided that the dual IT system is not merely exploited as a tax shelter.

17.1.4 Wealth and inheritance taxes Theoretical research on wealth taxes is relatively sparse. I am aware only of a recent article by Grüner and Schils (2007) which endogenises choices about wealth redistribution via a wealth tax in a political-economy voting model with borrowing constraints. These authors point out that a progressive redistributive wealth tax hits the upper class the hardest and increases the amount of capital they have to borrow in order to become entrepreneurs. Anticipating that larger loan sizes give entrepreneurs greater incentives to engage in morally hazardous behaviours (see chapter 7), lenders have to reduce interest rates to contain the moral hazard problem. But lower interest rates hurt the poor and middle class who receive lower returns on their savings. On the other hand, the poor and middle classes gain from having more capital from the redistribution, which might enable them to overcome borrowing constraints and share the entrepreneurial rents enjoyed by the upper class. The overall effect on the voting outcomes of the politically decisive middle class is therefore ambiguous, though greater wealth inequality amplifies the interest-rate effect and may ultimately result in less redistribution in equilibrium (Grüner and Schils, 2007). Inheritance tax (IHT) has been more widely studied, but chiefly from an empirical standpoint. Small business owners often complain that IHT can destroy family businesses, because heirs who lack funds have to liquidate their inherited business to pay the IHT due on it. For the same reason, IHT can act as a disincentive to grow an enterprise; and it can require entrepreneurs to engage in costly estate planning. The evidence certainly suggests that IHT has a negative impact on entrepreneurs, despite its relatively modest yield to the Treasury. For example, it has been found that small business owners’ employment growth rates are reduced by their expected IHT liabilities (Holtz-Eakin et al., 2001). The evidence also shows that many business owners who anticipate IHT lack the liquid assets to cover fully their future liabilities, despite purchasing more insurance than non-owners do (Holtz-Eakin, 1999; Holtz-Eakin et al., 2001). Consequently, one would expect sales of going concerns to be greater when IHT rates are higher. In accordance with this prediction, Brunetti’s (2006) analysis of San Francisco probate records combined with variations in state and federal tax liabilities reveals a significant positive impact of IHT on business sales.

438

17.2

Public policy

Labour and product market schemes

This section reviews the evidence about three different public policies designed to promote entrepreneurship. These are employment assistance schemes which encourage the unemployed to start their own businesses; procurement and affirmative action schemes which favour particular groups of business owners perceived to be at a competitive disadvantage; and publicly funded business advice and assistance schemes. 17.2.1

Employment assistance schemes

Governments which operate employment assistance schemes have several goals, including unemployment reduction, job creation and the fostering of enterprise. Governments might also anticipate other benefits, such as self-sufficiency of participants, economic development of high-unemployment areas, and the provision of opportunities for socially excluded groups (OECD, 1998). To date, the largest employment assistance schemes have operated in the UK, France, Spain, Germany and Denmark. According to Bruce and Schuetze (2004), ten states in the USA have ‘Self-Employment Assistance Programs’ (SEAs) intended to help unemployed workers create their own jobs by starting small businesses (see Fairlie, 2005b, for details about these and other US micro-business programmes). Most countries’ schemes share several features in common. Rather than attempt to describe them all, I will instead only sketch the UK’s Enterprise Allowance Scheme (EAS).11 The EAS, established in 1982, offered income support of £40 per week to unemployed people with £1,000 to invest in a new business for up to the whole of their first year in selfemployment. The idea was to partially compensate participants for the loss of state benefits entailed by becoming self-employed. Participants were offered advice on running a small business, but were not screened for eligibility to join the scheme. At its peak in 1987–88, 106,000 people were on the EAS, with take-up rates declining continuously thereafter. In most years, the number of EAS recipients comprised a small fraction of the annual unemployment flow, although they accounted for up to 30 per cent of new business starts (Bendick and Egan, 1987). Most EAS businesses were concentrated in services, construction and retail commerce. They typically involved limited amounts of investment, and operated on a small scale – possibly because of limited market opportunities in areas of high unemployment where many recipients were located. Perhaps unsurprisingly in the light of the findings reported in chapter 14, enterprises endowed with the most financial and human capital enjoyed the highest survival rates. The cost-effectiveness of EAS-type schemes is a key issue for its government sponsors. These schemes can be expected to involve a substantial deadweight cost, since some of the new businesses are likely to have started anyway in the absence of the scheme. An EAS may also support inefficient businesses and displace viable competitors. It has been suggested that one-off grants enabling new ventures to overcome entry barriers might reduce displacement effects relative to an EAS (Meager, 1994). Another problem is that EAS participants are a self-selected group of unemployed people, being

Taxation, regulation and other policies

439

disproportionately young, male and having been unemployed for less than six months prior to commencing in the EAS (Bendick and Egan, 1987). Hence an EAS might fail to help the neediest group, namely the long-term unemployed. The evidence reveals some of these concerns to be well founded. According to estimates published by Bendick and Egan (1987), 50 per cent of UK EAS-sponsored businesses would have started anyway; 50 per cent of those that did start displaced other businesses; about 50 per cent of the assisted firms survived for less than three years; and those that did start created only a fraction of one job in addition to the job of the proprietor. Only one-fifth of surviving firms created any new non-proprietor jobs, while more than 60 per cent of the jobs created in surviving firms after three years under the UK EAS were in 4 per cent of the businesses originally created (Bendick and Egan, 1987; Storey, 1994a). Overall, the evidence about the impact of EASs on business survival rates is rather mixed.12 Perhaps disappointingly, the business incomes of EAS recipients have been estimated as similar to the incomes they could have obtained in alternative occupations, although it is possible that the experience of entrepreneurship enhances participants’ future earnings (Bendick and Egan, 1987). Taking account of deadweight costs, displacement effects and the cost of administering the scheme, relative to the benefits generated by the new start-ups, Storey (1994a) estimated that after taking account of the fact that the government would have paid unemployment benefits anyway, the UK EAS was essentially cost-effective, even though its effects on job creation and unemployment reduction were slight. Similar conclusions emerge from an evaluation exercise of US EAS-type schemes compiled by the US Department of Labour (Blanchflower, 2004). This is where it helps to be clear about the objectives of EAS-type schemes. Schemes targeted at the unemployed face a trade-off between economic objectives such as high survival rates and employment creation, and social objectives such as putting to work the hardest to employ. Welfare recipients in particular lack (virtually by definition) human and financial resources and so are especially hard to move into sustainable businesses (Dennis, 1998). The social benefits of facilitating the creation of a few successful businesses by poor welfare recipients are likely to be high, as deadweight costs are low; but the economic costs are also high because this group is prone to the highest business closure rates. Therefore a well-targeted EAS is likely to create mainly economically marginal new businesses which have little impact on regional growth rates (MacDonald and Coffield, 1991; Greene et al., 2004). Thus it has been said of the British and French EASs that they ‘have succeeded in turning less than one per cent of transfer payment recipients into entrepreneurs, and an even smaller proportion into successful ones. They cannot be said to have contributed greatly to solving either social or economic problems, let alone both’ (Bendick and Egan, 1987, p. 540).13 Storey (1994a) concluded that the EAS probably had a greater political than economic impact. EAS participants in Germany appear to be self-selected from among the ablest of the unemployed. This is reflected in levels of human capital and rates of business survival

440

Public policy

and employment growth among scheme recipients which are not significantly different from those of non-recipients, as well as favourable performance outcomes.14 Overall, though, one is led to conclude that, despite favourable publicity, schemes designed to promote enterprise among the unemployed have had only a very limited economic impact in practice. 17.2.2

Procurement and affirmative action schemes Procurement is an important channel through which governments can provide targeted support to small businesses. The US federal government has systematically allocated a share of its purchases to small businesses since 1942. According to SBA (2001), small businesses received $69.3 billion in fiscal year 1999 in federal government contracts. This amounted to 35.6 per cent of total federal procurement. The construction sector received most procurement money. The role of the SBIR in this regard has already been discussed, in chapter 16. On top of this, state governments in the USA also provide large sums of procurement money to the small-business sector. To counter the possibility of consumer discrimination, which reduces demand for the goods and services produced by entrepreneurs from ethnic and disadvantaged groups (see chapter 5), several US state and federal administrations passed affirmative action (AA) schemes in the 1970s and 1980s intended to ‘level the playing field’. In particular, the US Public Works Employment Act of 1977 required state and local governments to reserve 10 per cent of federal funds for public works for ethnic minority businesses (Levie and Smallbone, 2007). There is evidence that these AA programmes increased African-American participation rates in some types of entrepreneurial activity, notably the construction industry (Blanchflower and Wainwright, 2005). However, AA policies have come under prolonged legal challenge and have been gradually dismantled over the last twenty years. AA policies are controversial because while they promote diversity, they can become vehicles of favouritism and may end up replacing unsubsidised high-quality majority-owned businesses with subsidised low-quality minority-owned businesses. Blanchflower and Wainwright (2005) observe that following the removal of AA policies, minority and female business ownership rates declined, sometimes dramatically. They did so less dramatically among white women, in part because some women-owned companies were run by spouses as ‘fronts’ to take advantage of the AA programmes. This provides a striking example of how the private sector can subvert well-intentioned public policy initiatives. 17.2.3

Business advice and assistance

Many national governments subsidise agencies which provide information and advice to entrepreneurs, for both new start-ups and established businesses. Private sector organisations also provide these services. The oft-cited rationale for public intervention in this area is that entrepreneurs lack information about what it is like to start a business, how to develop a business idea and obtain finance, how to identify customers, and how

Taxation, regulation and other policies

441

regulations and state benefits affect them. They might also be naïve, proposing poorly thought-out business plans that convey to lenders and other stakeholders an inadvertently negative impression of their business idea. Worse still, they may be unaware of deficiencies in their own business skills. If private sector assistance is not readily available, these problems can cause inefficiency and excessive exit rates of otherwise viable businesses. Public policies in this area usually provide directly, or subsidise the outsourced provision of: information; mentoring and training; and advisory services, via public-sector agencies and/or subsidised private-sector advisers and consultants. Many programmes combine assistance in packages for target groups. This comprised a major component of entrepreneurship policy under the Thatcher and Blair administrations in the UK, for example (Della-Giusta and King, 2006). In the USA, Small Business Development Centres provide free training, counselling and technical assistance services to small businesses (see Chrisman and Katrishen, 1994, for details). As yet, however, we know little about their effectiveness. In other countries, subsidised information and advice for smaller firms is both financed and delivered by the state, as under Sweden’s ALMI scheme, for example. An alternative delivery mode is via private sector consultants, as in the UK’s Enterprise Investment (EI) scheme. The EI subsidised only a maximum of fifteen days’ use of consultants’ time per firm, to provide a ‘one off’ demonstration of the benefits of external advice. Arguably, private provision might be preferable to public provision of these services by being more incentivised, and of higher quality and greater relevance, than programmes overseen by civil servants. A serious practical limitation of formal public-sector assistance programmes is that they tend to be much less commonly used than private-sector assistance programmes. The latter are in turn less commonly used than informal sources of advice such as business and social contacts, family members and close friends.15 For example, while nearly one-half of Australian SME owners accessed an external accountant on more than three occasions during 1995–96, with bankers and lawyers coming next, public sources of advice were much less commonly used (Watson, 2007). This might in part reflect the proven effectiveness of private sector sources of support, which are associated with significantly higher survival and growth rates. Although there appears to be no significant gender difference in the utilisation or impact of external business advice (Robson et al., 2008), take-up rates of public assistance programmes are often lower among members of ethnic minorities and immigrants, who rely instead on informal assistance and self-help.16 It is unclear what explains this, although ‘colour blindness’ in the design and delivery of these programmes might be partly at fault (Ram and Smallbone, 2003). Despite their low take-up rates, there is growing evidence that publicly funded assistance programmes succeed in helping ventures survive.17 The same is true of private-sector advice too (Bosma et al., 2004). This begs the question of why public intervention is needed when the market already provides effective private solutions. In an effort to answer this question, Wren and Storey (2002) evaluated the British EI scheme and found that it had a substantial impact on medium-sized (but not small)

442

Public policy

firms, increasing survival rates in the former by 4 per cent in the long run, and growth rates of surviving firms by up to 10 per cent per annum. But Storey (1994a) highlights several difficulties of trying to link assistance to improvements in performance by the recipients. In his words, ‘a substantial “unproven” verdict hangs over the policies in this area’ (1994a, p. 295). The Prince’s Trust is a novel British scheme which combines advice and mentoring with financial support. This scheme, initiated by HRH the Prince of Wales, targets disadvantaged young people who are 18–30 years old, especially members of ethnic minorities and individuals with criminal convictions. The typical start-up loan or grant offered under this scheme is £1,500–£3,500. Between 1983 and 1999, the Prince’s Trust disbursed almost £25 million in grants and £56 million in loans (OECD, 1998). The Prince’s Trust has been the focus of an independent evaluation study by Meager et al. (2003). These authors performed a Step V evaluation (in the terminology of chapter 15) by constructing a matched sample of 2,000 participants and 1,600 non-participants. They found little evidence that the Prince’s Trust affected participants’ employability or earnings, in contrast to other evaluations which used simpler methods and detected larger effects (e.g. Shutt and Sutherland, 2003; Storey, 2006). Nevertheless, Meager et al. (2003) detected some above-average persistence in self-employment propensities among Prince’s Trust programme participants.

17.3

Regulation

Even liberal developed economies are comprehensively regulated. One of the fastestgrowing areas of research within the economics of entrepreneurship is the theoretical and empirical analysis of different forms of regulation and legislation, and their impact on entrepreneurs. Because many regulations change only slowly, most empirical studies exploit cross-region and cross-country variations in regulation to identify its effects on entrepreneurship. Identification of impacts is possible because governments around the world choose different levels of regulation, reflecting different domestic political imperatives. The first type of regulation considered below relates to administrative requirements to start and register a business. A second type of regulation affects the ease with which entrepreneurs can exit. An important legal instrument in this respect is bankruptcy rules, which specify lenders’and entrepreneurs’entitlements to the assets of failed businesses. I then consider a related issue, namely laws which regulate corporate governance and credit markets, before turning to the effects on entrepreneurs of labour and product market regulations. After discussing these topics in turn, some conclusions are drawn. 17.3.1

The regulation of entry Entrepreneurs starting up a new firm invariably have to pay a fee to register their business and have to comply with several time-consuming administrative procedures. There are wide international variations in the costs of complying with entry regulations. For

Taxation, regulation and other policies

443

example, Djankov et al. (2002) observed that until fairly recently, entrepreneurs seeking to start a new firm in Italy had to follow no fewer than sixteen different procedures, pay nearly US$4,000 in fees, and wait at least sixty-two days to acquire the necessary permits. In contrast, in Canada there were only two procedures, costing US$280 and taking only two days. In a lively polemic, Dennis (1998) catalogues several examples of US entry regulations which impede company formations and early-stage growth. These include local zoning ordinances which designate home-based businesses illegal in some cities, or which restrict their operations. Other regulations place onerous ‘minimum training’ requirements on potential entrants or restrict the supply of licences needed to trade legally (e.g. for taxi-cabs). Dennis (1998) further observes that the costs of complying with these regulations fall heaviest on the poor, especially welfare recipients trying to start up new businesses. Regulations can also have perverse effects. German regulations designed to maintain the authenticity of Chinese restaurants by legally restricting who can work as chefs in these establishments have encouraged some restaurateurs to switch into fast food (Leung, 2003). This sector has lower average skill requirements and presumably even less ‘authenticity’, too. There are signs that policy-makers around the world are starting to simplify and even remove entry regulations. For example, between 1999 and 2005 the number of days needed to set up a business was reduced from 53 to 8 in France; from 82 to 47 in Spain; and from 62 (noted above) to 13 in Italy. One of the reasons for this policy change has been growing awareness about the damage that these policies can do to domestic entrepreneurship and the broader economy. That entry regulation is likely to impede entrepreneurship follows straightforwardly from the occupational choice model outlined in chapter 2.18 ‘Second-round’ effects on broader dimensions of economic performance are also likely to be negative, for at least two reasons. First, the number of advertised job vacancies declines when the number of entrepreneurs decreases. When unemployed workers face costs of searching for new jobs, the outcome is fewer successful matches between workers and employers, resulting in smaller inflows into employment from unemployment and hence higher levels of equilibrium unemployment. A second unfavourable outcome is that higher start-up costs not only hinder ‘mismatched’ workers (i.e. those unable to find a job) from starting their own new venture, but also weaken incentives for individuals to acquire the education needed to start and operate high-growth enterprises. Clearly, lower levels of education and fewer highgrowth firms carry adverse consequences for economic performance (Dulleck et al., 2006). And as well as enhancing efficiency via greater competition in the ways noted above, entry deregulation can counter the malign influence of powerful incumbents lobbying for entry regulations to protect themselves against competition. Deregulation reduces the returns incumbents receive from diverting effort into unproductive activities such as lobbying (see chapter 12). On the other hand, productive entrepreneurs can sometimes benefit from certain forms of entry regulation. For example, product certification can protect consumers

444

Public policy

from crooks and ‘cowboys’ who could otherwise set up firms to dupe their victims with scams or poor-quality workmanship. This benefit is especially valuable if quality is hard to measure, as in care homes, for example. By screening out potential frauds and cheats, entry regulations make consumers more willing to demand goods and services from legitimate entrepreneurs, thereby increasing aggregate demand. On a different tack, if entrepreneurs are over-optimistic (as the evidence in chapter 4 suggests), there can be too much entry for the social good. Hence regulations which make market entry costly could counteract this tendency. They might also increase the ex post welfare of over-optimists who are excluded from entrepreneurship (de Meza, 2002). The available evidence largely bears out theoretical predictions that entry regulations retard entrepreneurship. Within France, for example, Bertrand and Kramarz (2002) detected substantial negative effects on employment growth from legal changes requiring entrepreneurs to gain approval from regional zoning boards. And within the USA, there is evidence from the airline industry that deregulation increased the rate of firm formation (Kelly and Amburgey, 1991). Other evidence also suggests that entrepreneurs are sensitive to entry regulations. For example, following the removal of national barriers to cross-state listings within the EU in 2001, there was a six-fold increase (from 4,400 to over 28,000) in the annual number of new private Continental European firms listing in the UK, where incorporation costs are low (Becht et al., 2008). Many researchers who explore the entrepreneurship–regulation relationship empirically do so using cross-country data. These data usually measure the degree of entry regulation in terms of the number of administrative procedures for starting a new business, and/or the number of weeks to set one up. Cross-country correlations between entry regulations and new firm entry rates (or inflows from employment to self-employment, or employment growth) tend to be negative.19 An exception is Desai et al. (2003), who reported positive correlations. However, that study relies on data from the World Bank ‘Doing Business’ (WBDB) survey. These data cover only limited-liability companies employing between five and fifty employees one month after start-up. Because the vast majority of entrants have no employees initially, these data are unsuitable for analysing effects of policies on new venture creation, broadly defined. Likewise van Stel et al. (2007) detected no effects of WBDB start-up procedures on new business starts using GEM data from thirty-nine countries over 2000–05. Perhaps the weightiest contribution to the empirical literature on entry regulations and entrepreneurship is the important article by Klapper et al. (2006). A problem with using simple country variations in regulations is that the effects of the regulations are likely to depend on the industry structure of the country. After all, one would expect the negative impact of entry regulations on entry to be greatest in industries with relatively high rates of entry, since industries characterised by natural monopolies are unlikely to be much affected by further entry barriers. Furthermore, relying only on country variations in regulations runs the risk of conflating effects of entry regulations

Taxation, regulation and other policies

445

with other country-specific factors. Such factors include legal systems, conventions of transparency and ‘culture’ – as will be discussed further below. To address this issue, Klapper et al. (2006) include in their econometric model of entry an ‘interaction’ variable defined as the product of the ‘natural’ entry rate in each industry (proxied as those entry rates observed in that industry in the relatively barrierfree USA), and country-specific measures of entry regulation. One would expect this interaction term to be negatively associated with firm entry rates if entry regulations are important.20 Klapper et al. (2006) measure entry regulations in terms of both dollar entry costs and the number of entry procedures. They use the comprehensive Amadeus database on European firms over 1998–99. Their dependent variable is the entry rate of ‘new’ incorporated firms, defined as ventures no more than two years old. They report a significant negative coefficient on the interaction between natural industry entry rates and national measures of entry regulation, implying that costly entry regulations do indeed hamper the creation of new firms, in industries that should ‘naturally’ (i.e. in the USA) experience high rates of entry. Furthermore, these negative effects are estimated to be greatest in richer countries; in industries with larger industry shares of SMEs; and in countries which are not corrupt and where regulations are more likely to be enforced. Klapper et al. (2006) also report that entry regulations are associated with larger average sizes of new entrants and slower rates of growth among incumbents in high-entry industries. This could be because lower rates of entry reduce competition and therefore shield incumbents from pressure to improve productivity.21 Cross-country evidence suggests that entrepreneurs allocate less effort towards highgrowth activities in countries where there is more corruption (Bowen and de Clercq, 2008). On the other hand, building on a point made above, corruption can mitigate the negative impact of regulations on entrepreneurship by ‘greasing the wheels’ of business, since bribes can speed up the granting of permits to entrepreneurs, and even improve the quality of civil servants (Méon and Sekkat, 2005). This hypothesis can be tested by constructing an explanatory variable which interacts the regulatory burden with a measure of corruption. If the hypothesis is correct, entry rates should be negatively related to entry regulations and positively related to this interaction variable. Evidence consistent with this hypothesis emerges from GEM data for forty-three countries over 2003–05 (Dreher and Gassebner, 2007). Also consistent with this hypothesis, countries with more burdensome entry regulations tend to experience greater levels of corruption. Finally, it is worth emphasising that entry regulations not only impede entrepreneurship, but also fail to achieve their desired goal of enhancing the quality of goods. Nor are they associated with lower pollution levels or better health outcomes (Djankov et al., 2002). One is left wondering whether entry regulations originate from government efforts to nurture national champions or from efforts by incumbents to protect their markets from competition. Either way, regulations can gain historical legitimacy and foster special-interest groups determined to preserve them.

446

Public policy

17.3.2 The regulation of exit (bankruptcy) A different class of regulations governs the allocation of assets remaining after businesses fail. As noted in chapter 14, most business closures are not failures. But when ventures do fail, bankruptcy laws stipulate how remaining business assets are divided between entrepreneurs and external stakeholders such as suppliers, customers, lenders and the tax authorities. Anticipating the risk of future failure, bankruptcy rules can affect entrepreneurs’ and stakeholders’ incentives to engage with an entrepreneurial venture in the first place. Below, I first explain several salient aspects of bankruptcy law and then analyse the incentives they can create for entrepreneurs and related stakeholders. The discussion concludes with a look at the empirical evidence and some policy implications. An important historical argument in favour of leniency towards bankrupt entrepreneurs can be found in the well-known Local Loan v. Hunt case, where the US Supreme Court (in 1934) ruled that bankruptcy law ‘gives to the honest but unfortunate debtor…a new opportunity in life and a clear field for future effort, unhampered by the pressure and discouragement of pre-existing debt’. This is the legal basis for giving unsuccessful entrepreneurs in America a ‘fresh start’. In practice, gentler bankruptcy laws involve more generous asset exemption levels for bankrupts, and/or a shorter time for entrepreneurs to become discharged of their obligations. At the outset, one should recognise that bankruptcy provisions vary both between and within countries. Regarding cross-country variations, bankrupts enjoy no discharge at all in Austria, Sweden, Spain or Italy, compared with immediate discharge in the USA. Recently, the UK reduced the time to discharge from three years to just one year. Regarding within-country variations, consider the ‘Chapter 7’ provision of the US bankruptcy code, which discharges both business and personal debts. Under the provisions of Chapter 7, the entrepreneur relinquishes to creditors all of their assets in excess of an exemption level. All of their future earnings are exempt from any further obligation to repay debts. However, the entrepreneur’s credit record is marked by recourse to Chapter 7. From the perspective of the entrepreneur, corporate Chapter 11 is often a preferred route, as this can buy the entrepreneur time and possibly also liquidity, enabling them to renegotiate with banks and other creditors, and to avoid tax and other financial commitments while undergoing corporate restructuring. The problem with corporate Chapter 11 from an economic standpoint is that it can amount to a ‘subsidy for failure’, and may be especially inefficient if the subsidy encourages entrepreneurs to stick too long with businesses which are poor matches with their human capital (Baird and Morrison, 2005).22 Unusually in the economics of entrepreneurship, theory is virtually unanimous in predicting that gentler bankruptcy laws encourage entrepreneurship. One important reason relates to risk-taking. Akin to the effects of welfare benefits, discussed below, forgiving bankruptcy laws provide insurance and encourage risk-taking behaviour, enabling the economy to trade further along its risk–return locus (Fan and White, 2003; Lee et al., 2007). The insurance inherent in asset exemptions is especially important for over-optimistic entrepreneurs (see chapter 4) who sub-optimally choose excessively

Taxation, regulation and other policies

447

risky investment projects and so are more likely from an objective standpoint to need the insurance in the bad state, even though they did not expect to need it initially owing to their over-optimistic bias. Therefore gentler bankruptcy laws encourage marginal entrepreneurs to enter, and failed entrepreneurs able to exploit valuable business experience to re-enter, the market (Gladstone and Lane-Lee, 1995). An additional reason for favouring gentle bankruptcy laws is that they can mitigate problems arising from imperfect credit markets. Remaining with the case of over-optimistic entrepreneurs momentarily, banks can strengthen their defences against such borrowers in several ways, for example by stipulating stringent collateral requirements. But the forces of competition then induce banks to reduce interest rates to attract collateralised borrowers. Unfortunately, this further stimulates loan applications from over-optimistic entrepreneurs. It would therefore be socially efficient to somehow weaken bank defences, to reduce over-investment. A policy which achieves this is a laxer bankruptcy law, since that limits the amounts that banks retrieve from defaulters. By bearing more of the costs of default, banks must charge higher interest rates. This has the desired effect of discouraging over-optimists (Fan and White, 2003; Manove and Padilla, 1999). As noted in chapter 7, higher interest rates are a common solution to over-investment problems. It is also a neat policy response in the present context because problems caused by over-optimism are especially difficult to solve contractually because, by definition, optimists are unconscious of their own cognitive biases. Evidence cited below is consistent with the prediction that lax bankruptcy laws are indeed associated with higher costs of credit. Another credit market problem can arise from the game, described in chapter 7, in which able entrepreneurs signal their type by foregoing asset protection in return for a lower interest rate (Chamley, 1983). If less able types have to mimic able types and forgo asset protection as well in order to avoid revealing their unfavourable hidden borrower characteristics, then in equilibrium all entrepreneur types can end up bearing excessive costs ex post. Generous bankruptcy exemptions (e.g. restrictions preventing creditors from seizing specific assets, such as a person’s homestead) can therefore make both borrower types better off, while banks still break even (Aghion and Hermalin, 1990). Another possibility is that bankruptcy legislation which allows bankrupt entrepreneurs to make a fresh start, i.e. which does not encumber failed entrepreneurs with onerous liabilities, provides them with incentives to supply greater effort if they try again in business after failing. That is because they get to keep a greater fraction of the returns to their future effort. This can be expected to increase social welfare. On the other hand, while ex post effort is likely to increase, gentler bankruptcy provisions could dilute ex ante incentives to supply effort while operating a venture, since entrepreneurs anticipate a weaker penalty for failure (Ayotte, 2007). Several of the theoretical predictions above receive support from the data. More liberal bankruptcy rules have been found to promote entrepreneurship (Fan and White, 2003; Klapper et al., 2006) and to increase the amount of venture capital and number of VC financing rounds (Armour and Cumming, 2006; Hasan and Wang, 2008). For example, Fan and White (2003) used SIPP (Survey of Income and Program Participation)

448

Public policy

data over the period 1993–98 to exploit state-level variations in bankruptcy exemption levels within the USA. These authors report evidence of a strong positive relationship between entrepreneurship and asset exemptions, estimating the probability that a household owns a business is 35 per cent higher if they live in a state with unlimited rather than limited exemptions. On the other hand, more liberal bankruptcy rules also tend to be associated with higher interest rates, smaller loan sizes and restricted access to credit.23 Theory and evidence therefore agree that bankruptcy laws should err on the side of generosity towards entrepreneurs. However, there are both theoretical and practical caveats that argue against taking this policy recommendation too far. On the theoretical front, more liberal bankruptcy rules can reduce the costs of closing existing businesses and can thereby promote serial entrepreneurship (Landier, 2004). While this policy enhances efficiency in economies dominated by high-tech industries where high-value investment projects are plentiful, it does the opposite in economies in which highvalue entrepreneurial opportunities are scarce and where serial entrepreneurship wastes resources. An example of wasteful entrepreneurship is ongoing re-entry by persistently over-optimistic entrepreneurs who lack ability and fail repeatedly. Practical caveats are more prosaic, but no less important. For instance, since 2000, successive Enterprise Acts have eased bankruptcy rules in the UK. But consumers have exploited these changes to avoid repaying personal debts. According to an editorial in The Times of London, 3 January 2007, following the UK Enterprise Act in 2004, which reduced the time to discharge from three years to one year, there was a dramatic increase in the number of personal insolvencies. In 2006, the number of insolvencies soared to 100,000; yet only one-quarter of these had any connection with self-employment. Unscrupulous operators forming temporary enterprises designed to cheat the public also benefit from laxer bankruptcy laws. It is notable that in 2005, the US Bankruptcy Reform Act restricted access to fresh starts, the opposite direction to the UK move. 17.3.3

Property rights and corporate governance

Well-protected property rights promote entrepreneurship and innovation. Secure property rights prevent ad hoc expropriation of the fruits of entrepreneurship, which is necessary to make costly investments worthwhile. A stable legal framework promotes planning, resource acquisition and co-ordination that take time to put in place. It also reduces entrepreneurs’ idiosyncratic risk, inducing entrepreneurs to hire more capital and labour and build larger and more efficient entrepreneurial ventures. Chapter 12 cited some evidence from the transition economies of central and Eastern Europe showing that weak enforcement of property rights impedes entrepreneurship. There is also evidence that the risk of appropriation, measured in terms of the degree to which governments and institutions fail to protect economic freedoms, causes underpricing of IPOs listed in foreign countries (Bell et al., 2008).24 Regional and cross-country evidence bears out these concerns for developed economies too. In developed countries, stronger protection of property rights and the

Taxation, regulation and other policies

449

efficiency of courts are both associated with higher business entry rates, lower exit rates, smaller average firm sizes, older vintages of firms, superior venture growth performance and efficiency.25 For example, Jappelli et al. (2005) argue that improvements in judicial efficiency in enforcing debt contracts when borrowers can default opportunistically are likely to reduce credit constraints and increase lending. By increasing borrowers’ incentives to repay, creditors are more likely to extend credit ex ante. Conversely, if contracts are hard to enforce and banks deny credit to entrepreneurs wishing to invest in productivity-enhancing technologies, output and wages remain low. This can make operation of inefficient technologies optimal for entrepreneurs, who extract rents from the low wages and high prices that result. Hence low TFP is likely to be observed in economies with institutions supporting only modest levels of contract enforcement (Erosa and Cabrillana, 2008).26 On the other hand, Zazzaro (2005) points out that stricter enforcement of debt contracts under conditions of asymmetric information might actually worsen credit allocation to entrepreneurs if it weakens sufficiently banks’ incentives to screen borrowers. Instead, improvements in accounting standards (e.g. tightening accounting and reporting rules) can increase the amount of screening and hence improve credit allocation and social welfare. Reflecting the different theoretical positions on this issue, the available evidence is mixed. Using panel data from ninety-five Italian provinces, Jappelli et al. (2005) observe that significantly fewer business loans are extended in provinces with longer trials or large backlogs of pending trials. Judicial efficiency correlates positively with the volume of lending and negatively with proxies for credit constraints. Using a different sample of Italian data, however, Guiso et al. (2004) report that a measure of judicial inefficiency, namely the number of years it takes to reach a first-degree judgement in the province, has a negative but insignificant effect on Italian firm creation rates. Perhaps the way judicial efficiency is measured affects empirical estimates of its impact on entrepreneurship. It has been suggested that some cultures are characterised by low levels of corruption and judicial inefficiency. There is some evidence, for instance, that ‘AngloSaxon’ economies possessing an English common-law tradition protect investors and entrepreneurs better than economies with Continental legal traditions, such as the French civil code (Fogel et al., 2006). But, as noted in chapter 12, entrepreneurship in all countries can come under threat from rent-seeking by established incumbents. Political lobbying to restrict competition by regulation transcends cultures and legal traditions. The antidote is transparent governance, a system of checks and balances, and conventions of open reporting. Intellectual property rights (IPR) are especially important for entrepreneurs with high-technology ventures. IPR is a double-edged sword in the context of entrepreneurship. On the one hand, laws protecting IPR might appear to be anti-entrepreneurial, by restricting potential new entrants’ access to information embodied in current innovations which could generate new market opportunities. On the other hand, by protecting the fruits of entrepreneurs’ effort, IPR creates incentives for entrepreneurs to innovate

450

Public policy

ex ante (Claessens and Laeven, 2003; Mayer-Schönberger, 2003). Evidence from MIT shows that the more effective patent protection is, the greater the likelihood that new firms will exploit a patent (Shane, 2001). This is an example of a positive side to legal regulation. Policy-makers can strengthen property rights in several ways. Quite apart from providing and maintaining a stable, fair and transparent legal framework, suitable corporate governance and corporate reporting legislation can improve the flow of information to investors, helping them to avoid being expropriated. However, laws to this end, which are usually directed towards large companies, can also have serious and unintended negative effects on smaller entrepreneurial ventures. It has been argued, for example, that despite enhancing accountability and confidence among investors – which might be expected to stimulate their interest in entrepreneurial companies – the US Sarbanes– Oxley (SOX) Act may have imposed high costs on smaller companies. These include compliance costs which bear disproportionately on the smallest firms (see below); rules restricting managers’ freedom of manoeuvre, making management more complex and diverting entrepreneurs’ scarce attention away from building their businesses; and rules restricting the use of share options to attract scarce managerial talent. The latter can hit newer firms especially hard, since they are often cash-constrained and so have to rely on this incentive to motivate and reward employees (Caselli and Gennaioli, 2005; Eisenman et al., 2007). SOX therefore provides a good illustration of the pitfalls involved in framing policy in this area. 17.3.4

The regulation of employment Employment protection legislation (EPL) comprises various measures designed to protect the rights of employees at work. As defined by the OECD (1999, p. 51), employment protection refers to regulations about hiring (e.g. rules favouring disadvantaged groups, conditions for using temporary or fixed-term contracts, and training requirements) and firing (e.g. redundancy procedures, mandated pre-notification periods and severance payments, special requirements for collective dismissals, and short-time work schemes). EPL in turn refers to all types of employment protection measures, whether grounded primarily in legislation, court rulings, collectively bargained conditions of employment or customary practice. It is commonly argued that EPL retards entrepreneurship by disproportionately imposing burdens on the smallest firms which hire (or seek to hire) workers. These burdens are likely to be especially onerous in countries where EPL is implemented heavily, such as Sweden and Spain (OECD, 1998). Entrepreneurs running the smallest firms can least afford the hiring and firing costs imposed by EPL. Furthermore, the damage caused by having to retain poor performers in small firms where ‘every job counts’ is likely to be greater in small than in large firms. One might therefore expect EPL to discourage individuals from becoming entrepreneurs at all, and to decrease the survival prospects of entrepreneurs who employ outside workers. And as with government regulation more generally, there can also be disincentives for the growth of enterprises if small-firm exemptions are withdrawn once a given company size is attained.

Taxation, regulation and other policies

451

Some EPL protects workers with the longest tenure, as in Sweden for example (Henrekson, 2005). This creates a further disincentive for workers to try entrepreneurship, since those workers ‘lose their place in the queue’ if they ever want to close their venture and return to paid employment. EPL can also reduce worker mobility between firms. Yet worker mobility appears to play a crucial role in the success of entrepreneurial hotbeds like Silicon Valley, where it facilitates the rapid spread of technological information, knowledge spillovers and entrepreneurship (Saxenian, 1994; Davidsson and Henrekson, 2002). These arguments suggest a negative relationship between EPL and entrepreneurship. However, this relationship could be weakened if employers can circumvent EPL, for example by outsourcing work to self-employed contractors (Parker, 2007b). One might therefore expect a U-shaped relationship between EPL and entrepreneurship, as at low levels of EPL individuals have fewer incentives to be employees (since they enjoy less protection) while at high levels of EPL firms have greater incentives to replace employees with independent contractors (Müller and Arum, 2004; Parker, 2007b). The available evidence about whether EPL retards entrepreneurship is mixed. On the one hand, Kanniainen and Vesala (2005) ran regressions using pooled quinquennial data from several OECD countries over the period 1978–98, and reported a significant negative relationship between self-employment rates and four different measures of EPL. In addition, Klapper et al. (2006) found that labour regulations have a dampening effect on entry in labour-intensive industries, while van Stel et al. (2007) reported that labour market regulations which restrict hiring and firing depress entrepreneurial activity. But other multivariate analyses based on cross-country data have found insignificant effects of EPL on rates of self-employment and market entry.27 It is unclear why evidence on this issue is mixed; perhaps less aggregated data are needed to dig deeper into the underlying causal mechanisms. A different but related policy is the minimum wage. It is possible that wage rigidities caused by minimum wages ration workers out of paid employment, thereby increasing the numbers entering entrepreneurship. On the other hand, a binding minimum wage increases labour costs to entrepreneurs who hire workers. This may hit small firms disproportionately hard because they tend to be more labour-intensive than large firms. In a time-series study based on aggregate US data over 1948–82, Blau (1987) found little evidence that the minimum wage had any net effect on the aggregate US selfemployment rate. Clearly, however, there is a serious paucity of research on this issue, and much more work remains to be done here. 17.3.5

The regulation of credit markets Public policies relating to venture capital and entrepreneurship were treated in the previous chapter. The brief discussion below focuses instead on bank credit. As noted earlier in this book, bank lending is the most important source of external finance for the majority of entrepreneurial ventures. Regulations which restrict competition among small business lenders may therefore impede access to finance for entrepreneurial ventures, and so decrease the equilibrium amount of entrepreneurship.

452

Public policy

When credit markets are characterised by imperfect competition, informationsharing between banks weakens inter-bank competition for entrepreneur borrowers. That makes entrepreneurship costlier for able aspiring entrepreneurs who lack a track record. By injecting greater competition into the credit market, bank deregulation can overcome this problem and promote entrepreneurship (Gehrig and Stenbacka, 2007). On the other hand, the absence of regulations which protect investors can also constitute a barrier to new firm creation. In countries with strong creditor protection, it is relatively easy for unlisted firms investing in intangible (and hence non-collateralisable) assets to obtain loans. At the same time, institutional investors appear to be more willing to participate in private equity funds when the private equity market is regulated.28 To date, most empirical studies have reported a positive association between credit market deregulation and entrepreneurship.According to one study which utilised a panel of data on manufacturing establishments across US states between 1977 and 1994, US localities with less concentrated credit markets and looser regulations relating to bank expansion were characterised by greater numbers of small ventures and smaller average firm sizes (Cetorelli and Strahan, 2006). Consistent with these findings, cross-country evidence shows that greater bank concentration ratios are associated with more selfassessed obstacles to finance by owners of small and large firms (especially the former) in countries with low levels of economic and institutional development. More legal restrictions, government interference in credit markets and state ownership of banks strengthen the association between bank concentration and obstacles to finance in less developed countries (Beck et al., 2004). Greater bank competition appears not to be an unmixed blessing, however. Using data from 22 industries across 103 Italian provinces, Bonaccorsi di Patti and Dell’Ariccia (2004) investigate the role of bank competition on firm creation, and find a bell-shaped relationship. In contrast to the conventional view that bank competition always promotes access to credit, these authors argue that credit competition can sometimes worsen asymmetric information and hold-up problems, leading to restricted credit. And there is disagreement about the effects of deregulation on the volume of bank lending. While Black and Strahan (2004) claim that deregulation of American bank branching laws and interstate banking laws has significantly increased the growth of business lending and new business incorporations, a re-examination of these data reveals that deregulation was associated with increases in new firm incorporations in some regions but with decreases in others (Wall, 2004). This suggests a more nuanced relationship between bank competition and entrepreneurship. Other evidence shows that volumes of small business lending decrease following bank mergers. New and incumbent lenders pick up some, but not all, of the slack, and cannot compensate for the higher search costs and costs of credit caused by severance of long-standing banking relationships which can follow mergers and acquisitions (Berger and Udell, 2003). So the financial authorities need to be vigilant about ongoing consolidation in the small business loan market. Finally, a different linkage between financial regulation and start-ups relates to the housing market. Deregulation of the mortgage market which increases the number of

Taxation, regulation and other policies

453

owner-occupiers, as happened in Britain in the 1980s for example, can be expected to increase the availability of small business credit, since housing is often used as collateral by aspiring entrepreneurs (Henley, 2005).

17.3.6 Costs of compliance with regulations The number and scope of government regulations have increased steadily since the 1970s. Regulation is widely believed to put small firms at a disadvantage relative to their larger counterparts by imposing fixed compliance costs which the latter are able to spread over greater output. It is straightforward to modify the now familiar occupational choice model of chapter 2 to show that the imposition of a regulatory fixed cost causes the smallest firms to exit (see note 18 of the present chapter). Large firms benefit from the lower wages and higher prices which ensue, even making it possible for them to gain on net from the regulation if these benefits outweigh the higher regulatory costs. Thus large firms can have an incentive to apply political pressure for greater regulation, which can end up reducing entrepreneurship (see also Baumol, 1983). From an exhaustive analysis of the US evidence available at the time, Brock and Evans (1986) concluded that economies of scale do indeed exist for compliance with some forms of regulation. This is especially true of paperwork-based regulations, e.g. those relating to banking and pensions. Brock and Evans (1986) estimated compliance costs to be some ten times greater for small firms compared with their larger counterparts. Gurley-Calvez and Bruce (2008) estimated tax compliance costs per employee in small firms to be 1.8 times greater than those in large firms in the service industry and 4.5 times greater in manufacturing. Regulations can also harm small firms simply because they are more likely to engage in the proscribed activity (e.g. small mines involve more accidents than large mines). One policy which could limit the damage caused by regulation compliance costs incurred by small firms is the partial or complete exemption of small firms from regulation. This practice is known as ‘tiering’. It involves granting lighter regulatory burdens, or complete exemption, to small firms, in recognition of regulations’potentially adverse effects on competition, as discussed above. As Brock and Evans (1986, p. 74) put it: ‘tiering may occur in any of the major aspects of a regulatory program: in the substantive requirements; in reporting and record keeping requirements; or in enforcement and monitoring efforts. Tiering may entail less frequent inspections, lighter fines for non-compliance, exemptions, waivers, reduced requirements, or simpler reporting requirements for certain types of firms.’ Yet there are potential pitfalls with tiering regulations as well. Tiering may preserve inefficient firms in the market, and encourage larger, more efficient, firms to become smaller in order to qualify for a lighter regulatory touch. Brock and Evans (1986) concluded that there is little evidence that regulation has disproportionately harmed small American firms. Of course, this could simply reflect the success of tiering in protecting small firms from the worst excesses of government interference. In recognition of the regulatory damage that governments can inadvertently

454

Public policy

inflict on small businesses, the US legislature passed the Regulatory Flexibility Act. This instrument forces US government agencies to undertake a thorough analysis of the economic impact of their proposed regulations, and to consider alternatives that are less likely to harm small firms (SBA, 2001).

17.3.7

Conclusions Overall, despite ongoing disagreements among researchers, both theory and evidence point to somewhat stronger negative impacts on entrepreneurship and growth from regulation than positive or neutral effects. Yet governments the world over continue to tighten their regulatory grip on businesses while simultaneously espousing the benefits of entrepreneurship. It is unclear why this is so, though I will venture two conjectures. First, regulations such as clean air legislation are widely felt to be socially desirable even if they increase entrepreneurs’ production and operational costs. And it is even possible that these regulations harm large more than small firms, if the latter are better placed to respond to the demand for cleaner technologies by exploiting new opportunities created by environmental regulations (Dean and McMullen, 2007). In the case of the regulation of financial institutions, the benefits of prudent regulatory oversight which can avoid periods of credit crunch and financial crisis are even more obvious. However, most regulations promise less dramatic social benefits, while creating few obvious entrepreneurial opportunities. In fact, we still lack clear assessments of the impact of most regulations on the formation and early growth of businesses (Dennis, 1998). This evidence gap allows policy-makers enough ‘wriggle room’ to deny that their regulatory activities are retarding entrepreneurship at all. Second, there are sometimes powerful entrenched interests at work, such as incumbent firms, which have nothing to gain and much to lose from deregulation that opens them up to competition from vigorous new entrants. Tacit support by business people for further regulations can give a spurious legitimacy to claims that proposed regulations are ‘entrepreneur-neutral’ or even ‘entrepreneur-friendly’. It is important to remember, on the other hand, that entrepreneurs can sometimes be agents of institutional change. Entrepreneurs can help to deregulate centralised economies by becoming public advocates of change; by agitating ‘behind the scenes’ to persuade policy-makers to deregulate; and by evading restrictive laws to demonstrate to policy-makers the value of changing them (Acs, 1999; Li, Feng and Jiang, 2006). Such individuals are called ‘institutional entrepreneurs’. Some argue that institutional entrepreneurs play a particularly prominent role in China and India, where entrepreneurship cannot be disentangled from social and political constraints (Khanna, 2007). Despite the benefits to the economy as a whole that institutional entrepreneurship can bring, it can be privately risky, as the institutional entrepreneur not only has to cope with his or her customary business risk, but also bears the additional ‘political’ risk which arises if his or her reform efforts are unsuccessful. Li, Feng and Jiang (2006) provide several examples of institutional entrepreneurs who stimulated economically valuable reforms in China. Li, Meng and Zhang (2006) go on to find that entrepreneurs

Taxation, regulation and other policies

455

are more likely to become involved in institutional entrepreneurship in markets which are less developed, where government regulation is heavier, the informal tax burden is higher, and the rule of law is weaker. In most countries, though, individual entrepreneurs and even lobby groups are frequently unable to inhibit government regulation. Government spokespeople occasionally defend their actions by pointing out that regulations can open up opportunities for entrepreneurs to provide goods and services that make compliance with them easier, for example consultancy and legal services. This can happen at the same time that the regulations close off opportunities for other entrepreneurs on the grounds of cost or feasibility. A balanced assessment of regulations should therefore identify the winners as well as the losers. Overall though, economists usually argue that regulations which increase business costs for some companies generate net welfare losses even after factoring in the transfers from losers to winners. This recommends the general principle of minimising regulation wherever possible. Of course it should be recognised that government regulation can affect firms in a variety of ways, including ways which have scarcely been alluded to above. For example, Crafts (2006) argues that the main impact of regulation is felt less in terms of compliance costs than in terms of affecting incentives to invest and innovate. These impacts are difficult for researchers to detect and measure. As a result, we still lack hard evidence about their magnitude in the context of entrepreneurship. We also lack evidence about the effects of regulatory uncertainty on entrepreneurial firms, although there is some evidence that regulatory uncertainty leads smaller firms to select less risky strategies like invention (Birnbaum, 1984). Evidently much more research is needed on these issues. What does seem clear, however, is that the health of entrepreneurship and the business climate in general depends on having a transparent and fair legal justice system. If we are to have regulation it needs wherever possible to be simple, easy to understand and delineated in scope. Unfortunately, regulations appear to be becoming ever more complex and wide-ranging, a trend which may be set to accelerate following the 2008 financial crisis. An important challenge for policy-makers in the coming years is to regulate more sparingly, transparently, simply and appropriately, generating desired social benefits while minimising the burden on the businesses that generate the income and tax revenue on which governments depend. 17.4 The welfare state, trade unions, enterprise culture and instability

This section completes the discussion of specific public policies by examining relationships between entrepreneurship and several interlinked political, economic and cultural factors. 17.4.1 The size of the welfare state Although a large public sector can conceivably foster entrepreneurship, for example by improving infrastructure which can make enterprises more efficient, most

456

Public policy

commentators have focused on its likely negative impact on entrepreneurs. One challenge is that a large welfare state requires high taxes to finance it, which can crowd out private savings needed to leverage start-up finance and to exploit new business opportunities (Fölster, 2002; Henrekson, 2005). A second problem is that public-sector employment can push up wages, deterring entrepreneurs from hiring labour. Indeed, the attractiveness of public-sector jobs may have discouraged entrepreneurship among some African-Americans (Boyd, 1990; 1991). Third, public provision of goods and services can crowd out private-sector (entrepreneurial) provision of services. Davidsson and Henrekson (2002) and Henrekson (2005) argue that this problem is especially acute in high-tax Sweden. Furthermore, government fiscal expenditures can scarcely track changes in consumers’ tastes. So by crowding out the private sector, expansionary public expenditure can weaken entrepreneurs’ socially valuable role of predicting and adjusting to these changes. The outcome is lower average living standards (Takii, 2008). The other side of the argument, however, is that when the public sector does a good job of providing high-quality, centrally co-ordinated public services (e.g. health care) there is less need for private-sector entrepreneurship (Galbraith, 2006). Henrekson (2006) argues that a large welfare state may coexist with a low level of entrepreneurship as an endogenous outcome. Public provision of goods and services crowds out entrepreneurs from product markets, while high taxes needed to finance it crowd out savings and hence potential start-up finance. Large firms and trade unions then collude to perpetuate a large welfare state, since both parties have incentives to discourage competition from low-wage independent entrepreneurial ventures. Henrekson (2006) refers to this arrangement as the ‘Swedish model’. It corresponds to what has been called the ‘managed economy’, as opposed to the new ‘entrepreneurial economy’ which (it is claimed) has emerged in the United States (Audretsch and Thurik, 2001b). Certainly the lack of entrepreneurial dynamism in Sweden is quite striking. While onethird of the top 500 US firms is replaced within five years, the number of new top fifty private companies formed in Sweden over the entire 1970–2000 period turns out to have been precisely zero (Henrekson, 2005). In short, the Swedish economy appears to be characterised by ‘close to a complete void of spectacular high growth firms’ (Davidsson and Henrekson, 2002, p. 90). A growing body of cross-country evidence bears out these concerns. The size of the welfare state has been measured in several ways, including: the proportion of employees in the public sector; public expenditure as a percentage of GDP; absolute and relative levels of welfare expenditure; and state debt. To the best of my knowledge, the available cross-country evidence consistently identifies significant negative effects of these measures on aggregate cross-country rates of entrepreneurship.29 Similar findings have been observed using micro-data: public-sector workers are significantly and substantially less likely to transition to entrepreneurship than private-sector employees.30 On the other hand, US states which have larger highway expenditure budgets exhibit lower business failure rates, all else equal (Chen and Williams, 1999). This finding might reflect a positive facilitating role from basic public infrastructure, which of course is an altogether different category of public expenditure from welfare benefits.

Taxation, regulation and other policies

457

17.4.2 Social security transfers and trade unions The replacement rate is defined as the ratio of average out-of-work benefits (including unemployment benefits) to average earnings. A higher replacement ratio will decrease entrepreneurship if benefits discourage unemployed workers from moving off benefits – or if employees place an option value on these benefits which are unavailable to, or restricted for, entrepreneurs. Some evidence bears out these concerns. Using Spanish micro-data, Carrasco (1999) reported that unemployment benefits significantly and substantially decrease the probability of transitions from unemployment to self-employment, with a smaller impact on transitions from unemployment to paid employment. Studies based on multivariate analyses of aggregate data also detect a significant negative linkage between replacement rates and rates of entrepreneurship.31 A drawback with aggregate data though is that replacement rates might be correlated with other, omitted, factors which also retard entrepreneurship (e.g. a high social stigma of failure or institutional constraints on entrepreneurship) and which therefore exaggerate negative estimates of the effects of the replacement rate on entrepreneurship. Retirement benefits on the other hand might increase the attractiveness to employees of switching to part-time entrepreneurial activity as a form of partial retirement, especially if social security earnings tests penalise full-time paid work. Blau (1987) and Robson (1998b) found some evidence to support the notion that higher state retirement benefits promote self-employment. However, a welfare-state safety net might discourage individuals from providing the necessary effort required in entrepreneurship, which is a type of moral hazard (Sinn, 1995). Trade unions can magnify the adverse effects on entrepreneurship of social security benefits. The reason is that these benefits provide a floor on wages which unions can use to bargain for higher wages, thereby reducing profits of entrepreneurs who hire labour.32 On the other hand, trade unions can also create unemployment, which is a well-known point of entry into entrepreneurship (see chapter 4). The evidence broadly supports the notion that trade unions hamper entrepreneurship. Studies of this kind utilise several measures of trade union impact, including: union density; the number of workers covered by unions; bargaining coverage; and centralisation of wage bargaining.33 Also, Finnish micro-data show that trade union members are significantly less likely than non-union employees to transition into entrepreneurship (Hyytinen and Ilmakunnas, 2007). To the extent that trade unions succeed in extracting wage premiums from employers, this finding is precisely what one would expect.

17.4.3

Fostering an enterprise culture Some commentators have claimed that reforms to the UK labour market and welfare system in the 1980s created an ‘entrepreneurial culture’ in which self-employment and entrepreneurship were able to flourish. More generally, it has been argued, the term ‘enterprise culture’ is ‘associated with an ideological programme of reform intended to promote entrepreneurial values among citizens, businesses and in the welfare state’ (Della-Giusta and King, 2006, p. 630). In the 1980s, the British government not only

458

Public policy

deregulated markets, restricted the power of trade unions and introduced means-tested welfare benefits, but also initiated loan guarantee and enterprise allowance schemes; small business advice centres, grants and training programmes; tax reductions; and higher VAT (sales tax) registration thresholds – all in a concerted attempt to encourage enterprise. At the same time, the UK witnessed a dramatic increase in the selfemployment rate (see chapter 1), increases in business registrations and deregistrations, and strong growth of unlisted securities markets and venture capital activity (Bannock and Peacock, 1989). More detailed and careful analysis tends to rebut the view that the 1980s witnessed a renaissance in the British entrepreneurial spirit. According to the British Social Attitudes Survey, the proportion of employee respondents thinking about becoming self-employed did not change between 1983 and 1989 – a time when the renaissance of an enterprise culture was presumed to have occurred (Blanchflower and Oswald, 1990). Instead, Blanchflower and Oswald claimed that a combination of changing personal characteristics and favourable developments in the macroeconomic environment were able to explain the rise in the UK self-employment rate, without any need to rely on (unmeasured) factors like ‘entrepreneurial spirit’. Blanchflower and Freeman (1994) were also sceptical about an ‘enterprise culture’ effect, noting that although transitions from unemployment and non-labour force status to self-employment increased during the 1980s, transition rates from paid employment to self-employment did not. Blanchflower and Freeman declared that ‘it is hard to believe claims that an ‘enterprise culture’ has been established without some significant increase in this flow’. It is true that other international evidence suggests that regions and countries possessing cultural values which favour entrepreneurship enjoy higher rates of entrepreneurship (e.g. Swanson and Webster, 1992). But by omitting individual-level variables which are likely to be correlated with choices to become an entrepreneur, such studies are prone to overestimate the role of cultural values. Giannetti and Simonov (2004) deal with this problem in a study using Swedish municipal data from 1996–2000. Giannetti and Simonov (2004) instrument local rates of entrepreneurship with two variables thought to capture social norms: the proportion of pensioners who attend church and the extent of voting for right-wing (pro-business) political parties in the recent past.34 Giannetti and Simonov find that (instrumented) local rates of entrepreneurship significantly increase individual propensities to become self-employed while decreasing the incomes of the newly self-employed. They claim this supports the notion that social norms help drive entrepreneurial entry.

17.4.4

Political and economic instability

Episodes of political change and instability can be associated with higher rates of firm formation, though according to a century of evidence relating to the newspaper industry in the San Francisco Bay area, ventures created in these times tend to have lower survival rates than firms formed in more stable times (Carroll and Huo, 1986). Most economists believe that erratic macroeconomic policies discourage entrepreneurship by increasing

Taxation, regulation and other policies

459

risk and making it harder to discern whether transaction partners are behaving honestly or not. These problems are especially pronounced in transition economies (McMillan and Woodruff, 2002). A particular form of economic instability is inflation. Macroeconomic panel-data regressions from US states over 1986–95 reveal significant negative correlations between state inflation rates and state small-business employment shares (Robbins et al., 2000). This heightens the importance of monetary authorities using the interest rate as an anti-inflation instrument (see chapter 4). 17.5

Conclusion

The present and previous two chapters have critically evaluated the current state of theoretical and empirical knowledge about the impact of public policies towards entrepreneurship. It is no easy matter to draw conclusions from such a diverse body of work. But a few general principles and findings stand out. To start with, governments in several countries expend substantial resources on loan guarantee programmes and schemes to encourage the unemployed to become self-employed. Evaluation studies have not produced a ringing endorsement of these schemes. While they do not appear to misallocate resources badly, there is little evidence that they improve efficiency much either. These programmes have been rather limited in scope relative to the size of the potential markets they could serve. If governments are to be persuaded to enlarge the scope of these schemes, researchers must provide more convincing evidence that the schemes’ recipients generate substantial social and economic benefits which more than offset the costs. In fact, it is quite understandable that government policies towards entrepreneurship err on the side of modesty rather than extravagance. Governments invariably face conflicting aspirations and objectives. They want to target resources to achieve focus but are unable to pick winners; they want to make assistance selective to control budgetary costs but wish also to both remain inclusive and avoid spreading resources too thinly; and they want policies to make a big impact for political reasons while minimising costs and programme deadweight losses. These trade-offs are deep-rooted and probably inescapable. With respect to taxation, the rich theoretical literature presents an array of conflicting predictions and policy recommendations. Even the class of models that restricts attention to uniform linear income tax systems are unable to answer unambiguously such basic questions as ‘Will higher income tax rates increase or decrease the equilibrium number of entrepreneurs?’ and ‘What are the effects of different tax systems on entrepreneurs and on social welfare?’ Naturally, generalising these simple models to take account of other factors, including tax evasion, differential income taxation, non-linear (progressive) taxation, non-pecuniary aspects of work and social inequality-aversion can be expected to create even more conflicting recommendations. Tax evasion certainly seems to be a salient aspect of entrepreneurial behaviour. As a group, entrepreneurs systematically under-report their incomes and over-claim

460

Public policy

their business expenses to the tax authorities. However, it has not yet been convincingly established that pecuniary factors in general make individuals likelier to become entrepreneurs, let alone tax evasion opportunities in particular. An area where governments might be able to fine-tune their policies towards entrepreneurship is in the provision and dissemination of business information and advice. There might be a case, at least in principle, for governments to subsidise information flows to help the market function more efficiently. Economies of scale can be achieved in this area by concentrating provision in a few well-publicised advisory bodies. In fact, the problem in some countries seems to be not the absence of these agencies, but their proliferation and the sparse use made of them by entrepreneurs. Reynolds and White (1997) concluded that government policies and programmes have had very little direct effect on the start-up process because few ‘nascent’ entrepreneurs are aware of the large number of available programmes. This is reflected in take-up rates of advisory and assistance programmes of less than 10 per cent. Here too, governments need to spend public money wisely. Assistance directed towards enhancing general competitiveness and growth is likely to be more socially productive than trying to give targeted firms an advantage over their close competitors. That might accelerate the competitive race without increasing the number of winners. There is little point after all in helping some participants to play more effectively a zero-sum game. More generally, though, and at the risk of concluding this book on an overly normative note, there appear to be several major problems with the way that governments around the world currently conduct policy with regard to entrepreneurship. What follows are personal views based on a distillation of the literature examined in this book. First, it has not yet been clearly established that governments should encourage more entrepreneurship than is forthcoming from unfettered markets. The literature gives us no guidance about the optimal rate of business entry or entrepreneurship, concepts which might in any case elude satisfactory measurement. At any rate, though, there seems little point in encouraging entry for the sake of it since most new firms close before long anyway. Growth and survival are more economically valuable outcomes than entry per se. One is led to suspect that government involvement is often motivated by a reflexive pro-entrepreneurship ideology rather than by a dispassionate evaluation of the costs and benefits applying in each particular case. There is certainly no shortage of small-business practitioners and academics who have vested interests in encouraging this kind of ideology. But it rarely seems to be acknowledged in policy circles that entrepreneurial ventures might also possess drawbacks. For example, small firms provide less training and pay lower wages than large firms, so a policy of encouraging small firms at the expense of large ones could actually damage the national skill base. Second, the last few decades have witnessed an explosion of policy initiatives in many countries, which now resemble ‘a patchwork quilt of complexity and idiosyncrasy’ (Curran, 2000, p. 36).35 This has generated confusion among the intended beneficiaries. As noted above, that might help explain why take-up rates for public-sector

Taxation, regulation and other policies

461

information-based support services have been so low. A plethora of policy initiatives also complicates the task of evaluating specific policies, because policies tend to work jointly rather than in isolation. This problem adds to other difficulties of evaluating government policies towards entrepreneurship, that are chiefly attributable to a paucity of rigorous cost-benefit analyses. Third, most entrepreneurs and economists believe that governments regulate too much. Some regulation is undoubtedly necessary and welfare-improving (e.g. basic public health legislation), and can create entrepreneurial opportunities as well as constraints. But governments seem to find it just too tempting to devise complicated and onerous regulations which hinder entrepreneurship. By its very nature, regulation – and the mindset that accompanies it – is the antithesis of entrepreneurship; and bureaucrats, the handmaidens of regulation, can have vested interests in proliferating it. An extensive political economy and public choice literature demonstrates that politicians and interest groups have incentives to create and channel subsidies in ways that benefit themselves rather than society at large. As noted in previous chapters, recognising potential scope for welfare-improving government intervention in a particular situation does not automatically justify such intervention. Intervention invariably incurs direct costs, in addition to the opportunity cost of public funds; and bureaucrats cannot always be trusted to spend public money wisely. In a nutshell, not every problem is worth fixing. As the burden of red tape grows year on year, one is led to wonder whether governments are actually paying lip service to entrepreneurship rather than genuinely seeking to promote it. Public policy-makers might start to improve matters by being clearer about their objectives. Instead of formulating policies to boost employment creation and growth, they should address only specific and demonstrable market failures. Most importantly they should do less, funding fewer but better-publicised initiatives, and cutting back on onerous regulations. One practical suggestion in this regard is for governments outside the United States to emulate the US Regulatory Flexibility Act, which forces government agencies to consider the impact of proposed new regulations on small businesses, and to consider alternatives that are less likely to harm small firms. They might also insert ‘sunset clauses’ into regulations that affect business in general and entrepreneurship in particular. More fundamentally, governments should start to set aside their prejudice that there is always too little entrepreneurship, and at least consider the possibility that there could instead be too much of it. The last point is especially important because policies designed to help entrepreneurs can cause inefficiency and hinder successful enterprise. For example, a policy of taxfavouring small firms reduces the pre-tax rate of return entrepreneurs require to launch a new venture and can reduce incentives to grow if the tax benefit is withdrawn for large firms. Such a policy also encourages investment in inefficient projects, since rational investors might forsake investments with higher (pre-tax) rates of return in favour of less productive tax-favoured investments (Holtz-Eakin, 2000). Canadian Labour Sponsored Venture Capital Associations provide a good practical example of the inefficiency which can result in cases like these (Cumming and MacIntosh, 2006). Inefficiency

462

Public policy

is exacerbated if governments try to ‘play venture capitalist’ by picking projects for funding, despite having less experience than private-sector venture capitalists. If governments have to get involved with the funding of risky projects at all, they are probably best advised to stick with unfashionable but promising ventures rather than chic sectors which already enjoy abundant access to private finance (Gompers and Lerner, 1999). Finally, governments outside the United States might usefully look to the Small Business Innovation Research (SBIR) programme as an effective way of supporting innovative VC-backed entrepreneurship (see chapter 16). Above all else, it is incumbent on policy-makers and supporters of entrepreneurship to take into account private-sector reactions to public policies, and to identify clearly the social benefits motivating government intervention. At present there is only limited evidence about these alleged benefits. As with public policies in other areas, governments should legislate and intervene only when there is hard evidence that the benefits of their actions outweigh the costs.

Notes 1. To see this, consider again the Lucas (1978) model, in which individuals possess heterogeneous entrepreneurial abilities x. A cut-off level of ability x˜ separates those who become entrepreneurs (with x ≥ x˜ ) from those who become employees (with x < x˜ ). The cut-off ability is defined implicitly by the condition π(˜x) = w, where entrepreneurial profit π is an increasing function of ability (πx > 0) and w is the employee wage. A flat-rate (or excess profit) tax τ ∈ (0, 1) on entrepreneurs’ profits changes the cut-off ability to x˜˜ , defined by (1 − τ )π(x˜˜ ) = w. Evidently x˜˜ > x˜ , so entrepreneurs with abilities in the range x ∈ [˜x, x˜˜ ) close down their firms and switch to paid employment. The extra supply of labour decreases the employee wage, while the smaller number of firms implies excess demand for their goods, so increasing their prices. Thus larger firms gain from the tax at the expense of smaller ones, and may actually become better off as a result of the profit tax on net. 2. Note that individuals in this problem are choosing between expected utility of EU [(1 − τ )π()] in entrepreneurship and the certain utility of U [(1 + ς )w] in paid employment, where  is a random variable, τ is the tax rate and ς is the subsidy. 3. The insurance effect can be seen as Var [(1−τ )π()] = (1−τ )2 Var π() < Var π(). The effect on the number of entrepreneurs becomes ambiguous however if there are borrowing constraints. If the tax on entrepreneurs decreases the optimal capital stock in entrepreneurship, it is easier for entrants to overcome the borrowing constraints. In which case, the equilibrium number of entrepreneurs can actually increase (Kihlstrom and Laffont, 1982). 4. Now individuals choose between expected utility of EU [ς + (1 − τ )π()] in entrepreneurship and the certain utility of U [ς + (1 − τ )w] in paid employment. 5. A utilitarian social welfare function – a weighted sum of every individual’s utility – is increasing in efficiency (because output and utility are higher) and insurance (because lower risk-taking increases expected utility), and decreasing in inequality (because of diminishing marginal utility). 6. Specifically, Clemens and Heinemann (2006) consider the impact of the following redistributive tax system: yn = y1−γ yγ , where yn and y are post-tax and pre-tax income, respectively; y is a ‘threshold’ income level above (below) which individuals are net contributors to (recipients of) a revenue-neutral redistributive tax; and 0 < γ < 1 is the rate of progressivity of a progressive tax system. They show that the number of entrepreneurs increases (decreases) (is invariant) with respect to γ as the coefficient of relative risk aversion is greater than (less than) (equal to) 1.

Taxation, regulation and other policies

463

7. Some commentators contend that differential eligibility for welfare benefits is an ‘accidental’ feature of early legislation. Originally, many welfare policies were conceived for full-time workers in regular employment: the self-employed were effectively ‘left out’ (see Aronson, 1991). However, governments are also no doubt aware of potential moral hazard problems entailed by extending accident, sickness and unemployment insurance to the self-employed. Working for themselves without third-party oversight, the self-employed would presumably find it easier to spuriously claim benefits. 8. Examples of (i) include Long (1982a, 1982b) and Moore (1983a). For (ii), see Blau (1987), Yuengert (1995), Parker (1996), Cowling and Mitchell (1997), Robson (1998b), Robson and Wren (1999), Schuetze (2000), Bruce (2000), Fölster (2002), Parker and Robson (2004) and Bruce and Mohsin (2006). For (iii), see Bruce (2000). 9. Bruce (2000) and Stabile (2004) have adopted a similar strategy. 10. See Staber and Bogenhold (1993), Parker and Robson (2004) and Stabile (2004). Robson (2003) is a dissenting voice. 11. For a detailed summary of 1980s self-employment schemes for the unemployed in Germany, France, the UK and Denmark, see Meager (1993, 1994). 12. Storey (1994a) reported that around 14 per cent of EAS entrants failed in the first year of operation, during which time the EAS subsidy was received. Taylor (1999) cited a UK Department of Employment report which estimated that two-thirds of individuals completing their first year of the EAS were still in business two years later. These are higher survival rates than Taylor found for British self-employed workers overall. 13. Consistent with this, time-series econometric evidence from Parker (1996) suggests that the EAS increased UK self-employment rates only modestly and in the short run only. See Meager (1993, 1994) for time-series regression results for Germany. 14. See Hinz and Jungbauer-Gans (1999), Pfeiffer and Reize (2000) and Baumgartner and Caliendo (2008). 15. See Birley (1985), Bennett and Robson (1999) and Watson (2007). 16. For UK evidence see Fadahunsi et al. (2000) and Marlow (2003). For German evidence, see Kontos (2003). According to Robson et al. (2008), women business-owners make more frequent use of informal family-network contacts than their male counterparts do. 17. See Bendick and Egan (1987), Marshall et al. (1993) and Chen and Williams (1999) for evidence from France, Britain and the USA, respectively. Parker and Belghitar (2006) report that US formal business assistance programmes have a significant positive impact on the probability that nascent entrepreneurs actually launch their ventures. 18. This model predicts that entry costs c will (for a given average wage, w) increase the x˜ which defines the marginal entrepreneur: E(π(x)) − c = w. The number of entrepreneurs 1 − G(˜x) therefore decreases, where G(·) is the cumulative density function of ability. Only the ablest entrepreneurs remain. 19. See e.g. Fonseca et al. (2001), Fogel et al. (2006), Ciccone and Papioannou (2007) and Ho and Wong (2007). 20. The econometric model used by Klapper et al. (2006) is a double-censored tobit for the entry rate E (defined as the ratio of the number of new firms to the total number of firms in industry i in country j): Eij = α0 + α1
Di + α2
Dj + α3 ISij + α4 [Ei US × REGj ] + uij , where Di , Dj are industry and country dummies, respectively; ISij is a control variable measuring sales share of i in country j; and uij is a disturbance term. As noted in the text, the key variable is the interaction term, where Ei US is the ‘natural’ rate entry rate in industry i in the (relatively barrier-free) USA, and REGj is the size or impact of a particular regulation in European country j. Hence one would expect the key parameter α4 to be negative. Notice that by focusing on

464

21.

22.

23.

24. 25. 26. 27. 28.

29.

30. 31. 32.

33. 34.

Public policy interactions, country-level factors are absorbed into α2
Dj , and estimates of α4 instead identify the effects of entry regulations through industries that are most responsive to them. Using comparative data from England and Spain, Capelleras et al. (2008) claim that fewer but larger entrants are one of two responses to heavy regulation; another is a greater number of unregistered (informal) firms. To put this in perspective, Baird and Morrison (2005) point out that only 10–15 per cent of businesses which close file for bankruptcy; and only 62.5 per cent of those which file for bankruptcy are shut down and liquidated. So this is not a common exit route. See Scott and Smith (1986), Gropp et al. (1997), Fan and White (2003) and Berkowitz and White (2004). The evidence of a positive relationship between the borrowing rate and exemption levels is consistent with Chamley’s (1983) separating contract model mentioned in chapter 7. However, Berkowitz and White’s (2004) finding that banks treat incorporated and non-corporate firms in a similar way casts some doubt on the practical relevance of Chamley’s theory. In practice, lenders might blur the distinction between the two types of firm by requiring both to secure loans on personal assets. However, there is mixed evidence about correlations between economic freedom and entrepreneurship (c.f. Kreft and Sobel, 2005; Bjørnskov and Foss, 2008). See Kumar et al. (2002), Desai et al. (2003), Guiso et al. (2004), Jappelli et al. (2005), Fogel et al. (2006) and Laeven and Woodruff (2007). And by generating rents in this way, incumbent entrepreneurs also have a vested interest in preserving low contract enforcement regimes – see chapter 12. See Robson (2003), Torrini (2005) and Ciccone and Papioannou (2007). See Rajan and Zingales (2003), Giannetti (2003) and Cumming and Johan (2007). Another area where prudent regulation can conceivably help entrepreneurs is in relation to late payment by business customers. Cheap, effective and enforceable regulation of persistent late payment can in principle help entrepreneurs manage cash flow problems which might otherwise jeopardise their survival. See Ilmakunnas and Kanniainen (2001), Torrini (2005), Wennekers et al. (2005) and Bjørnskov and Foss (2008). Grant (1996) presents similar findings in a cross-state US study, while also noting that state-level economic development policies designed to attract new business investment were ineffective. See Boyd (1991), van Praag and van Ophem (1995) and Hyytinen and Ilmakunnas (2007). See Ilmakunnas and Kanniainen (2001), Parker and Robson (2004), Kanniainen and Vesala (2005), Torrini (2005) and Wennekers et al. (2005). Kanniainen and Vesala (2005) show this in a model where entrepreneurial firms hire labour and bargain with trade unions over wages. By increasing the union’s breakdown value relative to that of firms, unemployment compensation increases the bargained wage and therefore makes entrepreneurship less attractive on net (provided the second-round effect from lower labour demand does not depress the wage too much). Greater trade-union bargaining strength therefore depresses entrepreneurship. See Grant (1996), Kanniainen and Vesala (2005) and Torrini (2005). Interestingly, a sample of 14,846 responses to the 1990 European Values Survey reveals that self-employed people are more individually orientated than the rest of the population, believing strongly in the importance of individual responsibility and having a strong work ethic (Beugelsdijk and Noorderhaven, 2005). These values appear to be consistent across different ethnic groups (Morris and Schindehutte, 2005). The self-employed are also significantly more likely to favour a smaller state and less generous unemployment benefits (Beugelsdijk and Noorderhaven, 2005). Any causality lurking behind these associations is probably more likely to run from individuals’ political views to propensities to engage in entrepreneurship, rather than the contrary. Evidence shows that former communist party members are significantly less likely to become entrepreneurs

Taxation, regulation and other policies

465

than the average person after economies liberalise (see Robert and Bukodi, 2004, for evidence from Hungary; and Wu, 2006, for evidence from China). 35. Curran (2000) cites a report by Gavron et al. (1998) that examined no fewer than 200 UK government entrepreneurship initiatives, whose total cost amounted to £632 million in 1995–96 alone. Fuest et al. (2002) cite research identifying over 500 programmes to promote entrepreneurial entry in Germany.

References

Abell, P., R. Crouchley and C. Mills (2001). Social capital and entrepreneurship in Great Britain, Enterprise and Innovation Management Studies, 2(2), pp. 119–44 Abraham, K. G. and S. K. Taylor (1996). Firms’ use of outside contractors: theory and evidence, Journal of Labor Economics, 14, pp. 394–424 Acemoglu, D. (1998). Credit market imperfections and the separation of ownership from control, Journal of Economic Theory, 78, pp. 355–81 (2001). Credit market imperfections and persistent unemployment in Europe, European Economic Review, 45, pp. 665–79 (2008). Oligarchic versus demographic societies, Journal of the European Economic Association, 6(1), pp. 1–44 Acemoglu, D., P. Aghion and F. Zilibotti (2006). Distance to frontier, selection, and economic growth, Journal of the European Economic Association, 4(1), pp. 37–74 Acs, Z. J. (1999). Are Small Firms Important? Their Role and Impact, Boston MA, Kluwer Acs, Z. J. and C. Armington (2004a). The impact of geographic differences in human capital on service firm formation rates, Journal of Urban Economics, 56, pp. 244–78 (2004b). Employment growth and entrepreneurial activity in cities, Regional Studies, 38(8), pp. 911–27 (2006). Entrepreneurship, Geography, and American Economic Growth, New York, Cambridge University Press Acs, Z. J. and D. B. Audretsch (1987a). Innovation in large and small firms, Economics Letters, 23, pp. 109–12 (1987b). Innovation, market structure, and firm size, Review of Economics and Statistics, 49, pp. 567–74 (1988). Innovation in large and small firms: an empirical analysis, American Economic Review, 78, pp. 678–90 (1989). Small firm entry in manufacturing, Economica, 56, pp. 255–65 (1990). Innovation and Small Firms, Cambridge MA, MIT Press (1991). R&D, firm size and innovative activity, in Z. J. Acs and D. B. Audretsch (eds.), Innovation and Technological Change, Ann Arbor, University of Michigan Press, pp. 39–59 (1993). Small Firms and Entrepreneurship: An East–West Perspective, Cambridge University Press (2003). Innovation and technological change, in Z. J. Acs and D. B. Audretsch (eds.), Handbook of Entrepreneurship Research: An Interdisciplinary Survey and Introduction, Boston MA, Kluwer, pp. 55-79 Acs, Z. J. and S. Gifford (1996). Innovation of entrepreneurial firms, Small Business Economics, 8(3), pp. 203–18 466

References

467

Acs, Z. J. and P. Mueller (2008). Employment effects of business dynamics: mice, gazelles and elephants, Small Business Economics, 30(1), pp. 85–100 Acs, Z. J. and R. J. Phillips (2002). Entrepreneurship and philanthropy in American capitalism, Small Business Economics, 19(3), pp. 189–204 Acs, Z. J., and A. Varga (2005). Entrepreneurship, agglomeration and technological change, Small Business Economics, 24, pp. 323–34 Acs, Z. J., D. B. Audretsch and B. Carlsson (1990). Flexibility, plant size, and industrial restructuring, in Z. J. Acs and D. B. Audretsch (eds.), The Economics of Small Firms: A European Challenge, Dordrecht, Kluwer Acs, Z. J., D. B. Audretsch and D. S. Evans (1994). Why does the self-employment rate vary across countries and over time?, Discussion Paper No. 871, CEPR Acs, Z. J., D. B. Audretsch and M. P. Feldman (1994). R&D spillovers and recipient firm size, Review of Economics and Statistics, 100, pp. 336–67 Acs, Z. J., B. Carlsson and C. Karlsson (1999). The linkages among entrepreneurship, SMEs and the macroeconomy, in Z. J. Acs, B. Carlsson and C. Karlsson (eds.), Entrepreneurship, Small and Medium Sized Enterprises and the Macroeconomy, Cambridge University Press, Cambridge, pp. 3–42 Acs, Z. J., D. B. Audretsch, P. Braunerhjelm and B. Carlsson (2004). The missing link: the knowledge filter and entrepreneurship in endogenous growth, Discussion Paper No. 4783, CEPR Admati, A. and P. Pfleiderer (1994). Robust financial contracting and the role of venture capitalists, Journal of Finance, 49, pp. 371–402 Agarwal, R. (1996). Technological activity and the survival of firms, Economics Letters, 52, pp. 101–8 (1997). Survival of firms over the product life cycle, Southern Economic Journal, 97(3), pp. 571–84 (1998). Small firm survival and technological activity, Small Business Economics, 11, pp. 215–24 Agarwal, R. and D. B. Audretsch (2001). Does entry size matter? The impact of the life cycle and technology on firm survival, Journal of Industrial Economics, 49(1), pp. 21–43 Agarwal, R. and M. Gort (1996). The evolution of markets and entry, exit and survival of firms, Review of Economics and Statistics, 78, pp. 489–98 Agarwal, R., M. B. Sarkar and R. Echambadi (2002). The conditioning effect of time on firm survival: an industry life cycle approach, Academy of Management Journal, 45, pp. 971–94 Aghion, P. and P. Bolton (1997). A theory of trickle-down growth and development, Review of Economic Studies, 64, pp. 151–72 Aghion, P. and B. Hermalin (1990). Legal restrictions on private contracts can enhance efficiency, Journal of Law, Economics & Organisation, 6(2), pp. 381–409 Aghion, P. and P. Howitt (1992). A model of growth through creative destruction, Econometrica, 60, pp. 323–51 (1997). A Schumpeterian perspective on growth and competition, in D. M. Kreps and K. F. Wallis (eds.), Advances in Economics and Econometrics: Theory & Applications, Cambridge University Press, pp. 279–319 Aharonson, B. S., J. A. C. Baum and M. P. Feldman (2007). Desperately seeking spillovers? Increasing returns, industrial organisation and the location of new entrants in geographic and technological space, Industrial & Corporate Change, 16(1), pp. 89–130 Ahlin, C. and N. Jiang (2008). Can micro-credit bring development? Journal of Development Economics, 86(1), pp. 1–21 Ahlin, C. and R. M. Townsend (2007). Using repayment data to test across models of joint liability lending, Economic Journal, 117, pp. F11–51 Aidis, R. and M. van Praag (2007). Illegal entrepreneurship experience: does it make a difference for business performance and motivation? Journal of Business Venturing, 22, pp. 283–310

468

References

Aidis, R., S. Estrin and T. Mickiewicz (2008). Institutions and entrepreneurship development in Russia: a comparative perspective, Journal of Business Venturing, 23(6), pp. 656–72 Ajayi-obe O. and S. C. Parker (2005). The changing nature of work among the self-employed in the 1990s: evidence from Britain, Journal of Labor Research, 26, pp. 501–17 Akoten, J. E., Y. Sawada and K. Otsuka (2006). The determinants of credit access and its impacts on micro and small enterprises: the case of garment producers in Kenya, Economic Development & Cultural Change, 54(4), pp. 927–44 Alba-Ramirez, A. (1994). Self-employment in the midst of unemployment: the case of Spain and the United States, Applied Economics, 26, pp. 189–204 Alchian, A. and H. Demsetz (1972). Production, information costs, and economic organisation, American Economic Review, 62, pp. 777–95 Aldrich, H. E. (1999). Organizations Evolving, Newbury Park CA, Sage Aldrich, H. E. and P. H. Kim (2006). A life course perspective on occupational inheritance: selfemployed parents and their children, in M. Ruef and M. Lounsbury (eds.), Research in the Sociology of Organisations, Amsterdam, JAI Press (2007). Small worlds, infinite possibilities? How social networks affect entrepreneurial team formation and search, Strategic Entrepreneurship Journal, 1(1–2), pp. 147–65 Aldrich, H. E. and R. Waldinger (1990). Trends in ethnic businesses in the United States, in R. Waldinger, H. Aldrich and R. Ward (eds.), Ethnic Entrepreneurs, Newbury Park CA, Sage Aldrich, H. E., N. M. Carter and M. Ruef (2004). Teams, in Gartner et al (2004), pp. 299–310 Aldrich, H. E., L. A. Renzuli and N. Langton (1998). Passing on privilege: resources provided by selfemployed parents to their self-employed children, Research in Social Stratification and Mobility, 16, 291–317 Aldrich, H. E., J. Cater, T. Jones, D. McEvoy and P. Velleman (1985). Ethnic residential concentration and the protected market hypothesis, Social Forces, 63, pp. 996–1009 Alexander-Cook, K., D. Bernhardt and J. Roberts (1998). Riding free on the signals of others, Journal of Public Economics, 67, pp. 25–43 Almeida, P. and B. Kogut (1997). The exploration of technological diversity and the geographic localisation of innovation, Small Business Economics, 9(1), pp. 21–31 (1999). Localisation of knowledge and the mobility of engineers in regional networks, Management Science, 45(7), pp. 905–17 Almus, M. (2004). Job creation through public start-up assistance? Applied Economics, 36, pp. 2015–24 Almus, M. and E. A. Nerlinger (1999). Growth of new technology-based firms: which factors matter? Small Business Economics, 13, pp. 141–54 Alsos, G. A. and L. Kolvereid (1998). The business gestation process of novice, serial and parallel business founders, Entrepreneurship Theory & Practice, 22(4), pp. 101–14 Alsos, G. A. and E. Ljunggren (1998). Does the business start-up process differ by gender? A longitudinal study of nascent entrepreneurs, Frontiers of Entrepreneurship Research, Babson College, Wellesley MA Alsos, G. A., E. J. Iskasen and E. Ljunggren (2006). New venture financing and subsequent business growth in men- and women-led businesses, Entrepreneurship Theory & Practice, 30, pp. 667–86 Altman, E. (1983). Corporate Financial Distress, New York, Wiley-Interscience Alvarez, R. and G. Crespi (2003). Determinants of technical efficiency in small firms, Small Business Economics, 20, pp. 233–44 Alvarez, S. and J. Barney (2001). How entrepreneurial firms can benefit from alliances with larger partners, Academy of Management Executive, 15, pp. 139–48 Alvarez, A. and S. C. Parker (2009). Emerging firms and the allocation of control rights: a Bayesian approach, Academy of Management Review, 34(2), pp. 209–27

References

469

Amatucci, F. M. and J. E. Sohl (2007). Business angels: investment processes, outcomes and current trends, in A. Zacharakis and S. Spinelli Jr. (eds.), Entrepreneurship: The Engine of Growth, Volume 2, Westport CT, Praager, pp. 87–107 Amit, R., J. Brander and C. Zott (1998). Why do venture capital firms exist? Theory and Canadian evidence – an essay on the economics of imperfect information, Journal of Business Venturing, 13, pp. 441–46 Amit, R., L. Glosten and E. Muller (1990). Entrepreneurial ability, venture investments, and risk sharing, Management Science, 36(10), pp. 1232–45 (1993). Challenges to theory development in entrepreneurship research, Journal of Management Studies, 30, pp. 815–34 Amit, R., Muller, E. and I. Cockburn (1995). Opportunity costs and entrepreneurial activity, Journal of Business Venturing, 10, pp. 95–106 Amit, R., K. MacCrimmon, C. Zietsma and J. M. Oesch (2000). Does money matter? Wealth attainment as the motive for initiating growth-oriented technology ventures, Journal of Business Venturing, 16, pp. 119–43 Amossé, T. and D. Goux (2004). Entries and exits from self-employment in France over the last twenty years, in Arum and Müller (eds.) (2004), pp. 75–103 Anderson, E. and D. Schmittlein (1984). Integration of the sales force: an empirical examination, Rand Journal of Economics, 15(3), pp. 385–95 Anderson, R. C. and D. M. Reeb (2003). Founding-family ownership and firm performance: evidence from the S&P 500, Journal of Finance, 58, pp. 1301–28 Anderson, S. and J.-M. Baland (2002). The economics of Roscas and intrahousehold resource allocation, Quarterly Journal of Economics, 117(3), pp. 963–95 Andersson, P. and E. Wadensjoe (2007). Do the unemployed become successful entrepreneurs? International Journal of Manpower, 28(7), pp. 604–26 Anton, J. J. and D.A. Yao (1995). Start-ups, spin-offs, and internal projects, Journal of Law, Economics, & Organization, 11, pp. 362–78 Antunes, A., T. Cavalcanti and A. Villamil (2008). The effect of financial repression and enforcement on entrepreneurship and economic development, Journal of Monetary Economics, 55, pp. 278–97 Appelbaum, E. and E. Katz (1986). Measures of risk aversion and comparative statics of industry equilibrium, American Economic Review, 76, pp. 524–29 Arabsheibani, G., D. de Meza, J. Maloney and B. Pearson (2000). And a vision appeared unto them of a great profit: evidence of self-deception among the self-employed, Economics Letters, 67, pp. 35–41 Arauzo-Carod, J. M. and A. Segarra-Blasco (2005). The determinants of entry are not independent of start-up size: some evidence from Spanish manufacturing, International Journal of Industrial Organization, 27(2), pp. 147–65 Arenius, P. and M. Minniti (2005). Perceptual variables and nascent entrepreneurship, Small Business Economics, 24, pp. 233–47 Armendáriz de Aghion, B. (1999). On the design of a credit agreement with peer monitoring, Journal of Development Economics, 60, pp. 79–104 Armendáriz de Aghion, B. and C. Gollier (2000). Peer group formation in an adverse selection model, Economic Journal, 110, pp. 632–43 Armendáriz de Aghion, B. and J. Morduch (2000). Microfinance beyond group lending, Economics of Transition, 8(2), pp. 401–20 Armington, C. and Z. J. Acs (2002). The determinants of regional variation in new firm formation, Regional Studies, 36(1), pp. 33–45 Armington, C. and M. Odle (1982). Small business: how many jobs? Brookings Review, pp. 14–17 Armour, J. and D. Cumming (2006). The legislative road to Silicon Valley, Oxford Economic Papers, 58, pp. 596–635

470

References

Arnott, R. J. and J. E. Stiglitz (1986). Moral hazard and optimal commodity taxation, Journal of Public Economics, 29, pp. 1–24 Aronson, R. L. (1991). Self Employment: A Labour Market Perspective, Ithaca NY, ILR Press Arrighetti, A. and M. Vivarelli (1999). The role of innovation in the post entry performance of new small firms: evidence from Italy, Southern Economic Journal, 65, pp. 927–39 Arrow, K. (1962). Economic welfare and the allocation of resources for invention, in R. Nelson (ed.), The Rate and Direction of Inventive Activity, Princeton NJ, Princeton University Press Arum, R. and W. Müller (eds.) (2004). The Reemergence of Self-employment: A Comparative Study of Self-employment Dynamics and Social Inequality, Princeton NJ, Princeton University Press Ashworth, J., P. Johnson and C. Conway (1998). How good are small firms at predicting employment? Small Business Economics, 10, pp. 379–87 Astebro, T. (2003). The return to independent invention: evidence of risk-seeking, extreme optimism or skewness-loving? Economic Journal, 113, pp. 226–39 Astebro, T. and I. Bernhardt (2003). Start-up financing, owner characteristics and survival, Journal of Economics and Business, 55, 303–19 (2005). The winner’s curse of human capital, Small Business Economics, 24(1), pp. 63–78 Astebro, T. and P. Thompson (2007). Entrepreneurs: jacks of all trades or hobos? Unpublished paper, University of Toronto Atrostic, B. K. (1982). The demand for leisure and non-pecuniary job characteristics, American Economic Review, 72, pp. 428–40 Audretsch, D. B. (1991). New-firm survival and the technological regime, Review of Economics and Statistics, 73, pp. 441–50 (1995a). Innovation and Industry Evolution, Cambridge MA, MIT Press (1995b). Innovation, growth and survival, International Journal of Industrial Organization, 13, pp. 441–57 (2001). Research issues relating to structure, competition and performance of small technologybased firms, Small Business Economics, 16, pp. 37–51 (2003). Entrepreneurship policy and the strategic management of places, in D. M. Hart (ed.), The Emergence of Entrepreneurship Policy, Cambridge University Press, pp. 20–38 (2007a). The Entrepreneurial Society, New York, Oxford University Press (2007b). Entrepreneurship capital and economic growth, Oxford Review of Economic Policy, 23(1), pp. 63–78 Audretsch, D. B. and Z. J. Acs (1994). New firm start-ups, technology and macroeconomics fluctuations, Small Business Economics, 6, pp. 439–49 Audretsch, D. B. and M. P. Feldman (1996). R & D spillovers and the geography of innovation and production, American Economic Review, 86, pp. 630–40 Audretsch, D. B. and M. Fritsch (1994a). The geography of firm births in Germany, Regional Studies, 28, pp. 359–65 (1994b). On the measurement of entry rates, Empirica, 21, pp. 105–113 (1996). Creative destruction: turbulence and economic growth in Germany, in E. Helmstadter and M. Perlman (eds.), Behavioural Norms, Technological Progress and Economic Dynamics, Ann Arbor MI, University of Michigan Press, pp. 137–50 (2002). Growth regimes over time and space, Regional Studies, 36, pp. 113–24 Audretsch, D. B. and J. Jin (1994). A reconciliation of the unemployment-new firm startup paradox, Small Business Economics, 6, pp. 381–85 Audretsch, D. B. and M. Keilbach (2004a). Entrepreneurship capital and economic performance, Regional Studies, 38(8), pp. 949–59 (2004b). Does entrepreneurship capital matter?, Entrepreneurship Theory & Practice, 28(5), pp. 419–29

References

471

(2007a). The theory of knowledge spillover entrepreneurship, Journal of Management Studies, 44(7), pp. 1242–54 (2007b). The localisation of entrepreneurship capital: evidence from Germany, Papers in Regional Science, 86(3), pp. 351–65 Audretsch, D. B. and E. Lehmann (2006). Location and new venture creation, in S. C. Parker (ed.), The Life Cycle of Entrepreneurial Ventures, Springer, New York, pp. 137–60 Audretsch, D. B. and T. Mahmood (1994). The rate of hazard confronting new firms and plants in US manufacturing, Review of Industrial Organisation, 9, pp. 41–56 (1995). New firm survival: new results using a hazard function, Review of Economics and Statistics, 77, pp. 97–103 Audretsch, D. B. and P. E. Stephan (1996). Company-scientist locational links: the case of biotechnology, American Economic Review, 86, pp. 641–52 Audretsch, D. B. and R. Thurik (2000). Capitalism and democracy in the 21st century: from the managed to the entrepreneurial economy, Journal of Evolutionary Economics, 10, pp. 17–34 (2001a). Linking entrepreneurship to growth, STI Working Paper 2081/2, OECD, Paris (2001b). What is new about the new economy: sources of growth in the managed and entrepreneurial economies, Industrial & Corporate Change, 10(1), pp. 267–315 Audretsch, D. B. and M. Vivarelli (1997). Determinants of new firm start-ups in Italy, Empirica, 23, pp. 91–105 Audretsch, D. B., W. Bönte and M. Keilbach (2008). Entrepreneurship capital and its impact on knowledge diffusion and economic performance, Journal of Business Venturing, 23, pp. 687–98 Audretsch, D. B., M. C. Keilbach and E. E. Lehmann (2006). Entrepreneurship and Economic Growth, New York, Oxford University Press Audretsch, D. B., E. Santarelli and M. Vivarelli (1999). Start-up size and industrial dynamics: some evidence from Italian manufacturing, International Journal of Industrial Organization, 17, pp. 965–83 Audretsch, D. B., M.A. Carree,A. J. van Stel andA. R. Thurik (2002). Impeded industrial restructuring: the growth penalty, Kyklos, 55, pp. 81–98 Audretsch, D. B., L. Klomp, E. Santarelli andA. R. Thurik (2004). Gibrat’s Law: are services different? Review of Industrial Organization, 24(3), pp. 301–24 Auerswald, P. E. (2006). The simple economics of technology entrepreneurship: market failure reconsidered, in D. B. Audretsch, I. Grilo and A. R. Thurik (eds.), Handbook of Entrepreneurship Policy, Cheltenham, Edward Elgar, pp. 18–35 Auerswald, P. E. and L. M. Branscomb (2003). Start-ups and spin-offs: collective entrepreneurship between invention and innovation, in D. M. Hart (ed.), The Emergence of Entrepreneurship Policy, Cambridge, Cambridge University Press, pp. 61–91 Ayotte, K. (2007). Bankruptcy and entrepreneurship: the value of a fresh start, Journal of Law, Economics and Organization, 23(1), pp. 161–85 Azevedo, J. P. (2006). Microfinance and poor entrepreneurs, in Parker, S. C. (ed.), The Life Cycle of Entrepreneurial Ventures, New York, Springer, pp. 301–34 Backes-Gellner, U. and A. Werner (2007). Entrepreneurial signalling via education: a success factor in innovative start-ups, Small Business Economics, 29(1–2), pp. 173–90 Baird, D. G. and E. R. Morrison (2005). Serial entrepreneurs and small business bankruptcies, Columbia Law Review, 105, pp. 2310–68 Baker, M. and P. A. Gompers (2003). The determinants of board structure at the initial public offering, Journal of Law and Economics, 66, pp. 569–98 Baker, P. (1993). Taxpayer compliance for the self-employed: estimates from household spending data, Working Paper No. W93/14, Institute for Fiscal Studies, London Baldwin, J. R. (1998). Were small producers the engines of growth in the Canadian manufacturing sector in the 1980s? Small Business Economics, 10(4), pp. 349–64

472

References

Baldwin, J. R. and G. Picot (1995). Employment generation by small producers in the Canadian manufacturing sector, Small Business Economics, 7, pp. 317–31 Baldwin, J. R. and M. Rafiquzzaman (1995). Selection versus evolutionary adaptation: learning and post-entry performance, International Journal of Industrial Organization, 13, pp. 501–22 Banerjee, A. V. and A. F. Newman (1993). Occupational choice and the process of development, Journal of Political Economy, 101, pp. 274–98 Banerjee, A. V., T. Besley and T. W. Guinnane (1994). Thy neighbour’s keeper: the design of a credit co-operative with theory and a test, Quarterly Journal of Economics, 109, pp. 491–515 Banerji, S. and N. van Long (2007). Moral hazards, bankruptcy costs, and international financial capital mobility, Review of Economic Dynamics, 11(2), pp. 369–84 Bania, N., R. Eberts and M. Fogerty (1993). Universities and the start-up of new companies: can we generalise from Route 128 and Silicon Valley? Review of Economics and Statistics, 75, pp. 761–66 Bank of England (1993). Bank Lending to Smaller Businesses, London, Bank of England (1999). The Financing of Ethnic Minority Firms in the United Kingdom, London, Bank of England (2001). Finance for Small Firms: Eighth Report, London, Bank of England Bankman, J. and R. J. Gilson (1999). Why start-ups? Stanford Law Review, 51, 289–308 Bannock, G. (1981). The Economics of Small Firms: Return from the Wilderness, Oxford, Basil Blackwell Bannock, G. and A. Peacock (1989). Government and Small Business, London, Paul Chapman Publishing Baptista, R., V. Escária and P. Madruga (2008). Entrepreneurship, regional development and job creation: the case of Portugal, Small Business Economics, 30, pp. 49–58 Barboza, G. A. and H. Barreto (2006). Learning by association: micro credit in Chiapas, Mexico, Contemporary Economic Policy, 24(2), pp. 316–31 Bardhan, P. (1997). Corruption and evidence: a review of issues, Journal of Economic Literature, 35, pp. 1320–46 Barkham, R., M. Hart and E. Hanvey (1996). Growth in small manufacturing firms: an empirical analysis, in R. Blackburn and P. Jennings (eds.), Small Firms’Contribution to Economic Regeneration, London, Paul Chapman Publishing, pp. 112–25 Barkham, R., G. Gudgin, M. Hart and E. Hanvey (1996). The Determinants of Small Firm Growth: An Inter-regional Study in the United Kingdom 1986–1990, London, Jessica Kingsley Barkman, R. (1994). Entrepreneurial characteristics and the size of the new firm: a model and an econometric test, Small Business Economics, 6, pp. 117–25 Barlevy, G. (2007). On the cyclicality of research and development, American Economic Review, 97(4), pp. 1131–64 Barnett, W. (1997). The dynamics of competitive intensity, Administrative Science Quarterly, 42, pp. 128–60 Baron, D. P. (2007). Corporate social responsibility and social entrepreneurship, Journal of Economics and Management Strategy, 16(3), pp. 683–717 Baron, R. A. (1998). Cognitive mechanisms in entrepreneurship: why and when entrepreneurs think differently from other people, Journal of Business Venturing, 13, pp. 275–94 (2000). Counterfactual thinking and venture formation – the psychology of new venture creation, Journal of Business Venturing, 15, pp. 79–91 Barreto, H. (1989). The Entrepreneur in Microeconomic Theory: Disappearance and Explanation, London, Routledge Barringer, B. R., F. F. Jones and D. O. Neubaum (2005). A quantitative content analysis of the characteristics of rapid-growth firms and their founders, Journal of Business Venturing, 20, pp. 663–87

References

473

Barro, R. J. (1976). The loan market, collateral, and rates of interest, Journal of Money, Credit and Banking, 8, pp. 439–56 Barron, D., E. West and M. Hannan (1994). ‘A time to grow and a time to die’: growth and mortality of credit unions in New York City, 1914–1990, American Journal of Sociology, 100, pp. 381–421 Bartelsman, E., S. Scarpetta and F. Schivardi (2005). Comparative analysis of firm demographics and survival: evidence from micro-level sources in OECD countries, Industrial & Corporate Change, 14, pp. 365–391 Barzel, Y. (1987). The entrepreneur’s reward for self-policing, Economic Inquiry, 25, pp. 103–16 Basu, A. (1998). An exploration of entrepreneurial activity among Asian small businesses in Britain, Small Business Economics, 10, pp. 313–26 Basu, A. and S. C. Parker (2001). Family finance and new business start-ups, Oxford Bulletin of Economics and Statistics, 63, pp. 333–58 Bates, T. (1984). A review of the Small Business Administration’s Major Loan Programs, in P. M. Horvitz and R. R. Pettit (eds.), Small Business Finance: Sources of Financing for Small Business, Part B, Greenwich CT, JAI Press, pp. 211–39 (1985). Entrepreneur human capital endowments and minority business viability, Journal of Human Resources, 20, pp. 540–54 (1990). Entrepreneur human capital inputs and small business longevity, Review of Economics and Statistics, 72, pp. 551–59 (1991). Commercial bank finance in white and black owned small business start ups, Quarterly Review of Economics & Business, 31, pp. 64–80 (1993). Banking on Black Enterprise, Joint Center for Politics and Economic Studies, Washington DC (1994a). Firms started as franchises have lower survival rates than independent small business start-ups, Centre for Economic Studies Discussion Paper, US Census Bureau, Washington DC (1994b). Social resources generated by group support networks may not be beneficial to Asian immigrant-owned small businesses, Social Forces, 72, 671–89 (1995). Self-employment entry across industry groups, Journal of Business Venturing, 10, pp. 143–56 (1997). Race, Self-employment and Upward Mobility: An Illusive American Dream, Baltimore, John Hopkins University Press (1998). Survival patterns among newcomers to franchising, Journal of Business Venturing, 13, pp. 113–30 (2003). Minority business assistance programs are not designed to produce minority business development, in D. M. Hart (ed.), The Emergence of Entrepreneurship Policy, Cambridge University Press, pp. 155–72 (2005). Analysis of young, small firms that have closed: delineating successful from unsuccessful closures, Journal of Business Venturing, 20, pp. 343–58 (2006). The urban development potential of black-owned businesses, Journal of the American Planning Association, 72(2), pp. 272–91 Bates, T. and W. Bradford (1979). Financing Black Economic Development, New York, Academy Press (1992). Factors affecting new firm success and their use in venture capital financing, Journal of Small Business Finance, 2, pp. 23–38 (2007). Traits and performance of the minority venture-capital industry, Annals of the American Academy of Political and Social Science, 613, pp. 95–107 Bates, T. and C. R. Dunham (1993). Asian American success in self-employment, Economic Development Quarterly, 7(2), pp. 199–214

474

References

Bates, T. and L. J. Servon (1998). Microenterprise as an exit route from poverty: recommendations for programs and policy makers, Journal of Urban Affairs, 20, pp. 419–41 (2000). Viewing self-employment as a response to a lack of suitable opportunities for wage work, National Journal of Sociology, 12, pp. 23–53 Batjargal, B. (2007). Internet entrepreneurship: social capital, human capital, and performance of Internet ventures in China, Research Policy, 36, pp. 605–18 Baum, J. and S. Mezias (1992). Localised competition and organisational failure in the Manhattan hotel industry, 1898–1990, Administrative Science Quarterly, 37, pp. 580–604 Baum, J., H. Korn and S. Kotha (1995). Dominant designs and population dynamics in telecommunications services, Social Science Research, 24, pp. 97–135 Baumgartner, H. J. and M. Caliendo (2008). Turning unemployment into self-employment: effectiveness of two start-up programmes, Oxford Bulletin of Economics and Statistics, 70(3), pp. 347–73 Baumol, W. J. (1968). Entrepreneurship in economic theory, American Economic Review, Papers and Proceedings, 58, pp. 64–71 (1983). Toward operational models of entrepreneurship, in J. Ronen (ed.), Entrepreneurship, Lexington, MA, Lexington Books, pp. 29–48 (1990). Entrepreneurship: productive, unproductive, and destructive, Journal of Political Economy, 98, pp. 893–921 (1993a). Entrepreneurship, Management and Structure of Payoffs, Cambridge MA, MIT Press (1993b). Formal entrepreneurship theory in economics: existence and bounds, Journal of Business Venturing, 8(3), pp. 197-210 (2007). Small firms: why market-driven innovation can’t get along without them, Paper presented at IFN Conference, Stockholm, September Bayus, B. L. and R. Agarwal (2007). The role of pre-entry experience, entry timing, and product technology strategies in explaining firm survival, Management Science, 53(12), pp. 1887–1902 Bearse, P. J. (1984). An econometric analysis of black entrepreneurship, Review of Black Political Economy, 12, pp. 117–34 Becht, M., C. Mayer and H. F. Wagner (2008). Where do firms incorporate? Deregulation and the costs of entry, Journal of Corporate Finance, 14, pp. 241–56 Beck, S. (1986). Mobility from pre-retirement to postretirement job, The Sociological Quarterly, 27, pp. 515–31 Beck, T., A. Demirgüc-Kunt, and V. Maksimovic (2004). Bank competition and access to finance: international evidence, Journal of Money, Credit and Banking, 36(3), pp. 627–48 Becker, E. (1984). Self-employed workers: an update to 1983, Monthly Labor Review, 107, pp. 14–18 Becker-Blease, J. R. and J. E. Sohl (2007). Do women-owned businesses have equal access to angel capital? Journal of Business Venturing, 22(4), pp. 503–21 Beckman, C. M., M. D. Burton and C. O’Reilly (2007). Early teams: the impact of team demography on VC financing and going public, Journal of Business Venturing, 22(2), pp. 147–73 Bednarzik, R. W. (2000). The role of entrepreneurship in US and European job growth, Monthly Labor Review, July, pp. 3–16 Begley, T. M. and D. P. Boyd (1987). Psychological characteristics associated with performance in entrepreneurial firms and smaller businesses, Journal of Business Venturing, 2, pp. 79–93 Belke, A., Fehn, R. and N. Foster (2005). VC investment and labour market performance: a panel data analysis, in V. Kanniainen and C. Keuschnigg (eds.), Venture Capital, Entrepreneurship and Public Policy, Cambridge MA, MIT Press, pp. 97–126 Bell, C. (1990). Interactions between institutional and informal credit agencies in rural India, World Bank Economic Review, 4, pp. 297–327 Bell, R. G., C. B. Moore and H. A. Al-Shammari (2008). Country of origin and foreign IPO legitimacy, Entrepreneurship Theory & Practice, 32(1), pp. 185–202

References

475

Ben-Ner, A. (1988). The life-cycle of worker-owned firms in market economies: a theoretical analysis, Journal of Economic Behavior & Organization, 10(3), pp. 287–313 Bendick, M. and M. L. Egan (1987). Transfer payment diversion for small business development: British and French experience, Industrial and Labor Relations Review, 40, pp. 528–42 Bennett, R. J. and P. J. A. Robson (1999). The use of external business advice by SMEs in Britain, Entrepreneurship & Regional Development, 11, pp. 155–80 Benz, M. and B. S. Frey (2004). Being independent raises happiness at work, Swedish Economic Policy Review, 11(2), pp. 97–134 (2008). Being independent is a great thing: subjective evaluations of self-employment and hierarchy, Economica, 75, pp. 362–83 Berchicci, L. and C. L. Tucci (2006). Entrepreneurship, technology and Schumpeterian innovation: entrants and incumbents, in M. Casson, B. Yeung, A. Basu and N. Wadeson (eds.), The Oxford Handbook of Entrepreneurship, Oxford University Press, pp. 332–50 Bergemann, D. and U. Hege (1998). Venture capital financing, moral hazard, and learning, Journal of Banking and Finance, 22, pp. 703–35 (2006). The financing of innovation: learning and stopping, Rand Journal of Economics, 36(4), pp. 719–52 Berger, A. N. and G. F. Udell (1990). Collateral, loan quality and bank risk, Journal of Monetary Economics, 25, pp. 21–42 (1992). Some evidence on the empirical significance of credit rationing, Journal of Political Economy, 100, pp. 1047–77 (1995). Relationship lending and lines of credit in small firm finance, Journal of Business, 68, pp. 351–81 (2003). Small business and debt finance, in Z. J. Acs and D. B. Audretsch (eds.), Handbook of Entrepreneurship Research: An Interdisciplinary Survey and Introduction, Boston MA, Kluwer, pp. 299–328 Berger, A. N., W. S. Frame and N. H. Miller (2005). Credit scoring and the availability, price and risk of small business credit, Journal of Money, Credit and Banking, 37, pp. 191–222 Berglöf, E. (1994). A control theory of venture capital finance, Journal of Law, Economics & Organisation, 10, pp. 247–67 Bergmann, H. and R. Sternberg (2007). The changing face of entrepreneurship in Germany, Small Business Economics, 28(2–3), pp. 205–21 Bergström, F. (2000). Capital subsidies and the performance of firms, Small Business Economics, 14, pp. 183–93 Berkowitz, D. and D. N. DeJong (2003). Policy reform and growth in post-Soviet Russia, European Economic Review, 47(2), pp. 337–52 (2005). Entrepreneurship and post-socialist growth, Oxford Bulletin of Economics and Statistics, 67(1), pp. 25–46 Berkowitz, D. and J. Holland (2001). Does privatisation enhance or deter small enterprise formation? Economics Letters, 74, pp. 53–60 Berkowitz, D. and J. E. Jackson (2006). Entrepreneurship and the evolution of income distributions in Poland and Russia, Journal of Comparative Economics, 34, pp. 338–56 Berkowitz, J. and M. J. White (2004). Bankruptcy and small firms’ access to credit, Rand Journal of Economics, 35, pp. 69–84 Bernanke, B. and M. Gertler (1990). Financial fragility and economic performance, Quarterly Journal of Economics, 105, pp. 87–114 Bernardo, A. E. and I. Welch (2001). On the evolution of overconfidence and entrepreneurs, Journal of Economics & Management Strategy, 10, pp. 301–30 Bernhardt, D. (2000). Credit rationing?, American Economic Review, 90, pp. 235–39

476

References

Bernhardt, I. (1994). Competitive advantage in self-employment and paid work, Canadian Journal of Economics, 27, pp. 273–89 Bertrand, M. and F. Kramarz (2002). Does entry regulation hinder job creation? Evidence from the French retail industry, Quarterly Journal of Economics, 117, pp. 1369–1413 Bertrand, M. and A. Schoar (2006). The role of family in family firms, Journal of Economic Perspectives, 20(2), pp. 73–96 Besanko, D. and A. V. Thakor (1987a). Collateral and rationing: sorting equilibria in monopolistic and competitive credit markets, International Economic Review, 28, 671–89 (1987b). Competitive equilibrium in the credit market under asymmetric information, Journal of Economic Theory, 42, pp. 167–82 Besim, M. and G. P. Jenkins (2005). Tax compliance: when do employees behave like the selfemployed? Applied Economics, 37, pp. 1201–1208 Besley, T. (1995). Nonmarket institutions for credit and risk sharing in low-income countries, Journal of Economic Perspectives, 9, pp. 115–27 Besley, T. and S. Coate (1995). Group lending, repayment incentives and social collateral, Journal of Development Economics, 46(1), pp. 1–18 Besley, T., S. Coate and G. Loury (1993). The economics of rotating savings and credit associations, American Economic Review, 83, pp. 792–810 (1994). Rotating savings and credit associations, credit markets and efficiency, Review of Economic Studies, 61(4), pp. 701–19 Bester, H. (1985a). Screening vs. rationing in credit markets with imperfect information, American Economic Review, 75, pp. 850–55 (1985b). The level of investment in credit markets with imperfect information, Journal of Institutional & Theoretical Economics, 141, pp. 503–15 (1987). The role of collateral in credit markets with imperfect information, European Economic Review, 31, pp. 887–99 (1994). The role of collateral in a model of debt renegotiation, Journal of Money, Credit and Banking, 26, pp. 72–86 Beugelsdijk, S. and N. Noorderhaven (2005). Personality characteristics of self-employed: an empirical study, Small Business Economics, 24, pp. 159–67 Bhide, A. V. (1994). How entrepreneurs craft strategies that work, Harvard Business Review, 72(2), 150–61 (2000). The Origin and Evolution of New Businesses, Oxford University Press Biais, B. and C. Gollier (1997). Trade credit and credit rationing, Review of Financial Studies, 10, pp. 903–37 Biais, B., T. Mariotti, G. Plantin and J.-C. Rochet (2007). Dynamic security design: convergence to continuous time and asset price implications, Review of Economic Studies, 74, pp. 345–90 Bianchi, M. and M. Henrekson (2005). Is neoclassical economics still entrepreneurless? Kyklos, 58(3), pp. 353–77 Bikhchandani, S., D. Hirshleifer and I. Welch (1998). Learning from the behaviour of others: conformity, fads and informational cascades, Journal of Economic Perspectives, 12, pp. 151–70 Bilsen, V. and J. Konings (1998). Job creation, job destruction, and growth of newly established, privatised and state-owned enterprises in transition economies, Journal of Comparative Economics, 26, pp. 429–45 Binks, M. and A. Jennings (1986). Small firms as a source of economic rejuvenation, in J. Curran et al (eds.), The Survival of the Small Firm, Vol. 1, Aldershot, Gower, pp. 19–37 Birch, D. L. (1979). The Job Generation Process, Cambridge MA, MIT Programme on Neighbourhood and Regional Change Birch, D. L., A. Haggerty and W. Parsons (1997). Who’s Creating Jobs, Cambridge MA, Cognetics Inc.

References

477

Birley, S. (1985). The role of networks in the entrepreneurial process, Journal of Business Venturing, 1, pp. 107–17 (1989). Female entrepreneurs: are they really different? Journal of Small Business Management, 27, pp. 32–37 Birley, S. and P. Westhead (1994). A taxonomy of business start-up reasons and their impact on firm growth and size, Journal of Business Venturing, 9, pp. 7–31 Birnbaum, P. H. (1984). The choice of strategic alternatives under increasing regulation in high technology companies, Academy of Management Journal, 27(3), pp. 489–510 Bitler, M. P., T. J. Moskowitz and A. Vissing-Jørgensen (2005). Testing agency theory with entrepreneur effort and wealth, Journal of Finance, 60, pp. 539–76 Bitler, M. P., A. Robb and J. Wolken (2001). Financial services used by small businesses: evidence from the 1998 survey of small business finances, Federal Reserve Bulletin, 87, 183–205 Bjørnskov, C. and N. J. Foss (2008). Economic freedom and entrepreneurial activity: some crosscountry evidence, Public Choice, 134(3–4), pp. 307–28 Black, B. S. and R. J. Gilson (1998). Venture capital and the structure of capital markets: banks versus stock markets, Journal of Financial Economics, 47, pp. 243–77 Black, J. and D. de Meza (1997). Everyone may benefit from subsidising entry to risky occupations, Journal of Public Economics, 66, pp. 409–24 Black, J., de Meza, D., and D. Jeffreys (1996). House prices, the supply of collateral and the enterprise economy, Economic Journal, 106, pp. 60–75 Black, S. E. and P. E. Strahan (2004). Business formation and the deregulation of the banking industry, in Holtz-Eakin, D. and H. S. Rosen (eds.), Public Policy and the Economics of Entrepreneurship, Cambridge MA, MIT Press, pp. 59–82 Blanchard, L., B. Zhao and J. Yinger (2008). Do lenders discriminate against minority and women entrepreneurs? Journal of Urban Economics, 63, pp. 467–97 Blanchflower, D. G. (2000). Self-employment in OECD countries, Labour Economics, 7, pp. 471–505 (2004). Self-employment: more may not be better, Swedish Economic Policy Review, 11(2), pp. 15–73 Blanchflower, D. G. and R. B. Freeman (1994). Did the Thatcher reforms change British labour market performance?, in R. Barrell (ed.), The UK Labour Market, Cambridge University Press, pp. 51–92 Blanchflower, D. G. and B. D. Meyer (1994). A longitudinal analysis of the young self-employed in Australia and the United States, Small Business Economics, 6, pp. 1–19 Blanchflower, D. G. and A. Oswald (1990). Self-employment and the enterprise culture, in R. Jowell, S. Witherspoon and L. Brook (eds.), British Social Attitudes: The 1990 Report, London, Gower (1998). What makes an entrepreneur?, Journal of Labor Economics, 16, pp. 26–60 Blanchflower, D. G. and J. Wainwright (2005). An analysis of the impact of affirmative action programs on self-employment in the construction industry, Discussion Paper No. 1856, IZA, Bonn Blanchflower, D. G., P. B. Levine and D. J. Zimmerman (2003). Discrimination in the small-business credit market, Review of Economics and Statistics, 85, pp. 930–43 Blanchflower, D. G., A. Oswald and A. Stutzer (2001). Latent entrepreneurship across nations, European Economic Review, 45, pp. 680–91 Blau, D. M. (1985). Self-employment and self-selection in developing country labour markets, Southern Economic Journal, 52, pp. 351–63 (1987). A time-series analysis of self-employment in the United States, Journal of Political Economy, 95, pp. 445–67 Bluedorn,A. C. and G. Martin (2008). The time frames of entrepreneurs, Journal of Business Venturing, 23(1), pp. 1–20

478

References

Blumberg, B. F. and P. M. de Graaf (2004). Dutch self-employment between 1980 and 1997, in Arum and Müller (eds.) (2004), pp. 104–34 Blumberg, B. F. and W. A. Letterie (2008). Business starters and credit rationing, Small Business Economics, 30, pp. 187–200 Blundell, R. and T. MaCurdy (1999). Labour supply: a review of alternative approaches, in O. Ashenfelter and D. Card (eds.), Handbook of Labour Economics, Volume 3A, Amsterdam, Elsevier, pp. 1559–1695 Boadway, R. and M. Keen (2006). Financing and taxing new firms under asymmetric information, FinanzArchiv, 62(4), pp. 471–502 Boadway, R. and M. Sato (1999). Information acquisition and government intervention in credit markets, Journal of Public Economic Theory, 1(3), pp. 283–308 Boadway, R. and J.-F. Tremblay (2005). Public economics and start-up entrepreneurs, in V. Kanniainen and C. Keuschnigg (eds.), Venture Capital, Entrepreneurship and Public Policy, Cambridge MA, MIT Press, pp. 181–219 Boadway, R., M. Marchand and P. Pestieau (1991). Optimal linear income taxation in models with occupational choice, Journal of Public Economics, 46, pp. 133–62 Boadway, R., N. Marceau, M. Marchand and M. Vigneault (1998). Entrepreneurship, asymmetric information and unemployment, International Tax and Public Finance, 5, pp. 307–27 Boden, R. J. (1996). Gender and self-employment selection: an empirical assessment, Journal of Socioeconomics, 25, pp. 671–82 (1999a). Flexible working hours, family responsibilities and female self-employment, American Journal of Economics & Sociology, 58, pp. 71–83 (1999b). Gender inequality in wage earnings and female self-employment selection, Journal of Socio-Economics, 28, pp. 351–64 Boden, R. J. and A. R. Nucci (1997). Counting the self-employed using household and business sample data, Small Business Economics, 9, pp. 427–36 (2000). On the survival prospects of men’s and women’s new business ventures, Journal of Business Venturing, 15, pp. 347–62 Bodie, Z., R. C. Merton and W. F. Samuelson (1992). Labor flexibility and portfolio choice in a life cycle model, Journal of Economic Dynamics and Control, 16, pp. 427–49 Bohá˘cek, R. (2006). Financial constraints and entrepreneurial investment, Journal of Monetary Economics, 53, pp. 2195–2222 (2007). Financial intermediation with credit constrained agents, Journal of Macroeconomics, 29, pp. 741–59 Böheim, R. and U. Muehlberger (2006). Dependent forms of self-employment in the UK: identifying workers on the border between employment and self-employment, Discussion Paper No. 1963, IZA, Bonn Bolger, F., B. D. Pulford and A. M. Colman (2008). Market entry decisions, Experimental Psychology, 55(2), pp. 113–20 Bonaccorsi di Patti, E. and G. Dell’Ariccia (2004). Bank competition and firm creation, Journal of Money, Credit and Banking, 36(2), pp. 225–51 Bond, E. W. (1986). Entrepreneurial ability, income distribution and international trade, Journal of International Economics, 20, pp. 343–56 Bond, P. and R. Townsend (1996). Formal and informal financing in a Chicago ethnic neighbourhood, Federal Reserve Bank of Chicago Economic Perspectives, 20, pp. 3–27 Bond, S. and J. Sales (2001). Household work in the UK: a household analysis of the BHPS 1994, Work, Employment & Society, 15, pp. 233–50 Bone, J. (2006). ‘The longest day’: flexible contracts, performance related pay and risk shifting in the UK direct selling sector, Work, Employment & Society, 20(1), pp. 109–27

References

479

Bonnett, C. and A. Furnham (1991). Who wants to be an entrepreneur? A study of adolescents interested in a Young Enterprise Scheme, Journal of Economic Psychology, 12, pp. 465–78 Boot, A. W. A. (1992). Why hang on to losers? Divestitures and takeovers, Journal of Finance, 47(4), pp. 1401–23 Boot, A. W. A. and A. V. Thakor (1994). Moral hazard and secured lending in an infinitely repeated credit market game, International Economic Review, 35, pp. 899–920 Boot, A. W. A., R. Gopalan and A. V. Thakor (2006). The entrepreneur’s choice between private and public ownership, Journal of Finance, 61(2), pp. 803–36 Boot, A. W. A., A. V. Thakor and G. F. Udell (1991). Secured lending and default risk: equilibrium analysis, policy implications and empirical results, Economic Journal, 101, pp. 458–72 Bopaiah, C. (1998). Availability of credit to family businesses, Small Business Economics, 11, pp. 75– 86 Borjas, G. J. (1986). The self-employment experience of immigrants, Journal of Human Resources, 21, pp. 485–506 (1990). Friends or Strangers: The Impact of Immigrants on the US Economy, New York, Basic Books Borjas, G. J. and S. G. Bronars (1989). Consumer discrimination and self-employment, Journal of Political Economy, 97, pp. 581–605 Borooah, V. K. and M. Hart (1999). Factors affecting self-employment among Indian and black Caribbean men in Britain, Small Business Economics, 13, pp. 111–29 Bosma, N., C. M. van Praag, R. Thurik and G. de Wit (2004). The value of human and social capital investments for the business performance of startups, Small Business Economics, 23, pp. 227–36 Bostic, R. W. and K. P. Lampani (1999). Racial differences in patterns of small-business finance: the importance of local geography, Federal Reserve Bank of Chicago, pp. 149–79 Boston, T. D. and L. R. Boston (2007). Secrets of gazelles: the differences between high-growth and low-growth business owned by African American entrepreneurs, Annals of the American Academy of Political and Social Science, 613, pp. 108–30 Bosworth, B. P., A. S. Carron and E. Rhyne (1987). The Economics of Government Credit, Brookings Institution, Washington DC Bottazzi, L. and da Rin, M. (2002). Venture capital in Europe: Euro.NM and the financing of European innovative firms, Economic Policy, 34, 231–69 (2005). Financing entrepreneurial firms in Europe: facts, issues, and research agenda, in V. Kanniainen and C. Keuschnigg (eds.), Venture Capital, Entrepreneurship and Public Policy, Cambridge MA, MIT Press, pp. 3–32 Bouchard, T. J., and J. C. Loehlin (2001). Genes, evolution, and personality, Behavioural Genetics, 31(3), 243–73 Bovaird, C. (1990). Introduction to Venture Capital Finance, London, Pitman Bowen, H. P. and D. de Clercq (2008). Institutional context and the allocation of entrepreneurial effort, Journal of International Business Studies, 39(4), pp. 747–67 Bowlin, O. D. (1984). Lease financing: an attractive method of financing for small firms, in P. M. Horvitz and R. R. Pettit (eds.), Small Business Finance: Sources of Financing for Small Business, Part B, Greenwich CT, JAI Press, pp. 155–62 Boyd, D. P. (1984). Type A behaviour, financial performance and organisational growth in small business firms, Journal of Occupational Psychology, 57, pp. 137–40 Boyd, R. L. (1990). Black and Asian self-employment in large metropolitan areas: a comparative analysis, Social Problems, 37, pp. 258–69 (1991). A contextual analysis of black self-employment in large metropolitan areas, 1970–1980, Social Forces, 70(2), pp. 409–29 (2006). Transformation of the black business elite, Social Science Quarterly, 87(3), pp. 602–17

480

References

Bracoud, F. and B. Hillier (2000). Equity or debt? Contracts in markets with asymmetric information, Manchester School, 68, pp. 1–23 Bradford, W. D. (2003). The wealth dynamics of entrepreneurship for black and white families in the US, Review of Income and Wealth, 49(1), pp. 89–116 Bratkowski, A., I. Grosfield and J. Rostowski (2000). Investment and finance in de novo private firms, Economics of Transition, 8(1), pp. 101–16 Braunerhjelm, P. and B. Borgman (2004). Geographical concentration, entrepreneurship and regional growth: Evidence from regional data in Sweden, 1975–99, Regional Studies, 38(8), pp. 929–47 Brav, A., and P. A. Gompers (1997). Myth or reality? The long-run underperformance of Initial Public Offerings: evidence from venture and nonventure capital-backed companies, Journal of Finance, 52, pp. 1791–1821 Bregger, J. E. (1996). Measuring self-employment in the United States, Monthly Labor Review, 119, pp. 3–9 Breschi, S. and F. Lissoni (2001). Knowledge spillovers and local innovation systems: a critical survey, Industrial & Corporate Change, 10, pp. 975–1005 Bridge, S., K. O’Neill and S. Cromie (1998). Understanding Enterprise, Entrepreneurship and Small Business, Basingstoke, Macmillan Briscoe, G., A. Dainty and S. Millett (2000). The impact of the tax system on self-employment in the British construction industry, International Journal of Manpower, 21(8), 596–613 Brito, P. and A. S. Mello (1995). Financial constraints and firm post-entry performance, International Journal of Industrial Organization, 13, pp. 543–65 Brixy, U., S. Kohaut and C. Schnabel (2007). Do newly founded firms pay lower wages? First evidence from Germany, Small Business Economics, 29, pp. 161–71 Brock, W. A., and D. S. Evans (1986). The Economics of Small Businesses: Their Role and Regulation in the US Economy, New York, Holmes and Meier (1989). Small business economics, Small Business Economics, 1(1), pp. 7–20 Brockhaus, R. H. (1980). Risk taking propensity of entrepreneurs, Academy of Management Journal, 23, pp. 509–20 Broussard, N., R. Chami and G. D. Hess (2003). (Why) do self-employed parents have more children?, Working Paper No. 1103, CESIfo, Munich Brown, C. and J. L. Medoff (1989). The employer size–wage effect, Journal of Political Economy, 97, pp. 1027–59 (2003). Firm age and wages, Journal of Labor Economics, 21(3), pp. 677–97 Brown, C., Hamilton, J. and J. Medoff (1990). Employers Large and Small, Cambridge MA, Harvard University Press Brown, S. and J. G. Sessions (1998). Education, employment status and earnings: a comparative test of the strong screening hypothesis, Scottish Journal of Political Economy, 45, pp. 586–91 (1999). Education and employment status: a test of the strong screening hypothesis in Italy, Economics of Education Review, 18, pp. 397–404 Brown, S., L. Farrell and J. G. Sessions (2006). Self-employment matching: an analysis of dual earner couples and working households, Small Business Economics, 26, pp. 155–72 Brown, S., L. Farrell, M. N. Harris and J. G. Sessions (2006). Risk preference and employment contract type, Journal of the Royal Statistical Society, Series A, 169(4), pp. 849–63 Bruce, D. (1999). Do husbands matter? Married women entering self-employment, Small Business Economics, 13, pp. 317–29 (2000). Effects of the United States tax system on transitions into self-employment, Labour Economics, 7, pp. 545–74 (2002). Taxes and entrepreneurial endurance: evidence from the self-employed, National Tax Journal, 55, pp. 5–24

References

481

Bruce, D. and M. Mohsin (2006). Tax policy and entrepreneurship: new time series evidence, Small Business Economics, 26, pp. 409–25 Bruce, D. and H. J. Schuetze (2004). The labour market consequences of experience in selfemployment, Labour Economics, 11, pp. 575–98 Bruce, D., D. Holtz-Eakin and J. Quinn (2000). Self-employment and labour market transitions at older ages, Working Paper, CRR WP 2000–13, Boston College Brüderl, J. and P. Preisendörfer (1998). Network support and the success of newly founded businesses, Small Business Economics, 10, pp. 213–25 Brüderl, J., P. Preisendörfer and R. Ziegler (1992). Survival chances of newly founded business organisations, American Sociological Review, 57, pp. 227–42 Brunetti, M. J. (2006). The estate tax and the demise of the family firm, Journal of Public Economics, 90, pp. 1975–1993 Brush, C. G. (1992). Research on women business owners: past trends, a new perspective and future directions, Entrepreneurship Theory & Practice, 4, pp. 5–30 (2006). Women entrepreneurs: a research overview, in M. Casson, B. Yeung, A. Basu and N. Wadeson (eds.), The Oxford Handbook of Entrepreneurship, Oxford University Press, pp. 611–28 Brush, C. G., N. M. Carter, E. Gatewood, P. G. Greene and M. M. Hart (2003). Venture capital access: is gender an issue?, in D. M. Hart (ed.), The Emergence of Entrepreneurship Policy, Cambridge University Press, pp. 141–54 Bruton, G., D. Ahlstrom and K. Obloj (2008). Entrepreneurship in emerging economies: where are we today and where should the research go in the future? Entrepreneurship Theory & Practice, 32(1), pp. 1–14 Bruton, G., V. Fried and R. D. Hisrich (1997). Venture capitalists and CEO dismissal, Entrepreneurship Theory & Practice, 21(3), pp. 41–54 (2000). CEO dismissal in venture capital-backed firms: evidence from an agency perspective, Journal of Business Venturing, 24(4), pp. 69–77 Brynjolfsson, E., T. W. Malone, V. Gurbaxani and A. Kambil (1994). Does information technology lead to smaller firms? Management Science, 40(12), pp. 1628–44 Budig, M. J. (2006). Intersections on the road to self-employment: gender, family and occupational class, Social Forces, 84(4), pp. 2223–39 Buenstorf, G. (2007). Evolution on the shoulders of giants: entrepreneurship and firm survival in the German laser industry, Review of Industrial Organization, 30(3), pp. 179–202 Bull, I. and F. Winter (1991). Community differences in business births and business growths, Journal of Business Venturing, 6, pp. 29–43 Burkart, M., F. Panunzi and A. Shleifer (2003). Family firms, Journal of Finance, 58, pp. 2167–2201 Burke, A. E. and A. Hanley (2006). Bank interest margins and business start-up collateral: Testing for convexity, Scottish Journal of Political Economy, 53(3), pp. 319–34 Burke, A. E., F. R. Fitz-Roy and M. A. Nolan (2000). When less is more: distinguishing between entrepreneurial choice and performance, Oxford Bulletin of Economics and Statistics, 62, pp. 565–87 (2002). Self-employment wealth and job creation: the roles of gender, non-pecuniary motivation and entrepreneurial ability, Small Business Economics, 19, pp. 255–70 (2008). What makes a die-hard entrepreneur? Beyond the ‘employee or entrepreneur’ dichotomy, Small Business Economics, 31(2), pp. 93–115 Burmeister, K. and C. Schade (2007). Are entrepreneurs’ decisions more biased? An experimental investigation of the susceptibility to status quo bias, Journal of Business Venturing, 22(3), pp. 340–62

482

References

Busenitz, L. and J. Barney (1997). Differences between entrepreneurs and managers in large organisations: biases and heuristics in strategic decision-making, Journal of Business Venturing, 12, pp. 9–30 Butler, J. S. and C. Herring (1991). Ethnicity and entrepreneurship in America, Sociological Perspectives, 34, pp. 79–94 Buttner, E. H. and D. P. Moore (1997). Womens’ organisational exodus to entrepreneurship: self-reported motivations and correlates with success, Journal of Small Business Management, 35(1), pp. 34–46 Buttner, E. H. and B. H. Rosen (1988). Bank loan officers’ perceptions of characteristics of men, women and successful entrepreneurs, Journal of Business Venturing, 3, pp. 249–58 (1992). Rejection in the loan application process: male and female entrepreneurs’ perceptions and subsequent intentions, Journal of Small Business Management, 30(1), pp. 58–65 Bygrave, W. and C. W. Hofer (1991). Theorising about entrepreneurship, Entrepreneurship Theory & Practice, 16(2), pp. 13–22 Bygrave, W. and M. Minitti (2000). The social dynamics of entrepreneurship, Entrepreneurship Theory & Practice, 24, pp. 25–36 Caballero, R. J. and M. L. Hammour (1994). The cleansing effect of recessions, American Economic Review, 84, pp. 1350–68 Cabeza Pereiro, J. (2008). The status of self-employed workers in Spain, International Labour Review, 147(1), pp. 91–99 Cabral, L. (1995). Sunk costs, firm size and firm growth, Journal of Industrial Economics, 43, pp. 161–72 Cagetti, M. and M. de Nardi (2006). Entrepreneurship, frictions, and wealth, Journal of Political Economy, 114(5), pp. 835–70 Caliendo, M. and A. S. Kritikos (2008). Is entrepreneurial success predictable? An ex ante analysis of the character-based approach, Kyklos, 61(2), pp. 189–214 Calvo, G. and S. Wellisz (1980). Technology, entrepreneurs, and firm size, Quarterly Journal of Economics, 95, pp. 663–77 Camerer, C. and D. Lovallo (1999). Overconfidence and excess entry: an experimental approach, American Economic Review, 89, pp. 306–18 Camerer, C., L. Babcock, G. Loewenstein and R. Thaler (1997). Labour supply of New York City cabdrivers: one day at a time, Quarterly Journal of Economics, 112, pp. 407–41 Campbell, T. S. (1979). Optimal investment financing decisions and the value of confidentiality, Journal of Financial & Quantitative Analysis, 14, pp. 913–24 Campbell, M. and M. Daly (1992). Self-employment: into the 1990s, Employment Gazette, 100, pp. 269–92 Canning, D., C. W. Jefferson, and J. E. Spencer (2003). Optimal credit rationing in not-for-profit financial institutions, International Economic Review, 44, pp. 243–61 Cantillon, R. (1755). Essai Sur la Nature du Commerce en Général, 1931 edition, London, Macmillan Capelleras, J.-L. and F. J. Greene (2008). The determinants and growth implications of venture creation speed, Entrepreneurship & Regional Development, 20, pp. 317–43 Capelleras, J.-L., K. F. Mole, F. J. Greene and D. J. Storey (2008). Do more heavily regulated economies have poorer performing new ventures? Evidence from Britain and Spain, Journal of International Business Studies, 39(4), pp. 688–704 Capone, C. A. Jr. and B. H. Capone (1992). Rational entry into risky markets with negative expected returns, Journal of Economic Behavior & Organization, 18, pp. 415–25 Caputo, R. K. and A. Dolinsky (1998). Women’s choice to pursue self-employment: the role of financial and human capital of household members, Journal of Small Business Management, 36, pp. 8–17

References

483

Carland, J. W., F. Hoy, W. R. Boulton and J. C. Carland (1984). Differentiating small business owners from entrepreneurs, Academy of Management Review, 9, pp. 354–59 Carlsson, B. (1989). The evolution of manufacturing technology and its impact on industrial structure – an international study, Small Business Economics, 1, pp. 21–37 (1992). Management of flexible manufacturing – an international comparison, Omega, 20, pp. 11–22 Carlsson, B., D. B. Audretsch and Z. J. Acs (1994). Flexible technology and plant size – United States manufacturing and metalworking industries, International Journal of Industrial Organization, 12, pp. 359–72 Carod, J. M. A., D. L. Solis and M. M. Bofarull (2008). New business formation and employment growth: some evidence for the Spanish manufacturing industry, Small Business Economics, 30, pp. 73–84 Carr, D. (1996). Two paths to self-employment? Women’s and men’s self-employment in the United States, 1980, Work & Occupations, 23, pp. 26–53 Carrasco, R. (1999). Transitions to and from self-employment in Spain: an empirical analysis, Oxford Bulletin of Economics and Statistics, 61, pp. 315–41 Carree, M. A. (2006). On factors promoting and hindering entry and exit, in S. C. Parker (ed.), The Life Cycle of Entrepreneurial Ventures, New York, Springer, pp. 161–83 Carree, M. A. and A. R. Thurik (1996). Entry and exit in retailing: incentives, barriers, displacement and replacement, Review of Industrial Organization, 11, pp. 155–72 (2003). The impact of entrepreneurship on economic growth, in Z. J. Acs and D. B. Audretsch (eds.), Handbook of Entrepreneurship Research, Boston MA, Kluwer, pp. 437–71 (2006). Understanding the role of entrepreneurship for economic growth, in M. Carree and A. R. Thurik (eds.), The Handbook of Entrepreneurship and Economic Growth, Cheltenham, Edward Elgar, pp. ix–xix (2008). The lag structure of the impact of business ownership on economic performance in OECD countries, Small Business Economics, 30(1), pp. 101–10 Carree, M. A., A. van Stel, R. Thurik and S. Wennekers (2002). Economic development and business ownership: an analysis using data of 23 OECD countries in the period 1876–1996, Small Business Economics, 19, pp. 271–90 (2007). The relationship between economic development and business ownership revisited, Entrepreneurship & Regional Development, 19(3), pp. 281–91 Carrington, W. J., K. McCue, and B. Pierce (1996). The role of employer/employee interactions in labour market cycles: evidence from the self-employed, Journal of Labor Economics, 14, pp. 571–602 Carroll, C. D. (2002). Portfolios of the rich, in L. Guiso, M. Haliassos and T. Jappelli (eds.), Household Portfolios, Cambridge MA, MIT Press Carroll, G. R. and M. Hannan (2000). The Demography of Corporations and Industries, Princeton NJ, Princeton University Press Carroll, G. R. and P. Huo (1986). Organisational task and institutional environment in ecological perspective: findings from the local newspaper industry, American Journal of Sociology, 91, pp. 838-73 Carroll, G. R. and E. Mosakowski (1987). The career dynamics of self-employment, Administrative Sciences Quarterly, 32, pp. 570–89 Carroll, G. R., L. S. Bigelow, M-D. L. Seidel and L. B. Tsai (1996). The fates of de novo and de alio producers in the American automobile industry 1885–1981, Strategic Management Journal, 17, pp. 117–37 Carroll, R., D. Holtz-Eakin, M. Rider and H. S. Rosen (2000a). Income taxes and entrepreneurs’ use of labour, Journal of Labor Economics, 18, pp. 324–51

484

References

(2000b). Entrepreneurs, income taxes and investment, in J. B. Slemrod (ed.), Does Atlas Shrug? The Economic Consequences of Taxing the Rich, New York, Russell Sage, pp. 427–55 (2001). Personal income taxes and the growth of small firms, in J. Poterba (ed.), Tax Policy and the Economy, Cambridge MA, MIT Press pp. 121–47 Carter, N. M., W. B. Gartner and P. D. Reynolds (1996). Exploring start-up event sequences, Journal of Business Venturing, 11, pp. 151–66 Carter, N. M., M. Williams and P. D. Reynolds (1997). Discontinuance among new firms in retail: the influence of initial resources, strategy and gender, Journal of Business Venturing, 12, 125–45 Carter, N. M., W. B. Gartner, K. G. Shaver and E. J. Gatewood (2003). The career reasons of nascent entrepreneurs, Journal of Business Venturing, 18, pp. 13–39 Carter, S. and P. Rosa (1998). The financing of male- and female-owned businesses, Entrepreneurship & Regional Development, 10, pp. 225–41 Casamatta, C. (2003). Financing and advising: optimal financial contracts with venture capitalists, Journal of Finance, 58, pp. 2059–86 Casamatta, C. and C. Haritchabalet (2007). Experience, screening and syndication in venture capital investments, Journal of Financial Intermediation, 16(3), pp. 368–98 Caselli, F. and N. Gennaioli (2005). Credit constraints, competition, and meritocracy, Journal of the European Economic Association, 3(2–3), pp. 679–89 Casey, B. and S. Creigh (1988). Self-employment in Great Britain: its definition in the LFS, in tax and social security law and in labour law, Work, Employment & Society, 2, pp. 381–91 Cassar, A., L. Crowley and B. Wydick (2007). The effect of social capital on group loan repayment: evidence from field experiments, Economic Journal, 117, pp. F85–F106 Cassar, G. (2007). Money, money, money? A longitudinal investigation of entrepreneur career reasons, growth preferences and achieved growth, Entrepreneurship & Regional Development, 19, pp. 89–107 Cassar, G. and B. Gibson (2007). Forecast rationality in small firms, Journal of Small Business Management, 45(3), pp. 283–302 Cassiman, B. and M. Ueda (2006). Optimal project rejection and new firm start-ups, Management Science, 52(2), pp. 262–75 Casson, M. (1995). Entrepreneurship and Business Culture, Aldershot, Edward Elgar (2003). The Entrepreneur: An Economic Theory, Second Edition, Cheltenham, Edward Elgar Cavalluzzo, K. S. and L. C. Cavalluzzo (1998). Market structure and discrimination: the case of small businesses, Journal of Money, Credit and Banking, 30, pp. 771–92 Cavalluzzo, K. S. and J. D. Wolken (2005). Small business loan turndowns, personal wealth, and discrimination, Journal of Business, 78(6), pp. 2153–77 Cavalluzzo, L. C., J. D. Cavalluzzo and J. D. Wolken (2002). Competition, small-business financing, and discrimination: evidence from a new survey, Journal of Business, 75, pp. 641–80 Caves, R. E. (1998). Industrial organisation and new findings on the turnover and mobility of firms, Journal of Economic Literature, 96, pp. 1947–82 Cefis, E. and O. Marsili (2006). Survivor: the role of innovation in firms’ survival, Research Policy, 35, pp. 626–41 Cestone, G. and L. White (2003). Anti-competitive financial contracting: the design of financial claims, Journal of Finance, 58, pp. 2109–41 Cetorelli, N. and P. E. Strahan (2006). Finance as a barrier to entry: bank competition and industry structure in local US markets, Journal of Finance, 61(1), pp. 437–61 Chaganti, R. and S. Parasuraman (1996). A study of the impacts of gender on business performance and managerial patterns in small business, Entrepreneurship Theory & Practice, 21(2), 73–75

References

485

Chahine, S., I. Filatotchev and M. Wright (2007). Venture capitalists, business angels and performance of entrepreneurial IPOs in the UK and France, Journal of Business, Finance & Accounting, 34(3–4), pp. 505–28 Chamley, C. (1983). Entrepreneurial abilities and liabilities in a model of self-selection, Bell Journal of Economics, 14, pp. 70–80 Chan, Y-S. and G. Kanatos (1985). Asymmetric valuations and the role of collateral in loan agreements, Journal of Money, Credit and Banking, 17, pp. 84–95 Chandler, G. N. and S. H. Hanks (1998). An examination of the substitutability of founders’human and financial capital in emerging business ventures, Journal of Business Venturing, 13, pp. 353–69 Chandy, R. K. and G. J. Tellis (2000). The incumbent’s curse? Incumbency, size and radical product innovation, Journal of Marketing, 64(3), pp. 1–17 Charles, K. K. and E. Hurst (2003). The correlation of wealth across generations, Journal of Political Economy, 111, pp. 1155–82 Chell, E., C. Haworth, and S. Brearley (1991). The Entrepreneurial Personality: Concepts, Cases and Categories, London, Routledge Chen, J.-H. and M. Williams (1999). The determinants of business failures in the US low-technology and high-technology industries, Applied Economics, 31, pp. 1551–63 Cho, Y. (1986). Inefficiencies from financial liberalisation in the absence of well-functioning equity markets, Journal of Money, Credit and Banking, 18, pp. 191–99 Choi, Y. R. and P. H. Phan (2006). The influences of economic and technology policy on the dynamics of new firm formation, Small Business Economics, 26, pp. 493–503 Choi, Y. R., M. Lévesque and D. A. Shepherd (2008). When should entrepreneurs expedite or delay opportunity exploitation? Journal of Business Venturing, 23, pp. 333–55 Chowdury, P. R. (2005). Group lending: sequential financing, lender monitoring and joint liability, Journal of Development Economics, 77(2), pp. 415–39 Chrisman, J. J. and F. Katrishen (1994). The economic impact of small business development center counselling services in the United States: 1990–91, Journal of Business Venturing, 9, 271–80 Christiansen, C. (1993). The rigid disk drive industry: a history of commercial and technological turbulence, Business History Review, 67, pp. 531–88 Christiansen, C. and J. Bower (1996). Customer power, strategic investment, and the failure of leading firms, Strategic Management Journal, 17, pp. 197–218 Christiansen, L. R. (1971). Entrepreneurial income: how does it measure up?, American Economic Review, 71, pp. 575–85 Churchill, N. C. and V. L. Lewis (1983). The five stages of small business growth, Harvard Business Review, 61(3), 30–41 Ciccone, A. and E. Papaioannou (2007). Red tape and delayed entry, Journal of the European Economic Association, 5(2–3), pp. 444–58 Claessens, S. and L. Laeven (2003). Financial development, property rights and growth, Journal of Finance, 58(6), pp. 2401–36 Claessens, S., S. Djankov and L. Lang (2000). The separation of ownership and control in East Asian corporations, Journal of Financial Economics, 58, pp. 81–112 Clain, S. H. (2000). Gender differences in full-time self-employment, Journal of Economics & Business, 52, 499–513 Clark, A. E. and A. J. Oswald (1996). Satisfaction and comparison income, Journal of Public Economics, 61(3), pp. 359–81 Clark, K. and S. Drinkwater (1998). Ethnicity and self-employment in Britain, Oxford Bulletin of Economics and Statistics, 60, pp. 383–407 (2000). Pushed out or pulled in? Self-employment among ethnic minorities in England and Wales, Labour Economics, 7, pp. 603–28

486

References

(2002). Enclaves, neighbourhood effects and employment outcomes: ethnic minorities in England and Wales, Journal of Population Economics, 15, pp. 5–29 Clark, K., S. Drinkwater and D. Leslie (1998). Ethnicity and self-employment earnings in Britain 1973–95, Applied Economics Letters, 5, pp. 631–34 Clemens, C. (2006). Monopolistic competition and entrepreneurial risk-taking, Economics Letters, 93, pp. 68–74 (2008). Imperfect competition and growth with entrepreneurial risk, German Economic Review, 9(2), pp. 180–206 Clemens, C., and M. Heinemann (2006). On the effects of redistribution on growth and entrepreneurial risk-taking, Journal of Economics, 88(2), pp. 131–58 Clementi, G. L. and H. A. Hopenhayn (2006). A theory of financing constraints and firm dynamics, Quarterly Journal of Economics, 121, pp. 229–65 Clemenz, G. (1986). Credit Markets with Asymmetric Information, Lecture Notes in Economics and Mathematical Systems No. 272, Berlin, Springer-Verlag Co, C. Y., I. N. Gang and M-S. Yun (2005). Self-employment and wage earning in Hungary, Review of Development Economics, 9(2), pp. 150–65 Coase, R. (1937). The nature of the firm, Economica (N.S.), 4, pp. 386–405 Coate, S. and S. Tennyson (1992). Labour market discrimination, imperfect information and selfemployment, Oxford Economic Papers, 44, pp. 272–88 Coco, G. (1999). Collateral, heterogeneity in risk attitude and the credit market equilibrium, European Economic Review, 43, pp. 559–74 (2000). On the use of collateral, Journal of Economic Surveys, 14, pp. 191–214 Coelho, M. P. (2004). Unrealistic optimism, entrepreneurship and adverse selection, Unpublished doctoral dissertation, London School of Economics Coelho, M. P., D. de Meza and D. J. Reyniers (2004). Irrational exuberance, entrepreneurial finance and public policy, International Tax and Public Finance, 11, pp. 391–417 Cohen, W. M. and S. Klepper (1992). The trade-off between firm size and diversity in the pursuit of technological progress, Small Business Economics, 4, pp. 1–14 (1996). A reprise of size and R&D, Economic Journal, 106, pp. 925–51 Cohen, W. M. and D. Levinthal (1989). Innovation and learning, Economic Journal, 99, 569–96 Coleman, S. (2000). Access to capital and terms of credit: a comparison of men- and women-owned businesses, Journal of Small Business Management, 38, pp. 37–52 (2005). Is there a liquidity crisis for small, black-owned firms? Journal of Developmental Entrepreneurship, 10(1), pp. 29–47 (2007). The role of human and financial capital in the profitability and growth of women-owned small firms, Journal of Small Business Management, 45(3), pp. 303–19 Colombier, N. and D. Masclet (2008). Intergenerational correlation in self-employment: some further evidence from French ECHP data, Small Business Economics, 30, pp. 423–37 Colombo, M. G. and L. Grilli (2005). Start-up size: the role of external financing, Economics Letters, 88, pp. 243–50 Colombo, M. G., M. Delmastro and L. Grilli (2004). Entrepreneurs’ human capital and the startup size of new technology-based firms, International Journal of Industrial Organization, 22, pp. 1183-1211 Connelly, R. (1992). Self-employment and providing child care, Demography, 29, pp. 17–29 Constant, A. (2006). Female proclivity to the world of business, Kyklos, 59(4), pp. 465–80 Constant, A. and Y. Shachmurove (2006). Entrepreneurial ventures and wage differentials between Germans and immigrants, International Journal of Manpower, 27(3), pp. 208–29 Constant, A. and K. F. Zimmerman (2006). The making of entrepreneurs in Germany: are native men and immigrants alike? Small Business Economics, 26, pp. 279–300

References

487

Constant, A., Y. Shachmurove and K. F. Zimmerman (2007). What makes an entrepreneur and does it pay? Native men, Turks and other migrants in Germany, International Migration, 45(4), 71–100 Cook, L. (1999). Trade credit and bank finance: financing small firms in Russia, Journal of Business Venturing, 14, pp. 493–518 Cooper, A. C. (1985). The role of incubator organizations in the founding of growth-oriented firms, Journal of Business Venturing, 1, pp. 75–86 (1986). Entrepreneurship and high tech, in D. Sexton and R. Smilor (eds.), The Art and Science of Entrepreneurship, Cambridge MA, Ballinger Publishing Co., pp. 153–68 Cooper, A. C. and A. V. Bruno (1977). Success among high-technology firms, Business Horizons, 2, pp. 16–22 Cooper, A. C., T. B. Folta and C. Woo (1995). Entrepreneurial information search, Journal of Business Venturing, 10(2), pp. 107–20 Cooper, A. C., F. J. Gimeno-Gascon and C. Y. Woo (1991). A resource-based prediction of new venture survival and growth, Academy of Management Best Paper Proceedings, pp. 68–72 (1994). Initial human and financial capital as predictors of new venture performance, Journal of Business Venturing, 9, pp. 371–95 Cooper, A. C., C. Y. Woo and W. C. Dunkelberg (1988). Entrepreneurs’ perceived chances for success, Journal of Business Venturing, 3, pp. 97–108 (1989). Entrepreneurship and the initial size of firms, Journal of Business Venturing, 4, pp. 317–22 Cooper, A. C., T. Folta, J. Gimeno-Gascon and C. Y. Woo (1992). Entrepreneurs’ exit decisions: the role of threshold expectations, Academy of Management Proceedings, pp. 75–79 Cooper, A. C., M. Ramachandran and D. Schoorman (1997). Time allocation patterns of craftsmen and administrative entrepreneurs: implications for financial performance, Entrepreneurship Theory & Practice, 22, pp. 123–36 Cordes, J. J., C. E. Steuerle and E. Twombly (2004). Dimensions of nonprofit entrepreneurship: an exploratory essay, in D. Holtz-Eakin and H. S. Rosen (eds.), Public Policy and the Economics of Entrepreneurship, Cambridge MA, MIT Press, pp. 115–51 Cornelli, F. and O. Yosha (2003). Stage financing and the role of convertible securities, Review of Economic Studies, 70, pp. 1–32 Cosh, A. D. and A. Hughes (1994). Size, financial structure and profitability: UK companies in the 1980s, in A. Hughes and D. Storey (eds.), Finance and the Small Firm, London, Routledge, pp. 18–63 Cosh, A. D., A. Hughes and E. Wood (1999). Innovation in UK SMEs: causes and consequences for firm failure and acquisition, in Z. Acs and Carlsson, B. (eds.), Entrepreneurship, Small and Medium Sized Enterprises and the Macro Economy, Cambridge University Press Covas, F. (2006). Uninsured idiosyncratic production risk with borrowing constraints, Journal of Economic Dynamics and Control, 30, pp. 2167–90 Covick, O. (1983). Self-employment growth in Australia, in R. Blandy and O. Covick (eds.), Understanding Labour Markets in Australia, Sydney, Allen & Unwin, pp. 84–110 Cowling, M. (1998). Regional determinants of small firm loans under the UK Loan Guarantee Scheme, Small Business Economics, 11, pp. 155–67 (2000). Are entrepreneurs different across countries? Applied Economics Letters, 7, pp. 785–89 (2001). Fixed wages or productivity pay: evidence from 15 EU countries, Small Business Economics, 16(3), pp. 191–204 (2003). The contribution of the self-employed to employment in the EU, Report to the Small Business Service, DTI, London Cowling, M. and N. Clay (1995). Factors influencing take-up rates on the Loan Guarantee Scheme, Small Business Economics, 7, pp. 141–52

488

References

Cowling, M. and P. Mitchell (1997). The evolution of UK self-employment: a study of government policy and the role of the macroeconomy, Manchester School, 65, pp. 427–42 Cowling, M. and M. Taylor (2001). Entrepreneurial women and men: two different species?, Small Business Economics, 16, pp. 167–75 Cowling, M., P. Mitchell and M. Taylor (2004). Job creators, Manchester School, 72, 601–17 Crafts, N. (2006). Regulation and productivity performance, Oxford Review of Economic Policy, 22(2), pp. 186–202 Craggs, P. and P. Jones (1998). UK results from the Community Innovation Survey, Economic Trends, No. 539, HMSO, London Craig, B. R., W. E. Jackson III and J. B. Thomson (2007). Small firm credit discrimination, Small Business Administration guaranteed lending, and local market economic performance, Annals of the American Academy of Political and Social Science, 613, pp. 73–94 Cramer, J. S., J. Hartog, N. Jonker and C. M. van Praag (2002). Low risk aversion encourages the choice for entrepreneurship: an empirical test of a truism, Journal of Economic Behavior & Organization, 48, pp. 29–36 Creedy, J. and P. S. Johnson (1983). Firm formation in manufacturing industry, Applied Economics, 15, pp. 177–85 Cressy, R. C. (1993). The Startup Tracking Exercise: Third Year Report, National Westminster Bank, UK (1996). Are business start-ups debt-rationed?, Economic Journal, 106, pp. 1253–70 (1999). Small business failure: failure to fund or failure to learn? in Z. J. Acs, B. Carlsson and C. Karlsson (eds.), Entrepreneurship, Small and Medium Sized Enterprises and the Macroeconomy, Cambridge University Press, pp. 161–85 (2000). Credit rationing or entrepreneurial risk aversion? An alternative explanation for the Evans and Jovanovic finding, Economics Letters, 66, pp. 235–40 (2006a). Why do most firms die young?, Small Business Economics, 26, pp. 103–16 (2006b). Determinants of small firm survival and growth, in M. Casson, B. Yeung, A. Basu and N. Wadeson (eds.), The Oxford Handbook of Entrepreneurship, Oxford University Press, pp. 161–93 (2006c). Venture capital, in M. Casson, B. Yeung, A. Basu and N. Wadeson (eds.) The Oxford Handbook of Entrepreneurship, Oxford, Oxford University Press, pp. 353–86 Cressy, R. C. and C. Olofsson (1997). European SME financing: an overview, Small Business Economics, 9, pp. 87–96 Cressy, R. C. and O. Toivanen (1997). Is there adverse selection in the credit market?, Mimeo, Warwick Business School Crifo, P. and H. Sami (2008). Entrepreneurship turnover and endogenous returns to ability, Economic Modelling, 25(4), pp. 585–604 Cromie, S. (1987). Motivations of aspiring male and female entrepreneurs, Journal of Occupational Behavior, 8, pp. 251–61 Cueto, B. and J. Mato (2006). An analysis of self-employment subsidies with duration models, Applied Economics, 38, pp. 23–32 Cull, R., A. Demirgüc-Kunt and J. Morduch (2007). Financial performance and outreach: a global analysis of leading microbanks, Economic Journal, 117, pp. F107–F133 Cullen, J. B. and R. H. Gordon (2007). Taxes and entrepreneurial risk-taking: theory and evidence for the US, Journal of Public Economics, 91(7–8), pp. 1479–1505 Cumming, D. J. (2005).Agency costs, institutions, learning, and taxation in venture capital contracting, Journal of Business Venturing, 20(5), pp. 573–622 (2007). Government policy towards entrepreneurial finance: innovation investment funds, Journal of Business Venturing, 22, pp. 193–235 Cumming, D. J. and S. Johan (2007). Regulatory harmonisation and the development of private equity markets, Journal of Banking and Finance, 31(10), pp. 3218–50

References

489

Cumming, D. J., and J. G. MacIntosh (2006). Crowding out private equity: Canadian evidence, Journal of Business Venturing, 21, pp. 569–609 Curran, J. (2000). What is small business policy in the UK for? Evaluation and assessing small business policy, International Small Business Journal, 18, pp. 36–50 Curran, J. and R. A. Blackburn (1993). Ethnic Enterprise and the High Street Bank, Report, Kingston University (2001). Older people and the enterprise society: age and self-employment propensities, Work, Employment & Society, 15, pp. 889–902 Curran, J. and R. Burrows (1989). National profiles of the self-employed, Employment Gazette, 97, pp. 376–85 Curran, J., R. Burrows and M. Evandrou (1987). Small business owners and the selfemployed in Britain: an analysis of the GHS data, Small Business Research Trust, London da Rin, M., G. Nicodano and A. Sembenelli (2006). Public policy and the creation of active venture capital markets, Journal of Public Economics, 90, pp. 1699–1723 Dahlstrand, A. L. (1997). Growth and inventiveness in technology-based spin-off firms, Research Policy, 26, pp. 331–44 Daly, M. (1991). The 1980s – a decade of growth in enterprise, Employment Gazette, 99, pp. 109–34 Daniels, L. (1999). The role of small enterprises in the household and national economy in Kenya: a significant contribution or a last resort? World Development, 27(1), pp. 55–65 Daniels, L. and D. C. Mead (1998). The contribution of small enterprises to household and national income in Kenya, Economic Development & Cultural Change, 47(1), pp. 45–71 Dant, R. P. (1995). Motivation for franchising: rhetoric versus reality, International Small Business Journal, 14, pp. 10–32 Daripa, A. (2008). Optimal collective contract without peer information or peer monitoring, Journal of Development Economics, 86(1), pp. 147–63 Davidsson, P. (2006). Nascent entrepreneurship: Empirical studies and developments, Foundations & Trends in Entrepreneurship, 2(1), pp. 1–76 Davidsson, P. and M. Henrekson (2002). Determinants of the prevalence of start-ups and high-growth firms, Small Business Economics, 19, pp. 81–104 Davidsson, P. and B. Honig (2003). The role of social and human capital among nascent entrepreneurs, Journal of Business Venturing, 18, pp. 301–33 Davidsson, P., L. Achtenhagen and L. Naldi (2006). What do we know about small firm growth?, in Parker, S. C. (ed.), The Life Cycle of Entrepreneurial Ventures, New York, Springer, pp. 361–98 Davidsson, P, F. Delmar and J. Wiklund (2006). Entrepreneurship and the Growth of Firms, Cheltenham, Edward Elgar Davidsson, P., L. Lindmark and C. Olofsson (1994). New firm formation and regional development in Sweden, Regional Studies, 28, pp. 395–410 (1995). Small firms, business dynamics and differential development of economic well-being, Small Business Economics, 7, pp. 301–15 (1998). The extent of over-estimation of small firm job creation – an empirical examination of the regression bias, Small Business Economics, 11, pp. 87–100 Davis, S. J. and J. C. Haltiwanger (1992). Gross job creation, gross job destruction and employment reallocation, Quarterly Journal of Economics, 107, pp. 819–63 Davis, S. J., J. C. Haltiwanger and S. Schuh (1996a). Job Creation and Destruction, Cambridge MA, MIT Press (1996b). Small business and job creation: dissecting the myth and reassessing the facts, Small Business Economics, 8, pp. 297–315 de Clercq, D. and P. Arenius (2006). The role of knowledge in business start-up activity, International Small Business Journal, 24(4), pp. 339–58

490

References

de Clercq, D. and H. J. Sapienza (2005). When do venture capital firms learn from their portfolio companies? Entrepreneurship Theory & Practice, 29(4), pp. 517–35 de Fraja, G. (1996). Entrepreneur or manager: who runs the firm? Journal of Industrial Economics, 44(1), pp. 89–98 de Meza, D. (2002). Overlending? Economic Journal, 112, pp. F17–F31 de Meza, D. and C. Southey (1996). The borrower’s curse: optimism, finance and entrepreneurship, Economic Journal, 106, pp. 375–86 de Meza, D. and D. C. Webb (1987). Too much investment: a problem of asymmetric information, Quarterly Journal of Economics, 102, pp. 281–92 (1988). Credit market efficiency and tax policy in the presence of screening costs, Journal of Public Economics, 36, pp. 1–22. (1989). The role of interest rate taxes in credit markets with divisible projects and asymmetric information, Journal of Public Economics, 39, pp. 33–44 (1990). Risk, adverse selection and capital market failure, Economic Journal, 100, pp. 206–14 (1992). Efficient credit rationing, European Economic Review, 36, pp. 1277–90 (1999). Wealth, enterprise and credit policy, Economic Journal, 109, pp. 153–63 (2000). Does credit rationing imply insufficient lending?, Journal of Public Economics, 78, pp. 215–34 (2006). Credit rationing: Something’s gotta give, Economica, 73, pp. 563–78 Dean, T. J. and J. S. McMullen (2007). Toward a theory of sustainable entrepreneurship: reducing environmental degradation through entrepreneurial action, Journal of Business Venturing, 22(1), pp. 50–76 Dejardin, M. (2004). Sectoral and cross-sectoral effects of retailing firm demographies, Annals of Regional Science, 38, pp. 311–34 Dekle, R. and K. Hamada (2000). On the development of rotating credit associations in Japan, Economic Development & Cultural Change, 49(1), pp. 77–90 Della-Giusta, M. and Z. King (2006). Enterprise culture, in M. Casson, B. Yeung, A. Basu and N. Wadeson (eds.), The Oxford Handbook of Entrepreneurship, Oxford University Press, pp. 629–47 Delmar, F. and P. Davidsson (2000). Where do they come from? Prevalence and characteristics of nascent entrepreneurs, Entrepreneurship & Regional Development, 12, pp. 1–23 Delmar, F. and S. Shane (2003). Does business planning facilitate the development of new ventures? Strategic Management Journal, 24, pp. 1165–85 (2004). Legitimating first: organising activities and the survival of new ventures, Journal of Business Venturing, 19, pp. 385–410 Delmar, F. and J. Wiklund (2008). The effects of small business managers’ growth aspirations on firm growth: a longitudinal study, Entrepreneurship Theory & Practice, 32(3), pp. 437–57 DeMartino, R. and R. Barbato (2003). Differences between women and men MBA entrepreneurs: exploring family flexibility and wealth creation as career motivators, Journal of Business Venturing, 18, pp. 815–32 Demougin, D. and O. Fabel (2007). Entrepreneurship and the division of ownership in new ventures, Journal of Economics and Management Strategy, 16(1), pp. 111–28 Denis, D. J. (2004). Entrepreneurial finance: an overview of the issues and evidence, Journal of Corporate Finance, 10, pp. 301–26 Dennis, W. J. (1996). Self-employment: when nothing else is available? Journal of Labor Research, 17, pp. 645–61 (1997). More than you think: an illustrative estimate of business entries, Journal of Business Venturing, 12(3), pp. 175–96

References

491

(1998). Business regulation as an impediment to the transition from welfare to self-employment, Journal of Labor Research, 19, pp. 263–76 (1999). Research mimicking policy: entrepreneurship/small business policy in the United States, in D. L. Sexton and H. Landstrom (eds.), Handbook of Entrepreneurship, Oxford, Blackwells, pp. 64–82 Desai, M., P. Gompers and J. Lerner (2003). Institutions, capital constraints and entrepreneurial firm dynamics: evidence from Europe, Working Paper No. 10165, NBER Dessi, R. (2005). Start-up finance, monitoring and collusion, Rand Journal of Economics, 36(2), pp. 255–74 DeTienne, D. R., D. A. Shepherd and J. DeCastro (2008). The fallacy of ‘only the strong survive’: the effects of extrinsic motivation on the persistence decisions for underperforming firms, Journal of Business Venturing, 23(5), pp. 528–46 Devine, T. J. (1994a). Characteristics of self-employed women in the United States, Monthly Labor Review, 117, pp. 20–34 (1994b). Changes in wage-and-salary returns to skill and the recent rise in female self-employment, American Economic Review, Papers and Proceedings, 84, pp. 108–113 (1995). CPS earnings data for the self-employed: words of caution for use and interpretation, Journal of Economic and Social Measurement, 21, pp. 213–51 Dewatripont, M., P. Legros and S. Matthews (2003). Moral hazard and capital structure dynamics, Journal of the European Economic Association, 1, pp. 890–930 Diamond, D. (1984). Financial intermediation and delegated monitoring, Review of Economic Studies, 51, pp. 393–414 (1993). Seniority and maturity of debt contracts, Journal of Financial Economics, 33, 341–68 Dilnot, A. W. and C. N. Morris (1981). What do we know about the black economy?, Fiscal Studies, 2, pp. 58–73 Diochon, M., T. V. Menzies and Y. Gasse (2003). Insight into the dynamics of Canadian nascent entrepreneurs’ start-up effort and the role individual factors play in the process, Paper presented at BKERC Conference Dix, M. and N. Gandelman (2007). R & D institutional arrangements: start-up ventures versus internal labs, Manchester School, 75(2), pp. 218–36 Dixit,A. (1989). Entry and exit decisions under uncertainty, Journal of Political Economy, 97, pp. 620– 38 Dixit, A. and R. Rob (1994). Switching costs and sectoral adjustments in general equilibrium with uninsured risk, Journal of Economic Theory, 62, pp. 48–69 Dixon, R. (1991). Venture capital and the appraisal of investments, Omega, 19, pp. 333–44 Djankov, S., R. La Porta, F. Lopez-de-Silanes andA. Shleifer (2002). The regulation of entry, Quarterly Journal of Economics, 117, pp. 1–37 Djankov, S., Y. Qian, G. Roland and E. Zhuravskaya (2005). Who are China’s entrepreneurs? American Economic Review, Papers and Proceedings, 96(2), pp. 348–52 (2006). Entrepreneurship in China and Russia compared, Journal of the European Economic Association, 4(2–3), pp. 352–65 Djankov, S., E. Miguel, Y. Qian, G. Roland and E. Zhuravskaya (2005). Who are Russia’s entrepreneurs? Journal of the European Economic Association, 3(2–3), pp. 587–97 Dobrev, S. D. and W. P. Barnett (2005). Organisational roles and transition to entrepreneurship, Academy of Management Journal, 48, pp. 433-49 Doeringer, P. (1990). Economic security, labour market flexibility, and bridges to retirement, in P. Doeringer (ed.), Bridges to Retirement, Ithaca NY, ILR Press Dolton, P. J. and G. H. Makepeace (1990). Self-employment among graduates, Bulletin of Economic Research, 42, pp. 35–53

492

References

Domar, E. D. and R. A. Musgrave (1944). Proportional income taxation and risk taking, Quarterly Journal of Economics, 58, pp. 388–422 Doms, M., T. Dunne and M. J. Roberts (1995). The role of technology use in the survival and growth of manufacturing plants, International Journal of Industrial Organization, 13, pp. 523–42 Dosi, G. (1988). Sources, procedures, and microeconomic effects of innovation, Journal of Economic Literature, 26, pp. 1120–71 Dreher, A. and M. Gassebner (2007). Greasing the wheels of entrepreneurship? The impact of regulations and corruption on firm entry, KOF Working Paper No. 166, Zurich Drucker, P. (1990). Managing the Non-Profit Organization: Practices and Principles, New York, HarperCollins Du Rietz, A. and M. Henrekson (2000). Testing the female underperformance hypothesis, Small Business Economics, 14, pp. 1–10 Dulleck, U., P. Fritjers and R. Winter-Ebmer (2006). Reducing start-up costs for new firms: the double dividend on the labour market, Scandinavian Journal of Economics, 108(2), pp. 317–37 Dunkelberg, W., A. C. Cooper, C. Woo and W. Denis (1987). New firm growth and performance, in N. Churchill et al (eds.), Frontiers of Entrepreneurship Research, Babson College, pp. 307–21 Dunn, T. and D. Holtz-Eakin (2000). Financial capital, human capital and the transition to selfemployment: evidence from intergenerational links, Journal of Labor Economics, 18, pp. 282– 305 Dunne, P. and A. Hughes (1994). Age, size, growth and survival – UK companies in the 1980s, Journal of Industrial Economics, 42, pp. 115–40 Dunne, T., M. J. Roberts and L. Samuelson (1988). Patterns of firm entry and exit in US manufacturing industries, Rand Journal of Economics, 19(4), pp. 495–515 (1989a). The growth and failure of US manufacturing plants, Quarterly Journal of Economics, 104, pp. 495–515 (1989b). Plant turnover and gross employment flows in the US manufacturing sector, Journal of Labor Economics, 7, pp. 48–71 Duranton, G. and D. Puga (2001). Nursery cities: urban diversity, process innovation, and the life cycle of products, American Economic Review, 91(5), pp. 1454–77 Dushnitsky, G. and M. J. Lenox (2006). When does corporate venture capital investment create firm value? Journal of Business Venturing, 21(6), pp. 753–72 Dustmann, C. and O. Kirchkamp (2002). The optimal migration duration and activity choice after re-migration, Journal of Development Economics, 67, pp. 351–72 Dutz, M. A., J. A. Ordover and R. D. Willig (2000). Entrepreneurship, access policy and economic development: lessons from industrial organisation, European Economic Review, 44, pp. 739–47 Earle, J. S. and Z. Sakova (2000). Business start-ups or disguised unemployment? Evidence on the character of self-employment from transition economies, Labour Economics, 7, pp. 575–601 Easley, D. and M. O’Hara (1983). The economic role of the nonprofit firm, Bell Journal of Economics, 14, pp. 531–38 Edwards, L. N. and E. Field-Hendrey (2002). Home-based work and women’s labour force decisions, Journal of Labor Economics, 20, pp. 170–200 Eisenhardt, K. and K. Schoonhoven (1990). Organisational growth: linking founding team, strategy, environment and growth among US semiconductor ventures, 1978–1988, Administrative Science Quarterly, 35, pp. 504–29 Eisenhauer, J. G. (1995). The entrepreneurial decision: economic theory and empirical evidence, Entrepreneurship Theory & Practice, 20, pp. 67–79 Eisenman, E. J., M. P. Rice and P. Severino (2007). Public policy as an enabler or inhibitor of entrepreneurship: the case of Sarbanes–Oxley, in M. P. Rice and T. G. Habbershorn (eds.), Entrepreneurship: The Engine of Growth, Volume 3, Westport CT, Praeger, pp. 239–62

References

493

Ekelund, J., E. Johansson, M.-J. Järvelin and D. Lichtermann (2005). Self-employment and risk aversion – evidence from psychological test data, Labour Economics, 12, pp. 649–59 Elkan, W. (1988). Entrepreneurs and entrepreneurship in Africa, World Bank Research Observer, 3, pp. 171–85 Elston, J. A., G. W. Harrison and E. E. Ruström (2005). Characterising the entrepreneur using field experiments, Mimeo, University of Central Florida Emery, G. W. (1984). A pure financial explanation for trade credit, Journal of Financial & Quantitative Analysis, 19(3), pp. 271–85 (1987). An optimal financial response to variable demand, Journal of Financial & Quantitative Analysis, 22, pp. 209–25 Employment Department (1989). Small Firms in Britain, London, HMSO Ericson, R. and A. Pakes (1995). Markov-perfect industry dynamics: a framework for empirical work, Review of Economic Studies, 62, pp. 53–82 Eriksson, T. and J. M. Kuhn (2006). Firm spin-offs in Denmark 1981–2000: patterns of entry and exit, International Journal of Industrial Organization, 24, pp. 1021–40 Erosa, A. and A. H. Cabrillana (2008). On finance as a theory of TFP, cross-industry productivity differences, and economic rents, International Economic Review, 49(2), pp. 437–73 Esteve-Pérez, S. and J. A. Manez-Castillejo (2008). The resource-based theory of the firm and firm survival, Small Business Economics, 30, pp. 231–49 Estrin, S., K. E. Meyer and M. Bytchkova (2006). Entrepreneurship in transition economies, in M. Casson, B. Yeung, A. Basu and N. Wadeson (eds.), The Oxford Handbook of Entrepreneurship, Oxford University Press, pp. 693–725 European Commission (2003). Green Paper – Entrepreneurship in Europe, downloadable from http://eropa.eu.int/ Evans, D. S. (1987a). The relationship between firm growth, size and age: estimates for 100 manufacturing industries, Journal of Industrial Economics, 35, pp. 567–81 (1987b). Tests of alternative theories of firm growth, Journal of Political Economy, 95, pp. 657–74 Evans, D. S. and B. Jovanovic (1989). An estimated model of entrepreneurial choice under liquidity constraints, Journal of Political Economy, 97, pp. 808–27 Evans, D. S. and L. S. Leighton (1989a). The determinants of changes in US self-employment, 1968–1987, Small Business Economics, 1, pp. 111–19 (1989b). Some empirical aspects of entrepreneurship, American Economic Review, 79, pp. 519–35 (1990). Small business formation by unemployed and employed workers, Small Business Economics, 2, pp. 319–30 Evans, M. D. R. (1989). Immigrant entrepreneurship: effects of ethnic market size and isolated labour pool, American Sociological Review, 54, pp. 950–62 Everett, J. and J. Watson (1998). Small business failure and external risk factors, Small Business Economics, 11, pp. 371–90 Fadahunsi, A., D. Smallbone and S. Supri (2000). Networking and ethnic minority enterprise development: insights from a North London study, Journal of Small Business & Enterprise Development, 7, pp. 228–40 Fafchamps, M. (2000). Ethnicity and credit in African manufacturing, Journal of Development Economics, 61, pp. 205–35 Fain, T. S. (1980). Self-employed Americans: their number has increased, Monthly Labor Review, 103, pp. 3–8 Fairchild, G. B. (2008). Residential segregation influences on the likelihood of black and white selfemployment, Journal of Business Venturing, 23(1), pp. 46–74 Fairlie, R. W. (1999). The absence of the African-American owned business: an analysis of the dynamics of self-employment, Journal of Labor Economics, 17, pp. 80–108

494

References

(2002). Drug dealing and legitimate self-employment, Journal of Labor Economics, 20, pp. 538–67 (2004). Does business ownership provide a source of upward mobility for blacks and Hispanics?, in D. Holtz-Eakin and H. S. Rosen (eds.), Public Policy and the Economics of Entrepreneurship, Cambridge MA, MIT Press, pp. 153–79 (2005a). Self-employment, entrepreneurship and the NLSY79, Monthly Labor Review, Feb, pp. 40–47 (2005b). Entrepreneurship and earnings among young adults from disadvantaged families, Small Business Economics, 25, pp. 223–36 (2006a). The personal computer and entrepreneurship, Management Science, 52(2), pp. 187–203 (2006b). Entrepreneurship among disadvantaged groups: women, minorities and the less educated, in S. C. Parker (ed.), The Life Cycle of Entrepreneurial Ventures, New York, Springer, pp. 437–75 Fairlie, R. W. and H. A. Krashinsky (2006). Liquidity constraints, household wealth, and entrepreneurship revisited, Discussion Paper No. 2201, IZA, Bonn Fairlie, R. W. and B. D. Meyer (1996). Ethnic and racial self-employment: differences and possible explanations, Journal of Human Resources, 31, pp. 757–93 (1998). Does immigration hurt African-American self-employment?, in D. S. Hamermesh and F. D. Bean (eds.), Help or Hindrance? The Economic Implications of Immigration for AfricanAmericans, New York, Russell Sage Foundation, pp. 185–221 (2000). Trends in self-employment among white and black men during the twentieth century, Journal of Human Resources, 35, pp. 643–69 (2003). The effect of immigration on native self-employment, Journal of Labor Economics, 21, pp. 619–50 Fairlie, R.W. and A.M. Robb (2007a). Families, human capital, and small business: evidence from the Characteristics of Business Owners Survey, Industrial and Labor Relations Review, 2007, 60(2), pp. 225–45 (2007b). Why are black-owned businesses less successful than white-owned businesses? The role of families, inheritances and business human capital, Journal of Labor Economics, 25(2), pp. 289–323 Fan, W. and M. J. White (2003). Personal bankruptcy and the level of entrepreneurial activity, Journal of Law & Economics, 46, 543–67 Farber, H. S. (1999). Alternative and part-time employment arrangements as a response to job loss, Journal of Labor Economics, 17, pp. S142–S169 Fay, M. and L. Williams (1993). Gender bias and the availability of business loans, Journal of Business Venturing, 8, pp. 363–76 Fazzari, S., R. Hubbard and B. Petersen (1988). Financing constraints and corporate investment, Brookings Papers on Economic Activity, 2, pp. 141–95 Feeser, H. R. and G. R. Willard (1989). Incubators and performance: a comparison of high- and low-growth high-tech firms, Journal of Business Venturing, 4, pp. 429–42 (1990). Founding strategy and performance: a comparison of high and low growth high tech firms, Strategic Management Journal, 11(2), pp. 87–98 Feinberg, R. M. (1980). On the consistency of non-profit-maximising behaviour with perfect competition, Southern Economic Journal, 46, pp. 1163–66 (1987). An empirical investigation of owner-manager labour-supply behaviour, Managerial & Decision Economics, 8, pp. 213–16 Feldman, M. P. (1994). Knowledge complementarity and innovation, Small Business Economics, 6, pp. 363–72 (2001). The entrepreneurial event revisited: firm formation in a regional context, Industrial & Corporate Change, 10, pp. 861–91

References

495

(2003). Entrepreneurship and American research universities, in D. M. Hart (ed.), The Emergence of Entrepreneurship Policy, Cambridge University Press, pp. 92–112 Feldman, N. E. and J. Slemrod (2007). Estimating tax non-compliance with evidence from unaudited tax returns, Economic Journal, 117, pp. 327–52 Felstead, A. (1991). The social organisation of the franchise: a case of ‘controlled self-employment’, Work, Employment & Society, 5, pp. 37–57 (1992). Franchising, self-employment and the ‘enterprise culture’: a UK perspective, in P. Leighton and A. Felstead (eds.), The New Entrepreneurs: Self-employment and Small Business in Europe, London, Kogan Page, pp. 237–65 Fender, J. (2005). Self-employment, education and credit constraints: a model of interdependent credit rationing decisions, Journal of Macroeconomics, 27, pp. 31–51 Fender, J. and P. Sinclair (2006). On risk aversion and investment – a theoretical approach, Journal of Institutional & Theoretical Economics, 162(4), pp. 601–26 Ferber, M. A. and J. Waldfogel (1998). The long-term consequences of nontraditional employment, Monthly Labor Review, 121, pp. 3–12 Fernandez, M. and K. C. Kim (1998). Self-employment rates of Asian immigrant groups: an analysis of intragroup and intergroup differences, International Migration Review, 32, pp. 654–81 Fichman, M. and D. A. Levinthal (1991). Honeymoons and the liability of adolescence: a new perspective on duration dependence in social and organisational relationships, Academy of Management Review, 16, pp. 442–68 Fielden, S. L., M. J. Davidson and P. J. Makin (2000). Barriers encountered during micro and small business start-up in North West England, Journal of Small Business and Enterprise Development, 7, pp. 295–304 Fischer, E. M., A. R. Reuber and L. S. Dyke (1993). A theoretical overview and extension of research on sex, gender and entrepreneurship, Journal of Business Venturing, 8, pp. 151–68 Fisman, R. and I. Love (2003). Trade credit, financial intermediary development and industry growth, Journal of Finance, 58, pp. 353–74 Florida, R. (2003). Entrepreneurship, creativity, and regional economic growth, in D. M. Hart (ed.), The Emergence of Entrepreneurship Policy, Cambridge, Cambridge University Press, pp. 39–58 Florin, J., M. Lubatkin and W. Schulze (2003). A social capital model of high growth ventures, Academy of Management Journal, 46(3), pp. 374–84 Flota, C. and M. T. Mora (2001). The earnings of self-employed Mexican-Americans along the US–Mexico border, Annals of Regional Science, 35, 483–99 Foellmi, R. and M. Oechslin (2007). Who gains from non-collusive corruption? Journal of Development Economics, 82(1), pp. 95–119 Fogel, K., A. Hawk, R. Morck and B. Yeung (2006). Institutional obstacles to entrepreneurship, in M. Casson, B. Yeung, A. Basu and N. Wadeson (eds.), The Oxford Handbook of Entrepreneurship, Oxford University Press, pp. 540–79 Fölster, S. (2002). Do lower taxes stimulate self-employment?, Small Business Economics, 19, pp. 135–45 Fonseca, R., P. Lopez-Garcia and C. A. Pissarides (2001). Entrepreneurship, start-up costs and employment, European Economic Review, 45, pp. 692–705 Forbes, D. P. (2005). Are some entrepreneurs more overconfident than others? Journal of Business Venturing, 20, pp. 623–40 Foreman-Peck, J. (1985). Seedcorn or chaff? New firm formation and the performance of the inter-war economy, Economic History Review, 38, pp. 402–22 Foreman-Peck, J., G. Makepeace and B. Morgan (2006). Growth and profitability of small- and medium-sized enterprises: Some Welsh evidence, Regional Studies, 40(4), pp. 307–19 Form, W. H. (1985). Divided We Stand: Working Class Stratification in America, Urbana IL, University of Illinois Press

496

References

Formby, J. P. and E. L. Millner (1985). The convergence of utility and profit maximisation, Southern Economic Journal, 51, pp. 1174–85 Foss, K., N. J. Foss, P. G. Klein and S. K. Klein (2007). The entrepreneurial organisation of heterogeneous capital, Journal of Management Studies, 44(7), pp. 1165–86 Foti, A. and M. Vivarelli (1994). An econometric test of the self-employment model: the case of Italy, Small Business Economics, 6, pp. 81–93 Frame, W. S., A. Srinivasan and L. Woosley (2001). The effect of credit scoring on small-business lending, Journal of Money, Credit and Banking, 33, pp. 813–25 Francois, P. (2003). Not-for-profit provision of public services, Economic Journal, 113, pp. C53–C61 Frank, M. Z. (1988). An intertemporal model of industrial exit, Quarterly Journal of Economics, 103, pp. 333–44 Fraser, S. (2007). Finance for SMEs: comparisons of ethnic minority and white owned businesses, Paper presented at Durham ESRC–SBS Seminar on Entrepreneurial Finance, Warwick Business School Fraser, S. and F. J. Greene (2006). The effects of experience on entrepreneurial optimism and uncertainty, Economica, 73, pp. 169–92 Fredland, J. E. and R. D. Little (1981). Self-employed workers: returns to education and training, Economics of Education Review, 1, pp. 315–37 Freel, M. S. (2007). Are small innovators credit rationed? Small Business Economics, 28, pp. 23–35 Freeman, J. and J. S. Engel (2007). Models of innovation: startups and mature corporations, California Management Review, 50(1), pp. 94–119 Freeman, J., G. R. Carroll and M. T. Hannan (1983). The liability of newness: age dependence in organisational death rates, American Sociological Review, 48, pp. 692–710 Fried, J. and P. Howitt (1980). Credit rationing and implicit contract theory, Journal of Money, Credit & Banking, 12, pp. 471–87 Freytag, A. and R. Thurik (2007). Entrepreneurship and its determinants in a cross-country setting, Journal of Evolutionary Economics, 17(2), pp. 117–31 Friedman, M. (1953). The Methodology of Positive Economics, Chicago IL, University of Chicago Press Fritsch, M. (1997). New firms and regional employment change, Small Business Economics, 9, pp. 437–48 Fritsch, M. and O. Falck (2007). New business formation by industry over space and time: a multidimensional analysis, Regional Studies, 41(2), pp. 157–72 Fritsch, M. and P. Mueller (2004). Effects of new business formation on regional development over time, Regional Studies, 38(8), pp. 961–75 (2007). The persistence of regional new business formation-activity over time – assessing the potential of policy promotion programs, Journal of Evolutionary Economics, 17, pp. 299–315 (2008). The effect of new business formation on regional development over time: the case of Germany, Small Business Economics, 30, pp. 15–29 Fritsch, M. and A. Weyh (2006). How large are the direct employment effects of new businesses? An empirical investigation for West Germany, Small Business Economics, 27, pp. 245–60 Frye, T. and A. Shleifer (1997). The invisible hand and the grabbing hand, American Economic Review, 87, pp. 354–58 Fuchs, V. R. (1982). Self-employment and labour-force participation of older males, Journal of Human Resources, 17, pp. 339–57 Fuest, C., B. Huber and S. B. Nielsen (2002). Why is the corporation tax lower than the personal tax rate? The role of small firms, Journal of Public Economics, 87, pp. 157–74 Fujii, E. T. and C. B. Hawley (1991). Empirical aspects of self-employment, Economics Letters, 36, pp. 323–29 Futia, C. (1980). Schumpeterian competition, Quarterly Journal of Economics, 93, pp. 675–95

References

497

Gabszewicz, J. J. and D. Laussel (2006). Increasing returns, entrepreneurship and imperfect competition, Economic Theory, 30, pp. 1–19 Galbraith, J. K. (2006). Some notes on entrepreneurship and the welfare state, Industrial & Corporate Change, 15(1), pp. 203–207 Gale, D. and M. Hellwig (1985). Incentive-compatible debt contracts: the one-period problem, Review of Economic Studies, 52, pp. 647–63 Gale, W. G. (1990a). Collateral, rationing, and government intervention in credit markets, in R. G. Hubbard (ed.), Asymmetric Information, Corporate Finance, and Investment, University of Chicago Press, pp. 43–61 (1990b). Federal lending and the market for credit, Journal of Public Economics, 42, pp. 177–93 (1991). Economic effects of federal credit programs, American Economic Review, 81, pp. 133–52 Gangopadhyay, S., M. Ghatak and R. Lensink (2005). Joint liability lending and the peer selection effect, Economic Journal, 115, pp. 1005–15 Gans, J. S., D. H. Hsu and S. Stern (2002). When does start-up innovation spur the gale of creative destruction? Rand Journal of Economics, 33, pp. 571–86 Garcia-Mainar, I. and V. M. Montuenga-Gomez (2005). Education returns of wage earners and selfemployed workers: Portugal vs. Spain, Economics of Education Review, 24, pp. 161–70 Garcia-Penalosa, C. and J.-F. Wen (2008). Redistribution and entrepreneurship with Schumpeterian growth, Journal of Economic Growth, 13, pp. 57–80 Garen, J. (1998). Self-employment, pay systems and the theory of the firm: an empirical analysis, Journal of Economic Behavior & Organization, 36(2), pp. 257–74 (2006). Use of employees and alternative work arrangements in the United States: a law, economics, and organisations perspective, Labour Economics, 13, pp. 107–41 Garicano, L. and E. Rossi-Hansberg (2006). The knowledge economy at the turn of the twentieth century: the emergence of hierarchies, Journal of the European Economic Association, 4(2–3), pp. 396–403 Garmaise, M. J. (2008). Production in entrepreneurial firms: the effects of financial constraints on labour and capital, Review of Financial Studies, 21, pp. 543–77 Gartner, W. B. (1988). Who is an entrepreneur? Is the wrong question, American Journal of Small Business, 12, pp. 11–32 Gartner, W. B. and S. A. Shane (1995). Measuring entrepreneurship over time, Journal of Business Venturing, 10, pp. 283–301 Gartner, W. B., K. G. Shaver, N. M. Carter and P. D. Reynolds (2004). Handbook of Entrepreneurial Dynamics, Thousand Oaks CA, Sage Publications Garvin, D. A. (1983). Spin-offs and the new firm formation process, California Management Review, 25(2), pp. 3–20 Gavron, R., M. Cowling, G. Holtham and A. Westall (1998). The Entrepreneurial Society, London, IPPR Gehrig, T. and R. Stenbacka (2007). Information sharing and lending market competition with switching costs and poaching, European Economic Review, 51, pp. 77–99 van Gelderen, M., N. Bosma and R. Thurik (2001). Setting up a business in the Netherlands: Who starts, who gives up, who is still trying, Unpublished report, http://ssrn.com/abstract= 370874 van Gelderen, M., R. Thurik and N. Bosma (2006). Success and risk factors in the pre-startup phase, Small Business Economics, 26, pp. 319–35 Gentry, W. M. and R. G. Hubbard (2000). Tax policy and entrepreneurial entry, American Economic Review, 90, pp. 283–87 (2001). Entrepreneurship and household saving, NBER Working Paper 7894, National Bureau of Economic Research

498

References

(2004). Entrepreneurship and household saving, BEP Advances in Economic Analysis & Policy, 4(1), Article 8, pp. 1–55 Georgellis, Y. and H. J. Wall (2000). What makes a region entrepreneurial? Evidence from Britain, Annals of Regional Science, 34, pp. 385–403 (2006). Entrepreneurship and the policy environment, Federal Reserve Bank of St Louis Review, 88(2), pp. 95–111 Georgellis, Y., J. G. Sessions and N. Tsitsianis (2005). Windfalls, wealth, and the transition to self-employment, Small Business Economics, 25, pp. 407–28 Geroski, P.A. (1995). What do we know about entry? International Journal of Industrial Organization, 13, pp. 450–56 (2003). The Evolution of New Markets, Oxford University Press Geroski, P. A. and R. Pomroy (1990). Innovation and the evolution of market structure, Journal of Industrial Economics, 38, pp. 299–314 Gersbach, H. and H. Uhlig (2007). On the coexistence of banks and markets, Scandinavian Journal of Economics, 109(2), pp. 225–43 Gersovitz, M. (1983). Savings and nutrition at low incomes, Journal of Political Economy, 91, pp. 841–55 Gervais, S. and I. Goldstein (2006). The positive effects of biased self-perception in firms, Mimeo, Fuqua School of Business, Duke University Ghatak, M. (1999). Group lending, local information, and peer selection, Journal of Development Economics, 60, pp. 27–50 (2000). Screening by the company you keep: joint liability lending and the peer selection effect, Economic Journal, 110, pp. 601–31 Ghatak, M. and T. W. Guinnane (1999). The economics of lending with joint liability: theory and practice, Journal of Development Economics, 60, pp. 195–228 Ghatak, M., M. Morelli and T. Sjöström (2001). Occupational choice and dynamic incentives, Review of Economic Studies, 68, pp. 781–810 (2007). Entrepreneurial income, occupational choice, and trickle-up policies, Journal of Economic Theory, 137, pp. 27–48 Ghosal, V. (2007). Small is beautiful but size matters: the asymmetric impact of uncertainty and sunk costs on small and large businesses, MPRA Working Paper No. 5461 Giannetti, M. (2003). Do better institutions mitigate agency problems? Evidence from corporate finance choices, Journal of Financial & Quantitative Analysis, 38(1), pp. 185–212 Giannetti, M. and A. Simonov (2004). On the determinants of entrepreneurial activity: social norms, economic environment and individual characteristics, Swedish Economic Policy Review, 11(2), 269–313 Gibrat, (1931). Les Inégalités Économique, Paris, Siley Gifford, S. (1998). The Allocation of Limited Entrepreneurial Attention, Boston MA, Kluwer Gilbert, R. and D. Newbery (1982). Pre-emptive patenting and the persistence of monopoly, American Economic Review, 72, pp. 514–26 Gilson, R. J. (1999). The legal infrastructure of high technology industrial districts: Silicon Valley, Route 128, and covenants not to compete, New York University Law Review, 74, pp. 575–629 Gimeno, J., T. B. Folta, A. C. Cooper and C. Y. Woo (1997). Survival of the fittest? Entrepreneurial human capital and the persistence of underperforming firms, Administrative Science Quarterly, 42, pp. 750–83 Gittelman, M. (2006). National institutions, public–private knowledge flows, and innovation performance: A comparative study of the biotechnology industry in the US and France, Research Policy, 35, pp. 1052–68

References

499

Gladstone, R. and J. Lane-Lee (1995). The operation of the insolvency system in the UK: some implications for entrepreneurialism, Small Business Economics, 7, pp. 55–67 Glaeser, E. L. (2007). Entrepreneurship and the city, NBER Working Paper No. 13551 Glaeser, E. L. and A. Shleifer (2001). Not-for-profit entrepreneurs, Journal of Public Economics, 81, pp. 99–115 Glaeser, E. L., D. Laibson and B. Sacerdote (2002). An economic approach to social capital, Economic Journal, 112, pp. F437–F458 Glazer, N. and D. P. Moynihan (1970). Beyond the Melting Pot: The Negroes, Puerto Ricans, Jews, Italians and Irish of New York City, Second Edition, Cambridge MA, MIT Press Godley, A. (2006). Migration of entrepreneurs, in M. Casson, B. Yeung, A. Basu and N. Wadeson (eds.), The Oxford Handbook of Entrepreneurship, Oxford University Press, pp. 601–10 Goedhuys, M. and L. Sleuwaegen (2000). Entrepreneurship and growth of entrepreneurial firms in Côte d’Ivoire, Journal of Development Studies, 36, pp. 122–45 Goeggel, K., J. Graeb and F. Pfeiffer (2007). Self-employment in 15 European countries 1991–2003, Jahrbucher fir Nationalokonomie und Statistik, 227(2), pp. 153–67 Golec, J. and M. Tamarkin (1998). Betters love skewness, not risk, at the horse track, Journal of Political Economy, 106, pp. 205–25 Gollin, D. (2008). Nobody’s business but my own: self-employment and small enterprise in economic development, Journal of Monetary Economics, 55, pp. 219–33 Gomez, R. and E. Santor (2001). Membership has its privileges: the effect of social capital and neighbourhood characteristics on the earnings of microfinance borrowers, Canadian Journal of Economics, 34, pp. 943–66 Gompers, P. A. (1995). Optimal investment, monitoring, and the staging of venture capital, Journal of Finance, 50, pp. 1461–89 (1996). Grandstanding in the venture capital industry, Journal of Financial Economics, 42, pp. 133–56 Gompers, P. A. and J. Lerner (1998). What drives venture capital fundraising? Brookings Papers on Economic Activity, 1998, pp. 149–92 (1999). The Venture Capital Cycle, Cambridge MA, MIT Press (2001). The venture capital revolution, Journal of Economic Perspectives, 15(2), pp. 145–68 (2003). Equity financing, in Z. J. Acs and D. B. Audretsch (eds.), Handbook of Entrepreneurship Research: An Interdisciplinary Survey and Introduction, Boston MA, Kluwer, pp. 267–98 Gompers, P., J. Lerner and D. Scharfstein (2005). Entrepreneurial spawning: public corporations and the genesis of new ventures, 1986 to 1999, Journal of Finance, 60, pp. 577–614 Gompers, P., A. Kovner, J. Lerner and D. Scharfstein (2006). Skill vs. luck in entrepreneurship and venture capital: Evidence from serial entrepreneurs, NBER Working Paper 12592, National Bureau of Economic Research Goodman, A. and S. Webb (1994). For richer, for poorer: the changing distribution of income in the UK, 1961–91, Fiscal Studies, 15, pp. 29–62 Gordon, R. H. (1998). Can high personal tax rates encourage entrepreneurial activity?, IMF Staff Papers, 45, pp. 49–80 Gordus, J. P., P. Jarley and L.A. Ferman (1981). Plant Closings and Economic Dislocation, Kalamazoo MI, W. E. Upjohn Institute for Employment Research Gorman, M. and W. Sahlman (1989). What do venture capitalists do? Journal of Business Venturing, 4(4), pp. 231–48 Gort, M. and S. Klepper (1982). Time paths in the diffusion of product innovations, Economic Journal, 92, pp. 630–53 Gramm, C. L. and J. F. Schnell (2001). The use of flexible staffing arrangements in core production jobs, Industrial and Labor Relations Review, 54(2), pp. 245–58

500

References

Grant, D. (1996). The political economy of new business formation across the American states, 1970–1985, Social Science Quarterly, 77, 28–42 Gray, J. A. and Y. Wu (1995). On equilibrium credit rationing and interest rates, Journal of Macroeconomics, 17, pp. 405–20 Green, R. (1984). Investment incentives, debt and warrants, Journal of Financial Economics, 13, pp. 115–36 Greenbaum, S., G. Kantas and I. Venezia (1989). Equilibrium loan pricing under the bank–client relationship, Journal of Banking and Finance, 13, pp. 221–35 Greene, F. J, K. F. Mole and D. J. Storey (2004). Does more mean worse? Three decades of enterprise policy in the Tees Valley, Urban Studies, 41(7), pp. 1207–28 Greene, P. G., C. G. Brush and E. J. Gatewood (2007). Perspectives on women entrepreneurs: past findings and new directions, in M. Minniti (ed.) Entrepreneurship: The Engine of Growth, Volume 1, Westport CT, Praeger, pp. 181–204 Greene, W. H. (2003). Econometric Analysis, 5th Edition, New Jersey, Prentice-Hall Greenwald, B., J. E. Stiglitz and A. Weiss (1984). Informational imperfections in the capital market and macroeconomic fluctuations, American Economic Review, 74, pp. 194–99 Griliches, Z. (1992). The search for R & D spillovers, Scandinavian Journal of Economics, 94, pp. 29–47 Grilo, I. and J.-M. Irigoyen (2006). Entrepreneurship in the EU: to wish and not to be, Small Business Economics, 26(4), pp. 305–18 Grilo I. and A. R. Thurik (2006). Entrepreneurship in the old and the new Europe, in E. Santarelli (ed.), Entrepreneurship, Growth and Innovation, Berlin, Springer Science, pp. 75–103 Gromb, D. and D. Scharfstein (2002). Entrepreneurship in equilibrium, NBER Working Paper 9001, NBER, Cambridge MA Gropp, R., J. K. Scholz and M. J. White (1997). Personal bankruptcy and credit supply and demand, Quarterly Journal of Economics, 112, pp. 217–51 Grossman, G. M. (1984). International trade, foreign investment, and the formation of the entrepreneurial class, American Economic Review, 74, pp. 605–14 Grossman, G. M. and O. Hart (1986). The costs and benefits of ownership: a theory of vertical and lateral integration, Journal of Political Economy, 94, pp. 691–719 Grossman, G. M. and E. Helpman (1991). Growth and welfare in a small open economy, in E. Helpman and A. Razin (eds.), International Trade and Trade Policy, Cambridge MA, MIT Press, pp. 141–63 (2002). Integration versus outsourcing in industry equilibrium, Quarterly Journal of Economics, 117, pp. 85–120 Grossman, G. M. and C. Shapiro (1984). Informative advertising with differentiated products, Review of Economic Studies, 51, pp. 63–82 Gruber, J. and J. Poterba (1994). Tax incentives and the decision to purchase health insurance: evidence from the self-employed, Quarterly Journal of Economics, 109, pp. 701–33 Grüner, H. P. and R. Schils (2007). The political economy of wealth and interest, Economic Journal, 117, pp. 1403–22 Guariglia, A. and B. Y. Kim (2006). The dynamics of moonlighting in Russia, Economics of Transition, 14(1), 1–45 Guesnier, B. (1994). Regional variations in new firm formation in France, Regional Studies, 28, pp. 347–58 Guinnane, T. (1994). A failed institutional transplant: Raiffeisen’s credit co-operatives in Ireland, 1894–1914, Explorations in Economic History, 31, pp. 38–61 Guiso, L. (1998). High-tech firms and credit rationing, Journal of Economic Behavior & Organization, 35, pp. 39–59

References

501

Guiso, L. and F. Schivardi (2005). Learning to be an entrepreneur, Discussion Paper No. 5290, Centre for Economic Policy Research Guiso, L., P. Sapienza and L. Zingales (2004). Does local financial development matter? Quarterly Journal of Economics, 119, pp. 929–69 Gurley-Calvez, T. and D. Bruce (2008). Do tax cuts promote entrepreneurial longevity? National Tax Journal, 61(2), pp. 225–50 Gustaffson, B. and U. Kjulin (1994). Time use in child care and housework and the total cost of children, Journal of Population Economics, 7, pp. 287–306 Haber, S. and A. Reichel (2007). The cumulative nature of the entrepreneurial process: the contribution of human capital, planning and environment resources to small venture performance, Journal of Business Venturing, 22, pp. 119–45 Haber, S. E., E. J. Lamas and J. Lichtenstein (1987). On their own: the self-employed and others in private business, Monthly Labor Review, 110, pp. 17–23 Hagen, K. P. and J. G. Sannarnes (2007). Taxation of uncertain business profits, provate risk markets and optimal allocation of risk, Journal of Public Economics, 91, pp. 1507–17 Hakim, C. (1988). Self-employment in Britain: recent trends and current issues, Work, Employment & Society, 2, pp. 421–50 (1989a). New recruits to self-employment in the 1980s, Employment Gazette, 97, pp. 286–97 (1989b). Workforce restructuring, social insurance coverage and the black economy, Journal of Social Policy, 18, pp. 471–503 Halaby, C. N. (2003). Where job values come from: family and schooling background, cognitive ability, and gender, American Sociological Review, 68, pp. 251–78 Hall, B. H. (1987). The relationship between firm size and firm growth in the US manufacturing sector, Journal of Industrial Economics, 35, pp. 583–606 Hall, G. (1992). Reasons for insolvency amongst small firms – a review and fresh evidence, Small Business Economics, 4, pp. 237–50 Haltiwanger, J. (2006). Entrepreneurship and job growth, in D. B. Audretsch, R. Strom and Z. Acs (eds.), Entrepreneurship, Growth and Public Policy, Cambridge University Press Haltiwanger, J. and C. J. Krizan (1999). Small business and job creation in the United States: the role of new and young businesses, in Z. J. Acs (ed.), Are Small Firms Important? Their Role and Impact, Dordrecht, Kluwer Hamermesh, D. S. (1993). Labour Demand, Princeton University Press Hamilton, B. H. (2000). Does entrepreneurship pay? An empirical analysis of the returns to selfemployment, Journal of Political Economy, 108, pp. 604–31 Hamilton, R. T. (1989). Unemployment and business formation rates: reconciling time series and cross-section evidence, Environment and Planning C, 21, pp. 249–55 Hammarstedt, M. (2001). Immigrant self-employment in Sweden – its variation and some possible determinants, Entrepreneurship & Regional Development, 13, pp. 147–61 (2006). The predicted earnings differential and immigrant self-employment in Sweden, Applied Economics, 38, pp. 619–30 Hanley, A. and J. Crook (2005). The higher cost of follow-up loans, Small Business Economics, 24, pp. 29–38 Hannan, M. T., J. N. Barron and M. D. Burton (1996). Inertia and change in the early years: employment relations in young, high-technology firms, Industrial & Corporate Change, 5, pp. 503–35 Hansmann, H. (1987). Economic theories of non-profit organisation, in W. Powell (ed.), The Nonprofit Sector, New Haven CT, Yale University Press Hao, K. Y. andA. B. Jaffe (1993). Effect of liquidity in firms’R & D spending, Economics of Innovation & New Technology, 2, pp. 275–82 Harada, N. and K. Kijima (2005). Consumption-leisure preference structure: a new explanation of the Evans–Jovanovic results for entrepreneurial choice, Small Business Economics, 24, pp. 187–91

502

References

Harding, R. and M. Cowling (2004). Social Entrepreneurship Monitor United Kingdom, London Business School, www.gem.org Harhoff, D. and T. Körting (1998). Lending relationships in Germany – empirical evidence from survey data, Journal of Banking and Finance, 22, pp. 1317–53 Harris, J. R. and M. P. Todaro (1970). Migration, unemployment and development: a two-sector analysis, American Economic Review, 60, pp. 126–42 Harrison, B. (1994). The small firm myth, California Management Review, 36, pp. 142–58 Harrison, B. and M. R. Kelley (1993). Outsourcing and the search for flexibility, Work, Employment & Society, 7, pp. 213–35 Harrison, R. T. and C. M. Mason (2007). Does gender matter? Women business angels and the supply of entrepreneurial finance, Entrepreneurship Theory & Practice, 31(3), pp. 445–72 Harrison, R. T., C. M. Mason and P. Girling (2004). Financial bootstrapping and venture development in the software industry, Entrepreneurship & Regional Development, 16, pp. 307–33 Hart, D. M. (ed.) (2003). The Emergence of Entrepreneurship Policy, Cambridge University Press Hart, O. (1995). Firms, Contracts and Financial Structure, Oxford, Clarendon Press Hart, O. and J. Moore (1994). A theory of debt based on inalienability of human capital, Quarterly Journal of Economics, 109, pp. 841–79 (1998). Default and renegotiation: a dynamic model of debt, Quarterly Journal of Economics, 113, pp. 1–26 Hart, P. E. and N. Oulton (1996). Growth and size of firms, Economic Journal, 106, pp. 1242–52 Hart, P. E. and P. E. Prais (1956). The analysis of business concentration: a statistical approach, Journal of the Royal Statistical Society, 119, pp. 150–91 Hartarska, V. and C. Gonzales-Vega (2006). What affects new and established firms’ expansion? Evidence from small firms in Russia, Small Business Economics, 27, pp. 195–206 Harvey, M. (1995). Towards the insecurity society: the tax trap of self-employment, London, The Institute of Employment Rights Hasan, I. and H. Wang (2008). The US bankruptcy law and private equity financing: empirical evidence, Small Business Economics, 31(1), pp. 5–19 Hasle, P. and H. J. Limborg (2006). A review of the literature on preventive occupational health and safety activities in small enterprises, Industrial Health, 44(1), pp. 6–12 Haugh, H. (2006). Nonprofit social entrepreneurship, in S. C. Parker (ed.), The Life Cycle of Entrepreneurial Ventures, New York, Springer, pp. 401–36 Hause, J. C. and G. Du Rietz (1984). Entry, industry growth, and the microdynamics of industry supply, Journal of Political Economy, 92(4), pp. 733–57 Hawley, F. B. (1907). Enterprise and the Productive Process, G. P. Putnam’s Sons, New York Haynes, G. W. and D. C. Haynes (1999). The debt structure of small businesses owned by women in 1987 and 1993, Journal of Small Business Management, 37(2), pp. 1–19 Hayward, M. L. A., D. A. Shepherd and D. Griffin (2006). A hubris theory of entrepreneurship, Management Science, 52(2), pp. 160–72 Headd, B. (2003). Redefining business success: distinguishing between closure and failure, Small Business Economics, 21, pp. 51–61 Headen, A. E. (1990). Wage returns to ownership, and fee responses to physician supply, Review of Economics and Statistics, 72, pp. 30–37 Heaton, J. and D. Lucas (2000). Portfolio choice and asset prices: the importance of entrepreneurial risk, Journal of Finance, 55, pp. 1163–98 (2004). Capital structure, hurdle rates, and portfolio choice – interactions in an entrepreneurial firm, unpublished manuscript, University of Chicago Hébert, R. F. and A. N. Link (2006). Historical perspectives on the entrepreneur, Foundations and Trends in Entrepreneurship, 2(4), pp. 261–408 Heckman, J. J. (1979). Sample selection bias as a specification error, Econometrica, 47, pp. 475–92

References

503

Helfat, C. E. and M. B. Lieberman (2002). The birth of capabilities: market entry and the importance of pre-history, Industrial & Corporate Change, 11, pp. 725–60 Hellmann, T. (1998a). A theory of corporate venture investing, Working Paper, Stanford University (1998b). The allocation of control rights in venture capital contracts, Rand Journal of Economics, 29, pp. 57–76 (2002). A theory of strategic venture investing, Journal of Financial Economics, 64, pp. 285–314 (2006). IPOs, acquisitions, and the use of convertible securities in venture capital, Journal of Financial Economics, 81(3), pp. 649–79 (2007a). Entrepreneurs and the process of obtaining resources, Journal of Economics and Management Strategy, 16(1), pp. 81–109 (2007b). When do employees become entrepreneurs? Management Science, 53(6), pp. 919–33 Hellmann, T. and M. Puri (2000). The interaction between product market and financing strategy: the role of venture capital, Review of Financial Studies, 13, pp. 959–84 (2002). Venture capital and the professionalisation of start-ups: empirical evidence, Journal of Finance, 57, pp. 169–97 Hellmann, T. and J. Stiglitz (2000). Credit and equity rationing in markets with adverse selection, European Economic Review, 44, pp. 281–304 Hellwig, M. (1987). Some recent developments in the theory of competitive markets with adverse selection, European Economic Review, 31, pp. 319–25 Henderson, R. (1993). Underinvestment and incompetence as responses to radical innovation – evidence from the photolithographic alignment equipment industry, Rand Journal of Economics, 24(2), pp. 248–70 Henderson, R. and K. B. Clark (1990).Architectural innovation: the reconfiguration of existing product technologies and the failure of established companies, Administrative Science Quarterly, 35, pp. 9–20 Henley, A. (2004). Self-employment status: the role of state dependence and initial circumstances, Small Business Economics, 22, pp. 67–82 (2005). Job creation by the self-employed: the roles of entrepreneurial and financial capital, Small Business Economics, 25, pp. 175–96 (2007). Entrepreneurial aspiration and transition into self-employment: evidence from British longitudinal data, Entrepreneurship & Regional Development, 19(3), pp. 253–80 Henrekson, M. (2005). Entrepreneurship: a weak link in the welfare state? Industrial and Corporate Change, 14, pp. 437–67 (2006). Entrepreneurship and the welfare state: a reply, Industrial and Corporate Change, 15(3), pp. 579–93 Hermes, N. and R. Lensink (2007). The empirics of microfinance: what do we know? Economic Journal, 117, pp. F1–F10 Highfield, R. and R. Smiley (1987). New business starts and economic activity: an empirical investigation, International Journal of Industrial Organization, 5, pp. 51–66 Hildebrand, V. A. and D. R. Williams (2003). Self-employment and caring for children: evidence from Europe, IRISS Working Paper 2003–06 Hillier, B. (1998). The borrower’s curse: comment, Economic Journal, 108, pp. 1772–74 Hillier, B. and M. V. Ibrahimo (1992). The performance of credit markets under asymmetric information about projects means and variances, Journal of Economic Studies, 19, pp. 3–17 (1993). Asymmetric information and models of credit rationing, Bulletin of Economic Research, 45, pp. 271–304 Himmelberg, C. P. and B. C. Petersen (1994). R & D and internal finance: a panel study of small firms in high-tech industries, Review of Economics and Statistics, 76, pp. 38–51

504

References

Hintermaier, T. and T. Steinberger (2005). Occupational choice and the private equity premium puzzle, Journal of Economic Dynamics and Control, 29, pp. 1765–83 Hinz, T. and M. Jungbauer-Gans (1999). Starting a business after unemployment: characteristics and chances of success, Entrepreneurship & Regional Development, 11(4), pp. 317–33 Hippel, E. von (1994). Sticky information and the locus of problem solving, Management Science, 40, pp. 429–39 Hjerm, M. (2004). Immigrant entrepreneurship in the Swedish welfare state, Sociology, 38, pp. 739–56 HMSO (1971). Report on the Committee of Inquiry on Small Firms (Bolton Report), Cmnd 4811, HMSO, London (1979). Interim Report on the Financing of Small Firms (Wilson Report), Cmnd 7503, HMSO, London (1991). Constraints on the Growth of Small Firms, HMSO, London Ho., Y.-P. and P.-K. Wong (2007). Financing, regulatory costs and entrepreneurial propensity, Small Business Economics, 28(2–3), pp. 187–204 Hochberg, Y. (2002). Venture capital and corporate governance in the newly public firm, Mimeo, Stanford University Hochguertel, S. (2005a). The dynamics of self-employment and household wealth: new evidence from panel data, Mimeo, Free University of Amsterdam (2005b). Self-employment around retirement age in Europe, Mimeo, Free University of Amsterdam Hoff, K. and J. E. Stiglitz (1990). Imperfect information and rural credit markets: puzzles and policy perspectives, World Bank Economic Review, 4, pp. 235–51 Holbrook, D., W. M. Cohen, D. A. Hounshell and S. Klepper (2000). The nature, sources, and consequences of firm differences in the early history of the semiconductor industry, Strategic Management Journal, 21(10–11), pp. 1017–41 Holmes, T. (1999). Localisation of industry and vertical disintegration, Review of Economics and Statistics, 81, pp. 314–25 Holmes, T. J. and J. A. Schmitz (1990). A theory of entrepreneurship and its application to the study of business transfers, Journal of Political Economy, 98, pp. 265–94 (1996). Managerial tenure, business age, and small business turnover, Journal of Labor Economics, 14, pp. 79–99 (2001). A gain from trade: from unproductive to productive entrepreneurship, Journal of Monetary Economics, 47, pp. 417–46 Holmstrom, B. (1989). Agency costs and innovation, Journal of Economic Behavior & Organization, 12, pp. 305–27 Holmstrom, B. and P. Milgrom (1994). The firm as an incentive system, American Economic Review, 84(4), pp. 972–91 Holtz-Eakin, D. (1999). The death tax: investments, employment and entrepreneurs, Tax Notes, 84 (August 2) (2000). Public policy toward entrepreneurship, Small Business Economics, 15, pp. 283–91 Holtz-Eakin, D. and C. Kao (2003). Entrepreneurship and economic growth: the proof is in the productivity, Working Paper 50, Center for Policy Research, Syracuse University Holtz-Eakin, D., D. Joulfaian and H. S. Rosen (1994a). Entrepreneurial decisions and liquidity constraints, Rand Journal of Economics, 25, pp. 334–47 (1994b). Sticking it out: entrepreneurial survival and liquidity constraints, Journal of Political Economy, 102, pp. 53–75 Holtz-Eakin, D., J. Penrod and H. S. Rosen (1996). Health insurance and the supply of entrepreneurs, Journal of Public Economics, 62, pp. 209–35 Holtz-Eakin, D., J. W. R. Phillips and H. S. Rosen (2001). Estate taxes, life insurance, and small business, Review of Economics and Statistics, 83(1), pp. 52–63

References

505

Holtz-Eakin, D., H. S. Rosen and R. Weathers (2000). Horatio Alger meets the mobility tables, Small Business Economics, 14, pp. 243–74 Honig, B. (1996). Education and self-employment in Jamaica, Comparative Education Review, 40, pp. 177–93 (1998). What determines success? Examining the human, financial, and social capital of Jamaican microentrepreneurs, Journal of Business Venturing, 13, pp. 371–94 Honig, M. and G. Hanoch (1985). Partial retirement as a separate mode of retirement behaviour, Journal of Human Resources, 20, pp. 21–46 Honjo, Y. (2000a). Business failure of new software firms, Applied Economics Letters, 7, pp. 575–79 (2000b). Business failure of new firms: an empirical analysis using a multiplicative hazards model, International Journal of Industrial Organization, 18, pp. 557–74 Honkapohja, S. and A. Turunen-Red (2007). Increasing returns, learning, and beneficial tax competition, Journal of Public Economic Theory, 9(6), pp. 927–58 Hopenhayn, H. A. (1992). Entry, exit and firm dynamics in long run equilibrium, Econometrica, 60, pp. 1127–50 Hopenhayn, H. A. and G. Vereshchagina (2003). Risk taking by entrepreneurs, Working Paper No. 500, Rochester Center for Economic Research, University of Rochester Hout, M. and H. S. Rosen (2000). Self-employment, family background, and race, Journal of Human Resources, 35, pp. 670–92 Horvath, M. T. K. and M. J. Woywode (2005). Entrepreneurs and the choice of limited liability, Journal of Institutional & Theoretical Economics, 161, pp. 681–707 Horvitz, P. M. (1984). Problems in the financing of small business: introduction and summary, in P. M. Horvitz and R. R. Pettit (eds.), Small Business Finance: Sources of Financing for Small Business, Part B, Greenwich CT, JAI Press, pp. 3–18 Howitt, P. and P. Aghion (1998). Capital accumulation and innovation as complementary factors in long-run growth, Journal of Economic Growth, 3, pp. 111–30 Hsu, D. H. (2004). What do entrepreneurs pay for venture capital affiliation? Journal of Finance, 59, pp. 1805–44 Hsu, D. H., E. B. Roberts and C. E. Eesley (2007). Entrepreneurs from technology-based universities: evidence from MIT, Research Policy, 36, pp. 768–88 Hubbard, R. G. (1998). Capital-market imperfections and investment, Journal of Economic Literature, 36, pp. 193–225 Hubbard, R. G., K. N. Kuttner and D. N. Palia (2002). Are there bank effects in borrowers’ costs of funds? Evidence from a matched sample of borrowers and banks, Journal of Business, 75, pp. 559–81 Hubbard, T. N. (2001). Contractual form and market thickness in trucking, Rand Journal of Economics, 32, pp. 369–86 Hudson, J. (1987). The age, regional and industrial structure of company liquidations, Journal of Business, Finance & Accounting, 14, pp. 199–213 (1989). The birth and death of firms, Quarterly Review of Economics & Business, 29, pp. 68–86 Hudson, J. and K. Cuthbertson (1993). The determinants of bankruptcies in the UK: 1971–1988, Manchester School, 61, pp. 65–81 Hughes, A. (1992). The problems of finance for smaller businesses, in N. Dimsdale and M. Prevezer (eds.), Capital Markets and Company Success, Oxford University Press Hulme, D. and P. Mosley (1996). Finance Against Poverty, London, Routledge Hundley, G. (2001a). Why women earn less than men in self-employment, Journal of Labor Research, 22, pp. 817–29 (2001b). Why and when are the self-employed more satisfied with their work? Industrial Relations, 40, pp. 293–317

506

References

(2006). Family background and the propensity for self-employment, Industrial Relations, 45(3), pp. 377–92 Huppi, M. and G. Feder (1990). The role of groups and credit co-operatives in rural lending, The World Bank Research Observer, 5, pp. 187–204 Hurst, E. and A. Lusardi (2004). Liquidity constraints, household wealth, and entrepreneurship, Journal of Political Economy, 112, pp. 319–47 Huyghebaert, N., L. M. van de Gucht and C. van Hulle (2006). The choice between bank debt and trade credit in business start-ups: financing costs versus liquidation policy, Small Business Economics, 29(4), pp. 435–52 Hyytinen, A. and P. Ilmakunnas (2007). What distinguishes a serial entrepreneur? Industrial & Corporate Change, 16(5), pp. 793–821 Hyytinen, A. and M. Maliranta (2008). When do employees leave their job for entrepreneurship? Scandinavian Journal of Economics, 110(1), pp. 1–21 Hyytinen, A. and P. Rouvinen (2008). The labour market consequences of self-employment spells: European evidence, Labour Economics, 15, pp. 246–71 Hyytinen, A. and O.-P. Ruuskanen (2007). Time use of the self-employed, Kyklos, 60(1), pp. 105–22 Hyytinen, A. and L. Väänänen (2006). Where do financial constraints originate from? An empirical analysis of adverse selection and moral hazard in capital markets, Small Business Economics, 27, pp. 323–48 Iams, H. M. (1987). Jobs of persons working after receiving retired worker benefits, Social Security Bulletin, 50, pp. 4–19 Ijiri, Y. and H. A. Simon (1977). Skew Distributions and the Sizes of Business Firms, Amsterdam, North-Holland Ilahi, N. (1999). Return migration and occupational change, Review of Development Economics, 3(2), pp. 170–86 Ilmakunnas, P. and V. Kanniainen (2001). Entrepreneurship, economic risks, and risk insurance in the welfare state: results with OECD data 1978–93, German Economic Review, 2, pp. 195–218 Inderst, R. and H. M. Müller (2004). The effect of capital market characteristics on the value of start-up firms, Journal of Financial Economics, 72(2), pp. 319–56 Inderst, R., H. M. Müller and F. Münnich (2007). Financing a portfolio of projects, Review of Financial Studies, 20(4), pp. 1289–1325 Innes, R. (1991). Investment and government intervention in credit markets when there is asymmetric information, Journal of Public Economics, 46, pp. 347–81 (1992). Adverse selection, investment, and profit taxation, European Economic Review, 36, pp. 1427–52 (1993). Financial contracting under risk neutrality, limited liability and ex ante asymmetric information, Economica, 60, pp. 27–40 Iyigun, M. F. and A. L. Owen (1998). Risk, entrepreneurship and human capital accumulation, American Economic Review, Papers and Proceedings, 88, pp. 454–57 (1999). Entrepreneurs, professionals and growth, Journal of Economic Growth, 4, pp. 213–32 Jackson, J. and G. Rodney (1994). The attitudinal climate for entrepreneurial activity, Public Opinion Quarterly, 58, pp. 358–80 Jackson, L. F. (1984). Hierarchic demand and the Engle curve for variety, Review of Economics and Statistics, 66, pp. 8–15 Jackson, T. (1998). Inside Intel, New York, Penguin Jacoby, H. G. (1993). Shadow wages and peasant family labour supply – an econometric application to the Peruvian Sierra, Review of Economic Studies, 60, pp. 903–21 Jaffe, A. B. (1989). Real effects of academic research, American Economic Review, 79, pp. 957–70

References

507

Jaffe, A. B., M. Trajtenberg and R. Henderson (1993). Geographic localisation of knowledge spillovers as evidenced by patent citations, Quarterly Journal of Economics, 63, pp. 577–98 Jaffee, D. and J. E. Stiglitz (1990). Credit rationing, in The Handbook of Monetary Economics, Vol. II (eds. B. M. Friedman and F. H. Hahn), Amsterdam, North-Holland, pp. 838–88 Jain, B. and O. Kini (1995). Venture capital participation and the post-issue operating performance of IPO firms, Managerial and Decision Economics, 16, pp. 593–606 Jamal, M. (2007). Burnout and self-employment: a cross-country empirical study, Stress and Health, 23(4), pp. 249–56 James, E. (1998). Commercialisation among nonprofits: objectives, opportunities, and constraints, in B. A. Weisbrod (ed.), To Profit or Not to Profit, Cambridge University Press, pp. 271–85 Jappelli, T., M. Pagano and M. Bianco (2005). Courts and banks: Effects of judicial enforcement on credit markets, Journal of Money, Credit and Banking, 37(2), pp. 223–44 Jefferson, P. N. (1997). Unemployment and financial constraints faced by small firms, Economic Inquiry, 35, pp. 108–19 Jeng, L. A. and P. C. Wells (2000). The determinants of venture capital funding: evidence across countries, Journal of Corporate Finance, 6, pp. 241–89 Jenkins, S. P. (1995). Accounting for income inequality trends: decomposition analyses for the UK, 1971–1986, Economica, 62, pp. 29–63 Jianakoplos, A. A. and A. Bernasek (1998). Are women more risk averse? Economic Inquiry, 36(4), pp. 620–30 Jin, Y. and Z. Zeng (2007). Real estate and optimal public policy in a credit-constrained economy, Journal of Housing Economics, 16, pp. 143–66 Johansson, D. (2005). The turnover of firms and industry growth, Small Business Economics, 24(5), pp. 487–95 Johansson, E. (2000). Self-employment and liquidity constraints: evidence from Finland, Scandinavian Journal of Economics, 102, pp. 123–34 Johnson, P. (1981). Unemployment and self-employment: a survey, Industrial Relations Journal, 12, pp. 5–15 Johnson, P. and S. Parker (1994). The interrelationships between births and deaths, Small Business Economics, 6, pp. 283–90 (1996). Spatial variations in the determinants and effects of firm births and deaths, Regional Studies, 30, pp. 679–88 Johnson, P. and J. Rodger (1983). From redundancy to self-employment, Employment Gazette, 91, pp. 260–64 Johnson, P., S. C. Parker and F. Wijbenga (2006). Nascent entrepreneurship research: achievements and opportunities, Small Business Economics, 27(1), pp. 1–4 Johnson, S., J. McMillan and C. Woodruff (2000). Entrepreneurs and the ordering of institutional reform, Economics of Transition, 8, pp. 1–36 (2002). Property rights and finance, American Economic Review, 92, pp. 1335–56 Jones, T., D. McEvoy and G. Barrett (1994). Raising capital for the ethnic minority small firm, in A. Hughes and D. Storey (eds.), Finance and the Small Firm, London, Routledge, pp. 145–81 Joulfaian, D. and M. Rider (1996). Tax evasion in the presence of negative income tax rates, National Tax Journal, 49, pp. 553–70 (1998). Differential taxation and tax evasion by small business, National Tax Journal, 51, pp. 675–87 Jovanovic, B. (1982). Selection and the evolution of industry, Econometrica, 50, pp. 649–70 (1994). Firm formation with heterogeneous management and labour skills, Small Business Economics, 6, pp. 185–91 (2001). New technology and the small firms, Small Business Economics, 16(1), pp. 53–55

508

References

Jovanovic, B. and S. Lach (1989). Entry, exit, and diffusion with learning by doing, American Economic Review, 79, pp. 690–99 Jovanovic, B. and G. M. MacDonald (1994). The life cycle of a competitive industry, Journal of Political Economy, 102(2), pp. 322–47 Jovanovic, B. and Y. Nyarko (1996). Learning by doing and the choice of technology, Econometrica, 64(6), pp. 1299–1310 Jung, Y.H., A. Snow and G. A. Trandel (1994). Tax evasion and the size of the underground economy, Journal of Public Economics, 54, pp. 391–402 Jurik, N. (1998). Getting away and getting by: the experiences of self-employed homemakers, Work and Occupations, 25(11), pp. 7–35 Kalleberg, A. L. and K. T. Leicht (1991). Gender and organisational performance: determinants of small business survival and success, Academy of Management Journal, 34, pp. 136–61 Kalnins, A. and W. Chung (2006). Social capital, geography, and survival: Gujarati immigrant entrepreneurs in the US lodging industry, Management Science, 52(2), pp. 233–47 Kan, K. and W.-D. Tsai (2006). Entrepreneurship and risk aversion, Small Business Economics, 26, pp. 465–74 Kanbur, S. M. (1979). Of risk taking and the personal distribution of income, Journal of Political Economy, 87, pp. 769–97 (1980). A note on risk taking, entrepreneurship, and Schumpeter, History of Political Economy, 12, pp. 489–98 (1981). Risk taking and taxation: an alternative perspective, Journal of Public Economics, 15, pp. 163–84 Kangasharju, A. (2000). Regional variations in firm formation: panel and cross-sectional data evidence from Finland, Papers in Regional Science, 79, pp. 355–73 Kangasharju, A. and A. Moisio (1998). Births–deaths nexus of firms: estimating VAR with panel data, Small Business Economics, 11, pp. 303–13 Kangasharju, A. and S. Pekkala (2002). The role of education in self-employment success in Finland, Growth and Change, 33, pp. 216–37 Kanniainen, V. and C. Keuschnigg (2003). The optimal portfolio of start-up firms in venture capital finance, Journal of Corporate Finance, 9, pp. 521–34 (2004). Start-up investment with scarce venture capital support, Journal of Banking and Finance, 28, pp. 1935–59 Kanniainen, V. and T. Vesala (2005). Entrepreneurship and labour market institutions, Economic Modelling, 22, pp. 828–47 Kanniainen, V., S. Kari and J. Ylä-Liedenpohja (2007). Nordic dual income taxation of entrepreneurs, International Tax and Public Finance, 14(4), pp. 407–26 Kaplan, S. and P. Stromberg (2001). Venture capitalists as principals: contracting, screening and monitoring, American Economic Review, Papers and Proceedings, 91(2), pp. 426–30 (2003). Financial contracting meets the real world: an empirical analysis of venture capital contracts, Review of Economic Studies, 70(2), pp. 281–316 Kari, S. and H. Karikallio (2007). Tax treatment of dividends and capital gains and the dividend decision under dual income tax, International Tax and Public Finance, 14(4), pp. 427–56 Karlan, D. S. (2007). Social connections and group banking, Economic Journal, 117, F52–F84 Karoly, L. A. and J. M. Zissimopoulos (2004). Self-employment among older US workers, Monthly Labor Review, 127(7), pp. 24–47 Katz, J. A. (1990). Longitudinal analysis of self-employment follow-through, Entrepreneurship & Regional Development, 2, pp. 15–25 (1993). How satisfied are the self-employed? A secondary analysis approach, Entrepreneurship Theory & Practice, 17, pp. 35–51

References

509

Kaufmann, P. J. (1999). Franchising and the choice of self-employment, Journal of Business Venturing, 14(4), pp. 345–62 Kaufmann, P. J. and R. P. Dant (1998). Franchising and the domain of entrepreneurship research, Journal of Business Venturing, 14, pp. 5–16 Kaufmann, P. J. and F. Lafontaine (1994). Costs of control: the source of economic rents for McDonald’s franchisees, Journal of Law & Economics, 37, pp. 417–53 Kawaguchi, D. (2003). Human capital accumulation of salaried and self-employed workers, Labour Economics, 10, pp. 55–71 (2008). Self-employment rents: evidence from job satisfaction scores, Hitotsubashi Journal of Economics, 49(1), pp. 35–45 Keasy, K. and R. Watson (1995). The pricing of small firm bank finance: evidence from the UK, Applied Economics Letters, 2, pp. 208–10 Keeble, D. and S. Walker (1994). New firms, small firms, and dead firms: spatial patterns and determinants in the United Kingdom, Regional Studies, 28, pp. 411–27 Keeble, D., S. Walker and M. Robson (1993). New Firm Formation and Small Business Growth: Spatial and Temporal Variations and Determinants in the United Kingdom, Employment Department Research Series, 15, London, HMSO Keeton, W. R. (1979). Equilibrium Credit Rationing, New York, Garland Publishing Kelly, D. and T. L. Amburgey (1991). Organizational inertia and momentum: a dynamic model of strategic change, Academy of Management Journal, 34 (1), pp. 591–612 Kent, C. A. and F. W. Rushing (1999). Coverage of entrepreneurship in principles of economics textbooks: an update, Journal of Economic Education, 30, pp. 184–88 Kepler, E. and S. Shane (2007). Are male and female entrepreneurs really that different? SBA Office of Advocacy Working Paper No. 309, Washington DC Kerins, F., J. K. Smith and R. Smith (2004). Opportunity costs of capital for venture capital investors and entrepreneurs, Journal of Financial & Quantitative Analysis, 39(2), pp. 385–406 Kesselman, J. R. (1989). Income tax evasion: an inter-sectoral analysis, Journal of Public Economics, 38, pp. 137–82 Kessler,A. and H. Frank (2004). Nascent entrepreneurship in longitudinal perspective, Paper presented at BKERC 2004, Strathclyde Kets de Vries, M. F. R. (1977). The entrepreneurial personality: a person at the crossroads, Journal of Management Studies, 14, pp. 34–57 Keuschnigg, C. (2004a). Venture capital backed growth, Journal of Economic Growth, 9, pp. 239–61 (2004b). Taxation of a venture capitalist with a portfolio of firms, Oxford Economic Papers, 56, pp. 285–306 Keuschnigg, C. and S. B. Nielsen (2001). Public policy for venture capital, International Tax and Public Finance, 8, pp. 557–72 (2003). Tax policy, venture capital and entrepreneurship, Journal of Public Economics, 87, pp. 175–203 (2004a). Start-ups, venture capitalists, and the capital gains tax, Journal of Public Economics, 88, pp. 1011–42 (2004b). Progressive taxation, moral hazard and entrepreneurship, Journal of Public Economic Theory, 6, pp. 471–90 (2005). Public policy for start-up entrepreneurship with venture capital and bank finance, in V. Kanniainen and C. Keuschnigg (eds.), Venture Capital, Entrepreneurship and Public Policy, Cambridge MA, MIT Press, pp. 221–50 (2006). Public policy, start-up entrepreneurship and the market for venture capital, in S. C. Parker (ed.), The Life Cycle of Entrepreneurial Ventures, New York, Springer, pp. 227–57

510

References

Kevane, M. and B. Wydick (2001). Micro enterprise lending to female entrepreneurs: sacrificing economic growth for poverty alleviation? World Development, 29, pp. 1225–36 Khan, B. Z. and K. L. Sokoloff (1993). ‘Schemes of practical utility’: entrepreneurship and innovation among ‘great inventors’ in the United States, 1790–1865, Journal of Economic History, 53, pp. 289–307 Khandker, S. R. (2005). Microfinance and poverty: evidence using panel data from Bangladesh, World Bank Economic Review, 19(2), pp. 263–86 Khanna, T. (2007). Billions of Entrepreneurs: How China and India Are Reshaping Their Futures, and Yours, Boston MA, Harvard Business School Press Kidd, M. P. (1993). Immigrant wage differentials and the role of self-employment in Australia, Australian Economic Papers, 32, pp. 92–115 Kihlstrom, R. E. and J. J. Laffont (1979). A general equilibrium entrepreneurial theory of firm formation based on risk aversion, Journal of Political Economy, 87, pp. 719–49 (1982). A competitive entrepreneurial model of the stock market, in J. McCall (ed.), The Economics of Information and Uncertainty, University of Chicago Press (1983). Implicit labour contracts and free entry, Quarterly Journal of Economics Supplement, 98, pp. 55–105 Kilby, P. (1983). An entrepreneurial problem, American Economic Review, Papers and Proceedings, 73, pp. 107–11 Kim, D. Y. (2006). Stepping-stone to intergenerational mobility? The springboard, safety net, or mobility trap functions of Korean immigrant entrepreneurship for the second generation, International Migration Review, 40(4), pp. 927–62 Kim, G. O. (2006). Do equally owned small businesses have equal access to credit? Small Business Economics, 27, pp. 369–86 Kim, J.-Y. and A. S. Miner (2007). Vicarious learning from the failures and near-failures of others: evidence from the US commercial banking industry, Academy of Management Journal, 50(2), pp. 687–714 Kim, P. H. and H. E. Aldrich (2005). Social capital and entrepreneurship, Foundations & Trends in Entrepreneurship, 1(2), pp. 55–104 Kim, P. H., H. E. Aldrich and L. A. Keister (2006). Access (not) denied: the impact of financial, human, and cultural capital on entrepreneurial entry in the United States, Small Business Economics, 27, pp. 5–22 King, A. and C. L. Tucci (2002). Incumbent entry into new market niches: the role of experience and managerial choice in the creation of dynamic capabilities, Management Science, 48(2), 171–86 King, R. G. and R. Levine (1993). Finance, entrepreneurship and growth, Journal of Monetary Economics, 32, pp. 513–42 King, S. R. (1986). Monetary transmission: through bank loans or bank liabilities?, Journal of Money, Credit and Banking, 18, pp. 290–303 Kirchhoff, B. A. and P. G. Greene (1998). Understanding the theoretical and empirical content of critiques of US job creation research, Small Business Economics, 10, pp. 153–69 Kirilenko, A. (2001). Valuation and control in venture finance, Journal of Finance, 56, pp. 565–88 Kirzner, I. (1973). Competition and Entrepreneurship, University of Chicago Press (1985). Discovery and the Capitalist Process, Chicago, University of Chicago Press (1997). Entrepreneurial discovery and the competitive market process: an Austrian approach, Journal of Economic Literature, 35, pp. 60–85 Kitao, S. (2008). Entrepreneurship, taxation and capital investment, Review of Economic Dynamics, 11(1), pp. 44–69 Klapper, L., L. Laeven and R. Rajan (2006). Entry regulation as a barrier to entrepreneurship, Journal of Financial Economics, 82, pp. 591–629

References

511

Klasa, S. (2007). Why do controlling families of public firms sell their remaining ownership stake? Journal of Financial & Quantitative Analysis, 42(2), pp. 339–67 Klein, P. G. and M. L. Cook (2006). T. W. Schultz and the human capital approach to entrepreneurship, Review of Agricultural Economics, 28(3), pp. 344–50 Klepper, S. (1996). Entry, exit, growth and innovation over the product life cycle, American Economic Review, 86, pp. 562–83 (2001). Employee start-ups in high-tech industries, Industrial & Corporate Change, 10, pp. 639–74 (2002). Firm survival and the evolution of oligopoly, Rand Journal of Economics, 33(1), pp. 37–61 (2006). Strategic disagreements, spinoffs, and the evolution of Detroit as the capital of the US automobile industry, Paper presented at the Max Planck Institute Workshop on Entrepreneurship and Growth, Ringberg Castle, Germany (2007). A shop floor view of growth in turn-of-the-century Cleveland, Capitalism and Society, Berkeley Electronic Press, 1(3), article 6 Klepper, S. and E. Graddy (1990). The evolution of new industries and the determinants of market structure, Rand Journal of Economics, 21(1), pp. 27–44 Klepper, S. and K. L. Simons (2000a). The making of an oligopoly: firm survival and technological change in the evolution of the US tire industry, Journal of Political Economy, 108(4), pp. 728–60 (2000b). Dominance by birthright: entry of prior radio producers and competitive ramifications in the US television receiver industry, Strategic Management Journal, 21, pp. 997–1016 Klepper, S. and S. Sleeper (2005). Entry by spinoffs, Management Science, 51, 1291–1306 Klepper, S. and P. Thompson (2006). Submarkets and the evolution of market structure, Rand Journal of Economics, 37, pp. 861–86 Klette, T. J., J. Moen and A. Griliches (2000). Do subsidies to commercial R & D reduce market failure? Microeconometric evaluation studies, Research Policy, 29, pp. 471–95 Knaup, A. E. (2005). Survival and longevity in the Business Employment Dynamics data, Monthly Labor Review, May, pp. 50–56 Knight, F. H. (1921). Risk, Uncertainty and Profit, New York, Houghton-Mifflin Knight, G. and S. McKay (2000). Lifetime experiences of self-employment, Labour Market Trends, 108, pp. 470–72 Knight, K. E. and T. Dorsey (1976). Capital problems in minority business development: a critical analysis, American Economic Review, Papers and Proceedings, 66, pp. 328–31 Knight, R. M. (1985). The financing of small high-technology firms in Canada, Journal of Small Business and Entrepreneurship, 3, pp. 5–17 Kochar, A. (1997). An empirical investigation of rationing constraints in rural credit markets in India, Journal of Development Economics, 53, pp. 339–71 Koeller, C T. and T. G. Lechler (2006). Employment growth in high-tech new firms, Journal of Labor Research, 27(2), pp. 135–47 Koellinger, P. (2008). Why are some entrepreneurs more innovative than others? Small Business Economics, 31(1), pp. 21–37 Koellinger, P. and M. Minniti (2006). Not for lack of trying: American entrepreneurship in black and white, Small Business Economics, 27, pp. 59–79 Koellinger, P., M. Minniti and C. Schade (2007). ‘I think I can, I think I can’: overconfidence and entrepreneurial behaviour, Journal of Economic Psychology, 28, pp. 502–27 Kolvereid, L. and E. Isaksen (2006). New business start-ups and subsequent entry into selfemployment, Journal of Business Venturing, 21(6), pp. 866–85 Kon, Y. and D. J. Storey (2003). A theory of discouraged borrowers, Small Business Economics, 21, pp. 37–49 Konings, J. (1995). Gross job flows and the evolution of size in UK establishments, Small Business Economics, 7, pp. 213–20

512

References

Konings, J., H. Lehmann and M. E. Schaffer (1996). Job creation and job destruction in a transition economy: Ownership, firm size, and gross job flows in Polish manufacturing, 1988–91, Labour Economics, 3, pp. 299–317 Kontos, M. (2003). Self-employment policies and migrants’ entrepreneurship in Germany, Entrepreneurship & Regional Development, 15, pp. 119–35 Kortum, S. and J. Lerner (2000). Assessing the contribution of venture capital to innovation, Rand Journal of Economics, 31, pp. 674–92 Koskela, E. (1983). Credit rationing and non-price loan terms: a re-examination, Journal of Banking and Finance, 7, pp. 405–16 Koster, N. (2005). Spin-off firms and individual start-ups: are they really different? Mimeo, University of Groningen, Netherlands KPMG (1999). An Evaluation of the Small Firms Loan Guarantee Scheme, DTI, London, HMSO Kreft, S. F. and R. S. Sobel (2005). Public policy, entrepreneurship and economic freedom, Cato Journal, 25(3), 595–616 Krugman, P. (1991). Geography and Trade, Cambridge MA, MIT Press Kuhn, P. J. (1988). Unions in a general equilibrium model of firm formation, Journal of Labor Economics, 6, pp. 62–82 Kuhn, P. J. and H. J. Schuetze (2001). Self-employment dynamics and self-employment trends: a study of Canadian men and women, 1982–1998, Canadian Journal of Economics, 34, pp. 760–84 Kumar, K. B., R. G. Rajan and L. Zingales (1999). What determines firm size? NBER working paper series no. 7208, Cambridge, National Bureau of Economic Research Kuznets, S. (1966). Modern Economic Growth: Rate Structure and Spread, New Haven CT, Yale University Press La Porta, R., F. Lopez-de-Silanes and A. Shleifer (1999). Corporate ownership around the world, Journal of Finance, 54, pp. 471–517 Laeven, L. and C. Woodruff (2007). The quality of the legal system, firm ownership and firm size, Review of Economics and Statistics, 89(4), pp. 601–14 Laferrère, A. and P. McEntee (1995). Self-employment and intergenerational transfers of physical and human capital: an empirical analysis of French data, Economic & Social Review, 27, pp. 43–54 Laffont, J.-J. and D. Martimort (2002). The Theory of Incentives: The Principal-Agent Model, Princeton University Press Laffont, J.-J. and T. N’Guessan (2000). Group lending with adverse selection, European Economic Review, 44, pp. 773–84 Lafontaine, F. and S. Bhattacharyya (1995). The role of risk in franchising, Journal of Corporate Finance, 2(1–2), pp. 39–74 Lafontaine, F. and K. Shaw (1998). Franchising growth and franchisor entry and exit in the US market: myth and reality, Journal of Business Venturing, 13, pp. 95–112 Lafuente, E., Y. Vaillant and J. Rialp (2007). Regional differences in the influence of role models: comparing the entrepreneurial process of rural Catalonia, Regional Studies, 41(6), pp. 779–95 Lambert-Mogiliansky, A., M. Majumdar and R. Radner (2007). Strategic analysis of petty corruption: entrepreneurs and bureaucrats, Journal of Development Economics, 83, pp. 351–67 Landier, A. (2001). Start-up financing: Banks versus venture capital, Mimeo, MIT (2004). Entrepreneurship and the stigma of failure, Mimeo, Stern School of Business, New York University Landier, A. and D. Thesmar (2008). Financial contracting with optimistic entrepreneurs, Review of Financial Studies, 22(1), pp. 117–50. Langlois, R. N. (2007). The entrepreneurial theory of the firm and the theory of the entrepreneurial firm, Journal of Management Studies, 44(7), pp. 1107–24

References

513

Langlois, R. N. and M. M. Cosgel (1993). Frank Knight on risk, uncertainty, and the firm: a new interpretation, Economic Inquiry, 31(3), pp. 456–66 Langowitz, N. and M. Minniti (2007). The entrepreneurial propensity of women, Entrepreneurship Theory & Practice, 31(3), pp. 341–64 Laussel, D. and M. Le Breton (1995). A general equilibrium theory of firm formation based on individual unobservable skills, European Economic Review, 39, pp. 1303–19 Lazear, E. P. (2005). Entrepreneurship, Journal of Labor Economics, 23, pp. 649–80 Lazear, E. P. and R. L. Moore (1984). Incentives, productivity, and labour contracts, Quarterly Journal of Economics, 99, pp. 275–96 Le, A. T. (1999). Empirical studies of self-employment, Journal of Economic Surveys, 13, pp. 381–416 Le, A. T. (2000). The determinants of immigrant self-employment in Australia, International Migration Review, 34, pp. 183–214 Lee, D. Y. and E. W. K. Tsang (2001). The effects of entrepreneurial personality, background and network activities on venture growth, Journal of Management Studies, 38(4), pp. 583–602 Lee, L. and P.-K. Wong (2006). Entrepreneurship education – a compendium of related issues, in S. C. Parker (ed.), The Life Cycle of Entrepreneurial Ventures, New York, Springer, pp. 79–105 Lee, M. A. and M. S. Rendall (2001). Self-employment disadvantage in the working lives of blacks and females, Population Research and Policy Review, 20, pp. 291–320 Lee, S. H., M. W. Peng and J. B. Barney (2007). Bankruptcy law and entrepreneurship development: a real options perspective, Academy of Management Review, 32(1), pp. 257–72 Lee, S. Y., R. Florida and Z. J. Acs (2006). Creativity and entrepreneurship: a regional analysis of new firm formation, Regional Studies, 38(8), pp. 879–91 Leeth, J. D. and J. A. Scott (1989). The incidence of secured debt – evidence from the small business community, Journal of Financial & Quantitative Analysis, 24, pp. 379–94 Leff, N. H. (1979). Entrepreneurship and economic development: the problem revisited, Journal of Economic Literature, 17, pp. 46–64 Leibenstein, H. (1968). Entrepreneurship and development, American Economic Review, 58, pp. 72–83 Leighton, P. (1983). Employment and self-employment: some problems of law and practice, Employment Gazette, 91, pp. 197–203 Leland, H. E. and D. H. Pyle (1977). Informational asymmetries, financial structure and financial intermediation, Journal of Finance, 32, pp. 371–87 Leleux, B. and B. Surlemont (2003). Public versus private venture capital: seeding or crowding out? A pan-European analysis, Journal of Business Venturing, 18, pp. 81–104 Lensink, R. and E. Sterken (2001). Asymmetric information, uncertainty and credit rationing, Journal of Public Economics, 79, pp. 365–74 (2002). The option to wait to invest and equilibrium credit rationing, Journal of Money, Credit and Banking, 34(1), pp. 221–25 Lensink, R., S. Gangopadhyay and M. Ghatak (2005). On joint liability lending and the peer selection effect, Economic Journal, 115, pp. 1005–15 Lentz, B. F. and D. N. Laband (1990). Entrepreneurial success and occupational inheritance among proprietors, Canadian Journal of Economics, 23, pp. 563–79 Leonard, J. S. (1986). On the size distribution of employment and establishments, NBER Working Paper No. 1951, Cambridge MA Lerner, J. (1995). Venture capitalists and the oversight of private firms, Journal of Finance, 50, pp. 301–18 (1999). The government as venture-capitalist: the long-run impact of the SBIR program, Journal of Business, 72, pp. 285–318

514

References

(2002). When bureaucrats meet entrepreneurs: the design of effective ‘public venture capital’ programs, Economic Journal, 112, pp. F73–F84 Lerner, J., H. Shane and A. Tsai (2003). Do equity financing cycles matter? Evidence from biotechnology alliances, Journal of Financial Economics, 67, pp. 411–46 LeRoy, S. F. and L. D. Singell (1987). Knight on risk and uncertainty, Journal of Political Economy, 95, pp. 394–406 Leung, M. W. H. (2003). Beyond Chinese, beyond food: unpacking the regulated Chinese restaurant business in Germany, Entrepreneurship & Regional Development, 15, pp. 103–18 Levenson, A. R. and K. L. Willard (2000). Do firms get the financing they want? Measuring credit rationing experienced by small businesses in the US, Small Business Economics, 14, pp. 83–94 Lévesque, M. and K. R. MacCrimmon (1997). On the interaction of time and money invested in new ventures, Entrepreneurship Theory & Practice, 22, pp. 89–110 Lévesque, M. and M. Minniti (2006). The effect of aging on entrepreneurial behaviour, Journal of Business Venturing, 21(2), pp. 177–94 Lévesque, M., D. A. Shepherd and E. J. Douglas (2002). Employment or self-employment: a dynamic utility-maximising model, Journal of Business Venturing, 17, pp. 189–210 Levie, J. (2007). Immigration, in-migration, ethnicity and entrepreneurship in the United Kingdom, Small Business Economics, 28(2–3), pp. 143–69 Levie, J. and D. Smallbone (2007). Immigration, ethnicity and entrepreneurial behaviour, in M. Minniti (ed.), Entrepreneurship: The Engine of Growth, Volume 1, Westport CT, Praeger, pp. 157–80 Levy, B. (1993). Obstacles to developing indigenous small and medium enterprises: an empirical assessment, World Bank Economic Review, 7, pp. 65–83 Lewis, W. A. (1955). The Theory of Economic Growth, London, Allen & Unwin Li, D. D., J. Feng and H. Jiang (2006). Institutional entrepreneurs, American Economic Review, Papers and Procedings, 96(2), pp. 358–62 Li, H. B., L. S. Meng and J. S. Zhang (2006). Why do entrepreneurs enter politics? Evidence from China, Economic Inquiry, 44(3), pp. 559–78 Li, W. (2002). Entrepreneurship and government subsidies: a general equilibrium analysis, Journal of Economic Dynamics and Control, 26, pp. 1815–44 Li, W. and J. Weinberg (2003). Firm-specific learning and the investment behaviour of large and small firms, International Economic Review, 44, pp. 599–625 Li, Y. (2008). Duration analysis of venture capital staging: a real options perspective, Journal of Business Venturing, 23, pp. 497–512 Liao, J. and W. B. Gartner (2006). The effects of pre-venture plan timing and perceived environmental uncertainty on the persistence of emerging firms, Small Business Economics, 27, pp. 23–40 Licht, A. N. and J. I. Siegel (2006). The social dimensions of entrepreneurship, in M. Casson, B. Yeung, A. Basu and N. Wadeson (eds.), The Oxford Handbook of Entrepreneurship, Oxford University Press, pp. 512–39 Light, I. (1972). Ethnic Entrepreneurs in America, Berkeley CA, University of California Press (1984). Immigrant and ethnic enterprise in North America, Ethnic & Racial Studies, 7, pp. 195–216 Light, I. and E. Bonacich (1988). Immigrant Entrepreneurs: Koreans in Los Angeles 1965–1982, Berkeley CA, University of California Press Light, I. and C. Rosenstein (1995). Race, Ethnicity and Entrepreneurship in Urban America, New York, Aldine de Gruyter Lin, P.-C. and D. S. Huang (2008). Technological regimes and firm survival: evidence across sectors and over time, Small Business Economics, 30, pp. 175–86 Lin, Z., G. Picot and J. Compton (2000). The entry and exit dynamics of self-employment in Canada, Small Business Economics, 15, pp. 105–25

References

515

Lindh, T. and H. Ohlsson (1996). Self-employment and windfall gains: evidence from the Swedish lottery, Economic Journal, 106, pp. 1515–26 (1998). Self-employment and wealth inequality, Review of Income and Wealth, 44, pp. 25–42 Link, A. N. and C. Rees (1990). Firm size, university research and the returns to R & D, Small Business Economics, 2, pp. 25–33 Lipper, G. and B. Sommer (2002). Encouraging angel capital: what the US states are doing, Venture Capital, 4, pp. 357–61 Lloyd-Ellis, H. and D. Bernhardt (2000). Enterprise, inequality and economic development, Review of Economic Studies, 67, pp. 147–68 Lofstrom, M. (2002). Labour market assimilation and the self-employment decision of immigrant entrepreneurs, Journal of Population Economics, 15, pp. 83–114 Lofstrom, M. and C. Wang (2007). Mexican-Hispanic self-employment entry: the role of business start-up constraints, Annals of the American Academy of Political and Social Science, 613, pp. 32–46 Lohmann, H. and S. Luber (2004). Trends in self-employment in Germany: different types, different developments?, in Arum and Müller (eds.) (2004), pp. 36–74 Lombard, K. V. (2001). Female self-employment and demand for flexible, nonstandard work schedules, Economic Inquiry, 39, pp. 214–37 Long, J. E. (1982a). The income tax and self-employment, National Tax Journal, 35, pp. 31–42 (1982b). Income taxation and the allocation of market labour, Journal of Labor Research, 3, pp. 259–76 Longhofer, S. D. (1997). Absolute priority rule violations, credit rationing, and efficiency, Journal of Financial Intermediation, 6, pp. 249–67 Longstreth, M., K. Stafford and T. Mauldin (1987). Self-employed women and their families: time use and socio-demographic characteristics, Journal of Small Business Management, 25(3), pp. 30–37 Lopez, R. (1984). Estimating labour supply and production decisions of self-employed farm producers, European Economic Review, 24, pp. 61–82 Loscocco, K. A. (1997). Work–family linkages among self-employed women and men, Journal of Vocational Behaviour, 50, pp. 204–26 Loscocco, K. A. and J. Robinson (1991). Barriers to women’s small-business success in the United States, Gender and Society, 5, pp. 511–32 Lotti F., E. Santarelli and M. Vivarelli (2003). Does Gibrat’s Law hold among young, small firms? Journal of Evolutionary Economics, 13, pp. 213–35 Loufti, M. F. (1992). An overview of self-employment in Europe: nature, trends and policy issues, in P. Leighton and A. Felstead (eds.), The New Entrepreneurs: Self-employment and Small Business in Europe, London, Kogan Page, pp. 41–68 Lowe, R. A. and A. A. Ziedonis (2006). Over-optimism and the performance of entrepreneurial firms, Management Science, 52(2), pp. 173–86 Lu, D. (1994). The entrepreneurs who do both: production and rent-seeking, Journal of Economic Behavior & Organization, 23, pp. 93–98 Lucas, R. E. (1978). On the size distribution of business firms, Bell Journal of Economics, 9, pp. 508–23 Lundström, A. and L. A. Stevenson (2005). Entrepreneurship Policy: Theory and Practice, New York, Springer (2007). Dressing the emperor: the fabric of entrepreneurship policy, in D. B. Audretsch, I. Grilo and A. R. Thurik (eds.), The Handbook of Research on Entrepreneurship Policy, Cheltenham, Edward Elgar Lunn, J. and T. Steen (2005). The heterogeneity of self-employment: the example of Asians in the United States, Small Business Economics, 24, pp. 143–58 Lusardi, A. and O. Mitchell (2006). Planning and financial literacy: how do women fare? Working Paper No. 2006–136, University of Michigan Retirement Research Center

516

References

Lyssiotou, P., P. Pashardes and T. Stengos (2004). Estimates of the black economy based on consumer demand approaches, Economic Journal, 114, pp. 622–40 Macafee, K. (1982). A glimpse of the hidden economy in the national accounts of the United Kingdom, in V. Tanzi (ed.), The Underground Economy in the United States and Abroad, Lexington MA, Lexington Books, pp. 147–61 MacDonald R. and F. Coffield (1991). Risky Business? Youth and the Enterprise Culture, London, Falmer Press Macpherson, D. A. (1988). Self-employment and married women, Economics Letters, 28, pp. 281–84 Maddala, G. S. (1983). Limited-Dependent and Qualitative Variables in Econometrics, Cambridge University Press Maddison, A. (1991). Dynamic Forces in Capital Development, Oxford University Press Madill, J. J., G. H. Haines Jr. and A. L. Riding (2005). The role of angels in technology SMEs: a link to venture capital, Venture Capital, 7, pp. 107–29 Magnac, T. and and J.-M. Robin (1996). Occupational choice and liquidity constraints, Ricerche Economiche, 50(2), pp. 105–33 Mahmood, T. (1992). Does the hazard rate for new plants vary between low- and high-tech industries? Small Business Economics, 4(3), pp. 201–09 (2000). Survival of newly founded businesses: a log-logistic model approach, Small Business Economics, 14, pp. 223–37 Mandel, A. S. (1992). Small business, banks and SBA loan guarantees: comment, Small Business Economics, 4, pp. 169–70 Manove, M. (2000). Entrepreneurs, optimism and the competitive edge, Mimeo, Boston University Manove, M. and A. J. Padilla (1999). Banking (conservatively) with optimists, Rand Journal of Economics, 30, pp. 324–50 Manser, M. E. and G. Picot (1999). The role of self-employment in US and Canadian job growth, Monthly Labor Review, 122(4), pp. 10–25 Mansfield, E. (1991). Academic research and industrial innovation, Research Policy, 20, pp. 1–12 Manski, C. F. and J. D. Straub (2000). Worker perceptions of job insecurity in the mid-1990s: evidence from the Survey of Economic Expectations, Journal of Human Resources, 35, pp. 447–79 Marchionatti, R. (1999). On Keynes’ animal spirits, Kyklos, 52, pp. 415–39 Marlow, S. (2003). Take-up of business growth training schemes by ethnic minority-owned small firms, International Small Business Journal, 10, pp. 34–46 Marshall, A. (1930). Principles of Economics, London, Macmillan Marshall, J. N., N. Alderman, C. Wong and A. Thwaites (1993). The impact of government-assisted management training and development on small and medium-sized enterprises in Britain, Environment & Planning C, 11, pp. 331–48 Martin, R. E. (1988). Franchising and risk management, American Economic Review, 78, pp. 954–68 Martinez, J. A. B. (2005). Equilibrium entrepreneurship rates, economic development and growth: Evidence from Spanish regions, Entrepreneurship & Regional Development, 17, pp. 145–61 Marvel, M R. and G. T. Lumpkin (2007). Technology entrepreneurs’ human capital and its effects on innovation radicalness, Entrepreneurship Theory & Practice, 31(6), pp. 807–28 Mason, C. M. (2006). Informal sources of venture finance, in S. C. Parker (ed.), The Life Cycle of Entrepreneurial Ventures, New York, Springer, pp. 259–99 Mason, C. M. and R. T. Harrison (2000a). The size of the informal venture capital market in the United Kingdom, Small Business Economics, 15, pp. 137–48 (2000b). Venture capital: rationale, aims and scope, Venture Capital, 1, pp. 1-47 Masuda, T. (2006). The determinants of latent entrepreneurship in Japan, Small Business Economics, 26(3), pp. 227–40 Masuerl, E. and H. P. Smit (2000). Planning behaviour of small firms in central Vietnam, Journal of Small Business Management, 38(2), pp. 95–102

References

517

Mata, J. (1994). Firm growth during infancy, Small Business Economics, 6, pp. 27–39 (1996). Markets, entrepreneurs and the size of new firms, Economics Letters, 52, pp. 89–94 Mata, J. and J. A. F. Machado (1996). Firm start-up size: a conditional quantile approach, European Economic Review, 40, pp. 1305–23 Mata, J. and P. Portugal (1994). Life duration of new firms, Journal of Industrial Economics, 42, pp. 227–45 (2002). The survival of new domestic and foreign-owned firms, Strategic Management Journal, 23, pp. 323–43 Mata, J., P. Portugal and P. Guimaraes (1995). The survival of new plants: start-up conditions and post-entry evolution, International Journal of Industrial Organization, 13, pp. 459–81 Mathews, R. D. (2006). Strategic alliances, equity stakes, and entry deterrence, Journal of Financial Economics, 80, pp. 35–79 Matsuyama, K. (2006). The 2005 Lawrence R. Klein Lecture: emergent class structure, International Economic Review, 47(2), pp. 327–60 Mattesini, F. (1990). Screening in the credit market, European Journal of Political Economy, 6, pp. 1–22 Maxim, P. S. (1992). Immigrants, visible minorities, and self-employment, Demography, 29, pp. 181–98 Mayer-Schönberger, V. (2003). E-commerce, entrepreneurship and the law: reassessing a relationship, in D. M. Hart (ed.), The Emergence of Entrepreneurship Policy, Cambridge University Press, pp. 195–213 Mazzeo, M. J. (2004). Retail contracting and organisational form: alternatives to chain affiliation in the motel industry, Journal of Economics and Management Strategy, 13(4), pp. 599–615 McCarthy, A. M., D. A. Krueger and T. S. Schoenecker (1990). Changes in the time allocation patterns of entrepreneurs, Entrepreneurship Theory & Practice, 15, pp. 7–18 McClelland, D. C. (1961). The Achieving Society, Princeton NJ, Van Nostrand McCormick, B. and J. Wahba (2001). Overseas work experience, savings and entrepreneurship amongst return migrants to LDCs, Scottish Journal of Political Economy, 48(2), pp. 164–78 McGrath, R. G. (1999). Falling forward: real options reasoning and entrepreneurial failure, Academy of Management Review, 24(1), pp. 13–30 McKernan, S. M. (2002). The impact of microcredit programs on self-employment profits: do noncredit program aspects matter? Review of Economics and Statistics, 84, pp. 93–115 McManus, P. A. (2000). Market, state, and the quality of new self-employment jobs among men in the US and Western Germany, Social Forces, 78(3), pp. 865–905 McMillan, J. and C. Woodruff (1999). Interfirm relationships and informal credit in Vietnam, Quarterly Journal of Economics, 114(4), pp. 1285–1320 (2002). The central role of entrepreneurs in transition economies, Journal of Economic Perspectives, 16(3), pp. 153–70 McPherson, M. A. (1995). The hazards of small firms in Southern Africa, Journal of Development Studies, 32(1), pp. 31–54 (1996). Growth of micro and small enterprises in Southern Africa, Journal of Development Economics, 48(2), pp. 253–77 McPherson, M., L. Smith-Lovin and J. M. Cook (2001). Birds of a feather: homophily in social networks, Annual Review of Sociology, 27, pp. 415–44 Meager, N. (1992a). Does unemployment lead to self-employment? Small Business Economics, 4, pp. 87–103 (1992b). The characteristics of the self-employed: some Anglo-German comparisons, in P. Leighton and A. Felstead (eds.), The New Entrepreneurs: Self-employment and Small Business in Europe, London, Kogan Page, pp. 69–99

518

References

(1993). From unemployment to self-employment in the European Community, in F. Chittenden et al. (eds.), Small Firms: Recession and Recovery, London, Paul Chapman, pp. 27–53 (1994). Self-employment schemes for the unemployed in the European Community, in G. Schmid (ed.), Labour Market Institutions in Europe, New York, M.E. Sharpe, pp. 183–242 Meager, N. and P. Bates (2004). Self-employment in the United Kingdom during the 1980s and 1990s, in Arum and Müller (eds.) (2004), pp. 135–69 Meager, N., P. Bates and M. Cowling (2003). An evaluation of business start-up support for young people, National Institute Economic Review, 59, pp. 59–72 Meager, N., G. Court and J. Moralee (1994). Self-employment and the Distribution of Income, Brighton, The Institute for Manpower Studies (1996). Self-employment and the distribution of income, in J. Hills (ed.), New Inequalities, Cambridge University Press Meh, A. A. (2005). Entrepreneurship, wealth inequality and taxation, Review of Economic Dynamics 8, pp. 688–719 Mengistae, T. (2006). Competition and entrepreneurs’ human capital in small business longevity and growth, Journal of Development Studies, 42(5), pp. 812–36 Méon, P.-G. and K. Sekkat (2005). Does corruption grease or sand the wheels of growth? Public Choice, 122(1–2), pp. 69–97 Meredith, G. G., R. E. Nelson and P. A. Neck (1982). The Practice of Entrepreneurship, Geneva, ILO Mesnard,A. (2004). Temporary migration and capital market imperfections, Oxford Economic Papers, 56(2), pp. 242–62 Mesnard, A., and M. Ravallion (2006). The wealth effect of new business start-ups in a developing economy, Economica, 73, pp. 367–92 Metcalf, H., T. Modood and S. Virdee (1996). Asian Self-employment: The Interaction of Culture and Economics, London, Policy Studies Institute Meyer, B. (1990). Why are there so few black entrepreneurs? NBER Working Paper No. 3537, Cambridge MA Michael, S. C. (1996). To franchise or not to franchise: an analysis of decision rights and organisational form shares, Journal of Business Venturing, 11, pp. 59–71 Michael, S. C. and J. G. Combs (2008). Entrepreneurial failure: the case of franchisees, Journal of Small Business Management, 46(1), pp. 73–90 Michelacchi, C. (2003). Low returns in R & D due to the lack of entrepreneurial skills, Economic Journal, 113, pp. 207–25 Michelacci, C. and V. Quadrini (2005). Borrowing from employees: wage dynamics with financial constraints, Journal of the European Economic Association, 3(2–3), pp. 360–69 Michelacci, C. and O. Silva (2007). Why so many local entrepreneurs? Review of Economics and Statistics, 89(4), pp. 615–33 Michelacci, C. and J. Suarez (2004). Business creation and the stock market, Review of Economic Studies, 71(2), pp. 459–81 Milde, H. and J. G. Riley (1988). Signalling in credit markets, Quarterly Journal of Economics, 103, pp. 101–29 Milgrom, P. and J. Roberts (1992). Economics, Organization, and Management, Englewood Cliffs NJ, Prentice-Hall Miller, R. A. (1984). Job matching and occupational choice, Journal of Political Economy, 92, pp. 1086–1120 Miller, W. F. (2004). Fostering and sustaining entrepreneurial regions, International Journal of Technology Management, 28, pp. 324–35 Min, P. G. (1984). From white collar occupations to small business: Korean immigrants’ occupational adjustment, The Sociological Quarterly, 25, pp. 333–52

References

519

(1996). Caught in the Middle: Korean Merchants in America’s Multiethnic Cities, Berkeley CA, University of California Press Miner, J. B. and N. S. Raju (2004). Risk propensity differences between managers and entrepreneurs and between low- and high-growth entrepreneurs: a reply in a more conservative vein, Journal of Applied Psychology, 89(1), pp. 3–13 Minetti, R. (2007). Bank capital, firm liquidity and project quality, Journal of Monetary Economics, 54(8), pp. 2584–94 Minniti, M. (2005). Entrepreneurship and network externalities, Journal of Economic Behavior & Organization, 57, pp. 1–27 Minniti, M. and M. Lévesque (2008). Recent developments in the economics of entrepreneurship, Journal of Business Venturing, 23(6), pp. 603–12 Mirus, R. and R. S. Smith (1996). Self-employment, tax evasion and the underground economy: micro-based estimates for Canada, Working Paper 1002, International Tax Program, Harvard Law School Mocan, H. N. and E. Tekin (2003). Nonprofit sector and part-time work: an analysis of employeremployee matched data on child care workers, Review of Economics and Statistics, 85(1), pp. 38–50 Moehling, C. M. and R. H. Steckel (2004). Entrepreneurial activity and wealth inequality: a historical perspective, in D. Holtz-Eakin and H. S. Rosen (eds.), Public Policy and the Economics of Entrepreneurship, Cambridge MA, MIT Press, pp. 181–209 Mohapatra, S., S. Rozelle and R. Goodhue (2007). The rise of self-employment in rural China: development or distress? World Development, 35(1), pp. 163–81 Moore, B. (1994). Financial constraints to the growth and development of small high technology firms, in A. Hughes and D. Storey (eds.), Finance and the Small Firm, London, Routledge, pp. 112–44 Moore, C. S. and R. E. Mueller (2002). The transition from paid to self-employment in Canada: the importance of push factors, Applied Economics, 34, pp. 791–801 Moore, D. A., J. M. Oesch and C. Zietsma (2007). What competition? Myopic self-focus in marketentry decisions, Organization Science, 18(3), pp. 440–54 Moore, R. L. (1983a). Self-employment and the incidence of the payroll tax, National Tax Journal, 36, pp. 491–501 (1983b). Employer discrimination: evidence from self-employed workers, Review of Economics and Statistics, 65, pp. 496–501 Mora, M. T. (2006). Self-employed Mexican immigrants residing along the US–Mexico border: the earnings effect of working in the US versus Mexico, International Migration Review, 40(4), pp. 885–98 Mora, M. T. and A. Dávila (2006). Mexican immigrant self-employment along the US–Mexico border: an analysis of 2000 Census data, Social Sciences Quarterly 87, pp. 91–109 Moralee, L. (1998). Self-employment in the 1990s, Labour Market Trends, 106, pp. 121–30 Morck, R. and B. Yeung (2004). Family control and the rent-seeking society, Entrepreneurship Theory & Practice, 28(4), pp. 391–409 Morduch, J. (1999a). The microfinance promise, Journal of Economic Literature, 37, pp. 1569–1614 (1999b). The role of subsidies in microfinance: evidence from the Grameen Bank, Journal of Development Economics, 60(1), pp. 229–48 Moreno, A. M. and J. C. Casillas (2007). High-growth SMEs versus non-high-growth SMEs: a discriminant analysis, Entrepreneurship & Regional Development, 19, pp. 69–88 Moresi, S. (1998). Optimal taxation and firm formation: a model of asymmetric information, European Economic Review, 42, pp. 1525–51 Morgan, K. (1997). The learning region: institutions, innovation and regional renewal, Regional Studies, 31, pp. 491–503

520

References

Morris, M. and M. Schindehutte (2005). Entrepreneurial values and the ethnic enterprise: an examination of six subcultures, jsbm, 43(4), pp. 453–79 Moskowitz, T. J. and A. Vissing-Jørgensen (2002). The returns to entrepreneurial investment: a private equity premium puzzle? American Economic Review, 92, pp. 745–78 Mougin, F. (2007). Asymmetric information and legal investor protection, Economics Letters, 95, pp. 253–58 Mowery, D. C., R. R. Nelson, B. N. Sampat and A. A. Ziedonis (2004). Ivory Tower and Industrial Innovation: University-Industry Technology Transfer before and after the Bayh-Dole Act, Stanford University Press Mueller, P. (2006a). Exploring the knowledge filter: how entrepreneurship and university–industry relationships drive economic growth, Research Policy, 35(10), pp. 1499–1508 (2006b). Entrepreneurship in the region: breeding ground for nascent entrepreneurs? Small Business Economics, 27, pp. 41–58 (2007). Exploiting entrepreneurial opportunities: the impact of entrepreneurship on growth, Small Business Economics, 28(4), pp. 355–62 Müller, W. and R. Arum (2004). Self-employment dynamics in advanced economies, in Arum and Müller (eds.) (2004), pp. 1–35 Murphy, K. M., A. Shleifer and R. Vishny (1991). The allocation of talent: implications for growth, Quarterly Journal of Economics, 106, pp. 503–30 (1993). Why is rent-seeking so costly to growth? American Economic Review, Papers and Procedings, 83, pp. 409–14 Murray, M. P. (2006). Avoiding invalid instruments and coping with weak instruments, Journal of Economic Perspectives, 20(4), pp. 111–32 Myers, S. C. and N. S. Majluf (1984). Corporate financing and investment decisions when firms have information that investors do not have, Journal of Financial Economics, 13, pp. 187–221 Myrdal, G. (1944). An American Dilemma: The Negro Problem and Modern Democracy, New York, Harper Nafziger, E. W. and D. Terrell (1996). Entrepreneurial human capital and the long run survival of firms in India, World Development, 24, pp. 689–96 National Economic Research Associates (1990) An Evaluation of the Loan Guarantee Scheme, Department of Employment Research Paper No. 74, London, HMSO Naudé, W., T. Gries, E. Wood and A. Meintjies (2008). Regional determinants of entrepreneurial start-ups in a developing country, Entrepreneurship & Regional Development, 20, pp. 111–24 Neher, D. V. (1999). Staged financing: an agency perspective, Review of Economic Studies, 66, pp. 255–74 Nelson, R. and S. Winter (1982). An Evolutionary Theory of Economic Change, Cambridge MA, Belknap Press Nenadic, S. (1990). The life cycle of firms in nineteenth century Britain, in P. Jobert and M. Mors (eds.), The Birth and Death of Companies: an Historical Perspective, Lancs., Parthenon Newbert, S. L. (2005). New firm formation: A dynamic capability perspective, Journal of Small Business Management, 43(1), pp. 55–77 Newman, A. F. (2007). Risk-bearing and entrepreneurship, Journal of Economic Theory, 137(1), pp. 11–26 Nickell, S. J. (1996). Competition and corporate performance, Journal of Political Economy, 104, pp. 724–46 Nickell, S. J., P. Nicolitsas and N. Dryden (1997). What makes firms perform well? European Economic Review, 41, pp. 783–96 Nickerson, J. A. and B. S. Silverman (2003). Why aren’t all truck drivers owner-operators? Asset ownership and the employment relation in interstate for-hire trucking, Journal of Economics and Management Strategy, 12(1), pp. 91–118

References

521

Nicolaou, N., S. Shane, L. Cherkas and T. D. Spector (2008). The influence of sensation seeking in the heritability of entrepreneurship, Strategic Entrepreneurship Journal, 2(1), 7–21 Niittytkangas, H. and H. Tervo (2005). Spatial variations in intergenerational transmission of selfemployment, Regional Studies, 39, pp. 319–32 Nisjen, A. (1988). Self-employment in the Netherlands, International Small Business Journal, 7, pp. 52–60 Nocke, V. (2006). A gap for me: entrepreneurs and entry, Journal of the European Economic Association, 4(5), pp. 929–56 Noe, T. H. (1988). Capital structure and signalling game equilibria, Review of Financial Studies, 1, pp. 331–55 Nordhaus, W. D. (2004). Schumpeterian profits in the American economy: theory and measurement, NBER Working Paper No. 10433, Cambridge MA Norton, S. W. (1988). An empirical look at franchising as an organisational form, Journal of Business, 61(2), pp. 197–218 Norton, W. I. and W. T. Moore (2002). Entrepreneurial risk: have we been asking the wrong question? Small Business Economics, 18, pp. 281–87 Nucci, A. R. (1999). The demography of business closings, Small Business Economics, 12(1), pp. 25–39 Nykvist, J. (2008). Entrepreneurship and liquidity constraints: evidence from Sweden, Scandinavian Journal of Economics, 110(1), pp. 23–43 Nyström, K. (2007). An industry disaggregated analysis of the determinants of regional entry and exit, Annals of Regional Science, 41, pp. 877–96 Oakey, R. (1984). Finance and innovation in British small independent firms, Omega, 12, pp. 113–24 OECD (1986, 1994a, 1999). Employment Outlook, Paris, OECD Publications (1994b). Taxation and Small Businesses, Paris, OECD Publications (1998). Fostering Entrepreneurship, Paris, OECD Publications (2000). The partial renaissance of self-employment, in Employment Outlook, Paris, OECD Publications (2003). The Sources of Economic Growth in OECD Countries, Paris, OECD Publications O’Farrell, P. N. and A. R. Pickles (1989). Entrepreneurial behaviour within male work histories: a sector-specific analysis, Environment and Planning C, 21, pp. 311–31 Oi, W. Y. (1983). Heterogeneous firms and the organisation of production, Economic Inquiry, 21, pp. 147–71 Oi, W. Y. and T. L. Idson (1999). Firm size and wages, in O. Ashenfelter and D. Card (eds.), Handbook of Labour Economics, Vol. 3b, Amsterdam, North-Holland Oliveira, B., and A. Fortunato (2006). Firm growth and liquidity constraints: a dynamic analysis, Small Business Economics, 27, pp. 139–56 Ono, Y. (2007). Market thickness and outsourcing services, Regional Science and Urban Economics, 37, pp. 220–38 Ordover, J. and A. Weiss (1981). Information and the law: evaluating legal restrictions on competitive contracts, American Economic Review, Papers and Proceedings, 71, pp. 399–404 Orser, B. J., S. Hogarth-Scott and A. L. Riding (2000). Performance, firm size and management problem solving, Journal of Small Business Management, 38(4), pp. 42–58 Orser, B. J., A. L. Riding and K. Manley (2006). Women entrepreneurs and financial capital, Entrepreneurship Theory & Practice, 30(5), pp. 643–65 Orzach, R. and Y. Tauman (2005). Strategic dropouts, Games & Economic Behaviour, 50, pp. 79–88 Östberg, P. (2006). Disclosure, investment and regulation, Journal of Financial Intermediation, 5, pp. 285–306 Otani, K. (1996). A human capital approach to entrepreneurial capacity, Economica, 63, pp. 273–89

522

References

Padilla, A. J. and M. Pagano (1997). Endogenous communication among lenders and entrepreneurial incentives, Review of Financial Studies, 10(1), pp. 205–36 Pakes, A. and R. Ericson (1998). Empirical implications of alternative models of firm dynamics, Journal of Economic Theory, 79, pp. 1–45 Pakes, A. and S. Nitzan (1983). Optimum contracts for research personnel, research employment, and the establishment of ‘rival’ enterprises, Journal of Labor Economics, 1(4), pp. 345–65 Palich, L. E. and D. R. Bagby (1995). Using cognitive theory to explain entrepreneurial risk-taking: challenging conventional wisdom, Journal of Business Venturing, 10, pp. 425–38 Parasuraman, S. and C. A. Simmers (2001). Type of employment, work–family conflict and wellbeing: a comparative study, Journal of Occupational Behaviour, 22, pp. 551–68 Parente, S. and E. Prescott (1999). Monopoly rights: a barrier to riches, American Economic Review, 89, pp. 1216–33 Parker, S. C. (1996). A time series model of self-employment under uncertainty, Economica, 63, pp. 459–75 (1997a). The effects of risk on self-employment, Small Business Economics, 9, pp. 515–22 (1997b). The distribution of self-employment income in the United Kingdom, 1976–1991, Economic Journal, 107, pp. 455–66 (1999a). The optimal linear taxation of employment and self-employment incomes, Journal of Public Economics, 73, pp. 107–23 (1999b). The inequality of employment and self-employment incomes: a decomposition analysis for the UK, Review of Income and Wealth, 45, pp. 263–74 (2000). Saving to overcome borrowing constraints: implications for small business entry and exit, Small Business Economics, 15, pp. 223–32 (2001). Risk, self-employment and differential income taxation, Manchester School, 69, pp. 1–15 (2002). Do banks ration credit to new enterprises? And should governments intervene?, Scottish Journal of Political Economy, 49, pp. 162–95 (2003a). The distribution of wealth among older self-employed Britons, Fiscal Studies, 24, pp. 23–43 (2003b). Does tax evasion affect occupational choice?, Oxford Bulletin of Economics and Statistics, 65, pp. 379–94 (2003c). Asymmetric information, occupational choice and government policy, Economic Journal, 113, pp. 861–82 (2004). The Economics of Self-employment and Entrepreneurship, Cambridge University Press (2005a). The Economics of Entrepreneurship: what we know and what we don’t, Foundations & Trends in Entrepreneurship, 1(1), pp. 1–55 (2005b). Explaining regional variations in entrepreneurship as multiple occupational equilibria, Journal of Regional Science, 45, pp. 829–50 (2006a). Learning about the unknown: how fast do entrepreneurs adjust their beliefs? Journal of Business Venturing, 21(1), pp. 1–26 (2006b). Entrepreneurship, self-employment and the labour market, in M. Casson, B. Yeung, A. Basu and N. Wadeson (eds.), The Oxford Handbook of Entrepreneurship, Oxford University Press, pp. 435–60 (2007a). Policy-makers beware!, in D. B. Audretsch, I. Grilo and A. R. Thurik (eds.), The International Handbook of Entrepreneurship Policy Research, Cheltenham, Edward Elgar (2007b). Contracting out, public policy and entrepreneurship, Working paper, Durham University, England (2007c). Entrepreneurial learning and the existence of credit markets, Journal of Economic Behavior & Organization, 62(1), pp. 37–46 (2007d). Which firms do the entrepreneurs come from? Economics Bulletin, 10(10), pp. 1–9

References

523

(2008a). Entrepreneurship among married couples in the United States: a simultaneous probit approach, Labour Economics, 15(3), pp. 515–37 (2008b). The economics of formal business networks, Journal of Business Venturing, 23(6), pp. 627–40 (2008c). Social entrepreneurs: a neoclassical theory, in G. E. Shockley, P. M. Frank and R. R. Stough (eds.), Non-Market Entrepreneurship: Interdisciplinary Approaches, Cheltenham, Edward Elgar (2009a). Can cognitive biases explain venture team homophily? forthcoming, Strategic Entrepreneurship Journal (2009b). Why do small firms produce the entrepreneurs? Journal of Socio-Economics, 38(3), pp. 484–94 (forthcoming) Intrapreneurship or entrepreneurship? Journal of Business Venturing Parker, S. C. and Y. Belghitar (2006). What happens to nascent entrepreneurs?An econometric analysis of the PSED, Small Business Economics, 27, pp. 81–101 Parker, S. C. and C. M. van Praag (2006a). Schooling, capital constraints and entrepreneurial performance: the endogenous triangle, Journal of Business & Economic Statistics, 24, pp. 416–31 (2006b). The entrepreneur’s mode of entry: business takeover or new venture start? IZA Discussion Paper No. 2382, Bonn (2009). Group status and entrepreneurship, forthcoming, Journal of Economics & Management Strategy Parker, S. C. and M. T. Robson (2004). Explaining international variations in self-employment: evidence from a panel of OECD countries, Southern Economic Journal, 71, pp. 287–301 Parker, S. C. and J. Rougier (2007). Self-employment and the retirement decision, Applied Economics, 39, pp. 697–713 Parker, S. C. and A. van Witteloostuijn (2009). A general framework for estimating multidimensional contingency fit, Unpublished paper, University of Antwerp Parker, S. C., Y. Belghitar and T. Barmby (2005). Wage uncertainty and self-employed labour supply, Economic Journal, 115, pp. C190–C207 Parker, S. C., D. J. Storey and A. van Witteloostuijn (2005). What happens to gazelles? The importance of dynamic management strategy, Discussion Paper No. 12, Centre for Entrepreneurship, Durham Business School, England Parwada, J. T. (2008). The genesis of home bias? The location and portfolio choices of investment company start-ups, Journal of Financial & Quantitative Analysis, 43(1), pp. 245–66 Paulson, A. L. and R. Townsend (2004). Entrepreneurship and financial constraints in Thailand, Journal of Corporate Finance, 10, pp. 229–62 Paulson, A. L., R. Townsend and A. Karaivanov (2006). Distinguishing limited liability from moral hazard in a model of entrepreneurship, Journal of Political Economy, 114, pp. 100–144 Pavitt, K., M. Robson and J. Townsend (1987). Technological accumulation, diversification and organisation in UK companies, 1945–1983, Manchester School, 35, pp. 81–99 Paxton, J., D. Graham and C. Thraen (2000). Modelling group loan repayment behaviour: new insights from Burkina Faso, Economic Development & Cultural Change, 48, pp. 639–55 Pe’er, A. and I. Vertinsky (2008). Firm exits as a determinant of new entry: is there evidence of local creative destruction? Journal of Business Venturing, 23, pp. 280–306 Peneder, M. R. (2008). Firm entry and turnover: the nexus with profitability and growth, Small Business Economics, 30, pp. 327–44 Pennings, J. M., K. Lee and A. van Witteloostuijn (1998). Human capital, social capital, and firm dissolution, Academy of Management Journal, 41, pp. 425–40 Peretto, P. F. (1998). Technological change, market rivalry, and the evolution of the capitalist engine of growth, Journal of Economic Growth, 3, pp. 53–80

524

References

Pérotin, V. (2006). Entry, exit and the business cycle: are cooperatives different? Journal of Comparative Economics, 34, pp. 295–316 Perry, C. W. and H. S. Rosen (2004). The self-employed are less likely to have health insurance than wage earners. So what?, in D. Holtz-Eakin and H. S. Rosen (eds.), Public Policy and the Economics of Entrepreneurship, Cambridge MA, MIT Press, pp. 23–57 Pestieau, P. and U. M. Possen (1991). Tax evasion and occupational choice, Journal of Public Economics, 45, pp. 107–25 (1992). How do taxes affect occupational choice?, Public Finance, 47, pp. 108–119 Petersen, M. A. and R. G. Rajan (1994). The benefits of lending relationships: evidence from small firm business data, Journal of Finance, 49, pp. 3–37 (1997). Trade credit: theories and evidence, Review of Financial Studies, 10, pp. 661–69 (2002). Does distance still matter? The information revolution in small business lending, Journal of Finance, 57, pp. 2533–70 Petrunia, R. (2008). Does Gibrat’s Law hold? Evidence from Canadian retail and manufacturing firms, Small Business Economics, 30, pp. 201–14 Pfeiffer, F. and F. Reize (2000). Business start-ups by the unemployed – an econometric analysis based on firm data, Labour Economics, 7, pp. 629–63 Phillips, B. D. (1985). The effect of industry deregulation on the small business sector, Business Economics, 20, pp. 28–37 Phillips, B. D. and B. A. Kirchhoff (1989). Formation, growth and survival: small firm dynamics in the US economy, Small Business Economics, 1, pp. 65–74 Phillips, J. D. (1962). The Self-employed in the United States, Urbana IL, University of Illinois Phillips, P. C. B. (1986). Understanding spurious regressions in econometrics, Journal of Econometrics, 33, pp. 311–40 Piketty, T. (1997). The dynamics of the wealth distribution and the interest rate with credit rationing, Review of Economic Studies, 64, pp. 173–89 Pickles, A. R. and P. N. O’Farrell (1987). An analysis of entrepreneurial behaviour from male work histories, Regional Studies, 21, pp. 425–44 Pinfold, J. F. (2001). The expectations of new business founders: the New Zealand case, Journal of Small Business Management, 39, pp. 279–85 Piore, M. J. and C. F. Sabel (1984). The Second Industrial Divide: Possibilities for Prosperity, New York, Basic Books Pissarides, C. A. and G. Weber (1989). An expenditure-based estimate of Britain’s black economy, Journal of Public Economics, 39, pp. 17–32 Pitt, M. M. and S. R. Khandker (1998). The impact of group-based credit programs on poor households in Bangladesh: does the gender of participants matter? Journal of Political Economy, 106, pp. 958–96 Polkovnichenko, V. (2003). Human capital and the private equity premium, Review of Economic Dynamics, 6, pp. 831–45 Pollert, A. (1988). The ‘flexible firm’: fixation or fact? Work, Employment and Society, 2, pp. 281–316 Portes, A. and M. Zhou (1996). Self-employment and the earnings of immigrants, American Sociological Review, 61, pp. 219–30 Poterba, J. M. (1989a). Venture capital and capital gains taxation, in L. H. Summers (ed.), Tax Policy and the Economy, Cambridge MA, MIT Press, pp. 47–67 (1989b). Capital gains tax policy towards entrepreneurship, National Tax Journal, 42, pp. 375–89 Povel, P. and M. Raith (2004). Optimal debt with unobservable investments, Rand Journal of Economics, 35, pp. 599–616 Power, L. and M. Rider (2002). The effect of tax-based savings incentives on the self-employed, Journal of Public Economics, 85, pp. 33–52 Praag, C. M. van (1999). Some classic views on entrepreneurship, De Economist, 147, pp. 311–35

References

525

(2003). Business survival and success of young small business owners, Small Business Economics, 21, pp. 1–17 (2005). Successful Entrepreneurship: Confronting Economic Theory with Empirical Practice, Cheltenham, Edward Elgar Praag, C. M. van and J. S. Cramer (2001). The roots of entrepreneurship and labour demand: individual ability and low risk aversion, Economica, 68, pp. 45–62 Praag, C. M. van and H. van Ophem (1995). Determinants of willingness and opportunity to start as an entrepreneur, Kyklos, 48, pp. 513–40 Praag, C. M. van and P. H. Versloot (2007). What is the value of entrepreneurship? A review of recent research, Small Business Economics, 29(4), pp. 351–82 Preisendörfer, P. and T. Voss (1990). Organisational mortality of small firms: the effects of entrepreneurial age and human capital, Organizational Studies, 11, 107–29 Prevenzer, M. (1997). The dynamics of industrial clustering in biotechnology, Small Business Economics, 9, pp. 255–71 Prottas, D. J. and C. A. Thompson (2006). Stress, satisfaction, and the work–family interface: a comparison of self-employed business owners, independents, and organisational employees, Journal of Occupational Health Psychology, 11(4), pp. 366–78 Prusa, T. J. and J. A. Schmitz (1991). Are new firms an important source of innovation? Economics Letters, 35, pp. 339–42 Puri, M. and D. T. Robinson (2005). Optimism, entrepreneurship and economic choice, Working Paper, Duke University Quadrini, V. (1999). The importance of entrepreneurship for wealth concentration and mobility, Review of Income and Wealth, 45, pp. 1–19 (2000). Entrepreneurship, saving, and social mobility, Review of Economic Dynamics, 3, pp. 1–40 Quinn, J. F. (1980). Labour-force participation patterns of older self-employed workers, Social Security Bulletin, 43, pp. 17–28 Rafiq, M. (1992). Ethnicity and enterprise: a comparison of Muslim and non-Muslim owned Asian businesses in Britain, New Community, 19, pp. 43–60 Rahman, A. (1993). The general replicability of the Grameen Bank model, in A. N. M. Wahid (ed.), The Grameen Bank: Poverty Relief in Bangladesh, Boulder CO, Westview Press, pp. 209–22 Rai, A. S. and T. Sjöström (2003). Is Grameen lending efficient? Repayment incentives and insurance in village economies, Review of Economic Studies, 71, pp. 217–34 Rajan, R. and L. Zingales (2003). The great reversal: the politics of financial development in the 20th century, Journal of Financial Economics, 69, pp. 5–50 Ram, M. and D. Smallbone (2003). Policies to support ethnic minority enterprise: the English experience, Entrepreneurship & Regional Development, 15, pp. 151–66 Rampini, A. A. (2004). Entrepreneurial activity, risk and the business cycle, Journal of Monetary Economics, 51, pp. 555–73 Rau, R., K. Hoffmann, U. Metz, P. G. Richter, U. Roesler and U. Stephan (2008). Health risks of entrepreneurs, Zeitschrift fur Arbeits und Organisationpsychologie, 52(3), pp. 115–25 Rauch, A. and M. Frese (2000). Psychological approaches to entrepreneurial success: a general model and an overview of findings, International Review of Industrial and Organizational Psychology, 15, pp. 101–43 Ravasi, D. and C. Turati (2005). Exploring entrepreneurial learning: a comparative study of technology development projects, Journal of Business Venturing, 20, pp. 137–64 Ravid, S. A. and M. Spiegel (1997). Optimal financial contracts for a start-up with unlimited operating discretion, Journal of Financial & Quantitative Analysis, 32(3), pp. 269–86

526

References

Razin, E. and A. Langlois (1996). Metropolitan characteristics and entrepreneurship among immigrants and ethnic groups in Canada, International Migration Review, 30, pp. 703–27 Rees, H. and A. Shah (1986). An empirical analysis of self-employment in the UK, Journal of Applied Econometrics, 1, pp. 95–108 (1994). The characteristics of the self-employed: the supply of labour, in J. Atkinson and D. J. Storey (eds.), Employment, the Small Firm and the Labour Market, London, Routledge, pp. 317–27 Reid, G. C. (1991). Staying in business, International Journal of Industrial Organization, 9, pp. 545–56 (1993). Small Business Enterprise: An Economic Analysis, London, Routledge (1999). Capital structure at inception and the short-run performance of micro-firms, in Z. J. Acs, B. Carlsson and C. Karlsson (eds.), Entrepreneurship, Small and Medium Sized Enterprises and the Macroeconomy, Cambridge University Press, pp. 186–205 Reid, G. C. and L. R. Jacobsen (1988). The Small Entrepreneurial Firm, Aberdeen University Press Reid, G. C. and J. A. Smith (2000). What makes a new business start-up successful? Small Business Economics, 14, pp. 165–82 Reinganum, J. (1983). Uncertain innovation and the persistence of monopoly, American Economic Review, 73, pp. 741–48 Renzulli, L. A., H. Aldrich and J. Moody (2001). Family matters: gender, networks and entrepreneurial outcomes, Social Forces, 79(2), pp. 523–46 Repullo, R. and J. Suarez (2000). Entrepreneurial moral hazard and bank monitoring: a model of the credit channel, European Economic Review, 44, pp. 1931–50 Reynolds, P. D. (1987). New firms: societal contribution versus survival potential, Journal of Business Venturing, 2, pp. 231–46 (1994). Autonomous firm dynamics and economic growth in the United States, 1986–1990, Regional Studies, 28, pp. 429–42 (1997). Who starts new firms? Preliminary explorations of firms-in-gestation, Small Business Economics, 9, pp. 449–62 (1999). Creative destruction: source or symptom of economic growth? in Z. J. Acs, B. Carlsson and C. Karlsson (eds.), Entrepreneurship, Small and Medium Sized Enterprises and the Macroeconomy, Cambridge University Press, pp. 97–136 Reynolds, P. D. and S. B. White (1997). The Entrepreneurial Process: Economic Growth, Men, Women and Minorities, Westport CT, Quorum Books Reynolds, P. D., D. J. Storey and P. Westhead (1994). Cross-national comparisons of the variation in new firm formation rates, Regional Studies, 28, pp. 443–56 Reynolds, P. D., W. D. Bygrave, E. Autio, L. W. Cox and M. Hay (2002). GEM Global 2002 Executive Report, www.gemconsortium.org/ Reynolds, P. D., N. M. Carter, W. B. Gartner and P. G. Greene (2004). The prevalence of nascent entrepreneurs in the United States: evidence from the Panel Study of Entrepreneurial Dynamics, Small Business Economics, 23, pp. 263–84 Reynolds, P. D., N. Bosma, E. Autio, S. Hunt, N. De Bono and I. Servais (2005). Global Entrepreneurship Monitor: Data Collection Design and Implementation 1998–2003, Small Business Economics, 24, pp. 233–47 Rhyne, E. H. (1988). Small Business, Banks and SBA Loan Guarantees, New York, Quorum Books Riding, A. L. and G. Haines Jr. (2001). Loan guarantees: costs of default and benefits to small firms, Journal of Business Venturing, 16, pp. 595–612 Riding, A. L. and C. S. Swift (1990). Women business owners and terms of credit: some empirical findings of the Canadian experience, Journal of Business Venturing, 5, pp. 327–40 Riding, A. L., J. Madill and G. Haines Jr. (2007). Incrementality of SME loan guarantees, Small Business Economics, 29(1–2), pp. 47–61

References

527

Riley, J. (1979). Testing the educational screening hypothesis, Journal of Political Economy, 87, pp. S227–52 Ringuedè, S. (1998). An efficiency wage model for small firms: firm size and wages, Economics Letters, 59, pp. 263–68 Rissman, E. R. (2003). Self-employment as an alternative to unemployment, Working Paper No. 2003– 34, Federal Reserve Bank of Chicago Ritsila, J. and H. Tervo (2002). Effects of unemployment on new firm formation: Micro-level panel data evidence from Finland, Small Business Economics, 19, pp. 31–40 Robb, A. M. and R. W. Fairlie (2007). Access to financial capital among US businesses: the case of African American firms, Annals of the American Academy of Political and Social Science, 613, pp. 47–72 Robb, A. M. and J. Wolken (2002). Firm owner and financing characteristics: differences between male and female-owned small businesses, Working paper, Federal Board Reserve Board of Governors, USA Robbins, D. K., L. J. Pantuosco, D. F. Parker and B. K. Fuller (2000). An empirical assessment of the contribution of small business to US state economic performance, Small Business Economics, 15, pp. 293–302 Robert, P. and E. Bukodi (2004). Winners or losers? Entry and exit into self-employment in Hungary: 1980s and 1990s, in Arum and Müller (eds) (2004), pp. 245–76 Roberts, E. B. (1991). Entrepreneurs in High Technology, Oxford University Press Robinson, P. B. and E. A. Sexton (1994). The effect of education and experience on self-employment success, Journal of Business Venturing, 9, pp. 141–56 Robles, B. J. and H. Cordero-Guzman (2007). Latino self-employment and entrepreneurship in the United States: an overview of the literature and data sources, The Annals of the American Academy of Political and Social Science, 613, pp. 18–31 Robson, M. T. (1991). Self-employment and new firm formation, Scottish Journal of Political Economy, 38, pp. 352–68 (1996). Macroeconomic factors in the birth and death of UK firms: evidence from quarterly VAT registrations, Manchester School, 64, pp. 170–88 (1997). The relative earnings from self and paid employment: a time series analysis for the UK, Scottish Journal of Political Economy, 44, pp. 502–18 (1998a). Self-employment in the UK regions, Applied Economics, 30, pp. 313–22 (1998b). The rise in self-employment amongst UK males, Small Business Economics, 10, pp. 199–212 (2003). Does stricter employment protection legislation promote self-employment? Small Business Economics, 21, pp. 309–319 Robson, M. T. and C. Wren (1999). Marginal and average tax rates and the incentive for self-employment, Southern Economic Journal, 65, pp. 757–73 Robson, P. J. A. and R. J. Bennett (2000). SME growth: the relationship between business advice and external collaboration, Small Business Economics, 15(3), pp. 193–208 Robson, P. J. A., S. L. Jack and M. S. Freel (2008). Gender and the use of business advice: evidence from firms in the Scottish service sector, Environment & Planning C, 26, pp. 292–314 Rocha, H. O. and R. Sternberg (2005). Entrepreneurship: the role of clusters, theoretical perspectives and empirical evidence from Germany, Small Business Economics, 24, pp. 267–92 Rodriguez-Rodriguez, O. M. (2006). Trade credit in small and medium size firms: An application of the system estimator with panel data, Small Business Economics, 27, pp. 103–26 Rogerson, R. and P. Rupert (1993). On testing the intertemporal substitution theory of labour supply, Journal of Economic Dynamics and Control, 17, pp. 37–50

528

References

Romanelli, E. (1989). Environments and strategies of organisation start-up: effects on early survival, Administrative Science Quarterly, 34, pp. 369–87 Romanov, D. (2006). The corporation as a tax shelter: evidence from recent Israeli tax changes, Journal of Public Economics, 90, pp. 1939–54 Romer, P. M. (1986). Increasing returns and long-run growth, Journal of Political Economy, 94(5), pp. 1002–37 Ronstadt, R. (1986). Exit, stage left: why entrepreneurs end their entrepreneurial careers before retirement, Journal of Business Venturing, 1, pp. 323–38 Rosa, P. (1998). Entrepreneurial processes of business cluster formation and growth by ‘habitual’ entrepreneurs, Entrepreneurship Theory & Practice, 22(4), pp. 43–61 Rosa, P., S. Carter and D. Hamilton (1996). Gender as a determinant of small business performance: insights from a British study, Small Business Economics, 8, pp. 463–78 Rose-Ackerman, S. (1996). Altruism, nonprofits, and economic theory, Journal of Economic Literature, 34, pp. 701–28 Rosen, H. and P. Willen (2002). Risk, return and self-employment, Mimeo, University of Chicago / Princeton University Rosen, S. (1981). The economics of superstars, American Economic Review, 71, pp. 845–58 (1997). Austrian and neoclassical economics: any gains from trade? Journal of Economic Perspectives, 11(4), pp. 139–52 Rosenberg, N. (2003). America’s entrepreneurial universities, in D. M. Hart (ed.), The Emergence of Entrepreneurship Policy, Cambridge University Press, pp. 113–37 Rosenthal, S. S. and W. C. Strange (2003). Geography, industrial organisation, and agglomeration, Review of Economics and Statistics, 85, pp. 377–93 Ross, S. A. (1977). The determination of financial structure: the incentive-signalling approach, Bell Journal of Economics, 8, pp. 23–40 Rossi, S. E. (1998). Credit guarantee schemes: the case of Italy, in Storey (ed.) (2000), pp. 693–710 Rosti, L. and F. Chelli (2005). Gender discrimination, entrepreneurial talent and self-employment, Small Business Economics, 24, pp. 131–42 Rotefoss, B. and L. Kolvereid (2005). Aspiring, nascent and fledgling entrepreneurs: An investigation of the business start-up process, Entrepreneurship & Regional Development, 17, pp. 109–27 Rothaermel, F. T., S. D. Agung and L. Jiang (2007). University entrepreneurship: a taxonomy of the literature, Industrial & Corporate Change, 16(4), pp. 691–791 Rothschild, M. and J. E. Stiglitz (1971). Increasing risk: II. Its consequences, Journal of Economic Theory, 3, pp. 66–84 Rothwell, R. (1984). The role of small firms in the emergence of new technologies, Omega, 12(1), pp. 19–29 Rotter, J. B. (1982). The Development and Applications of Social Learning Theory: Selected Papers, New York, Praeger Roy, A. D. (1951). Some thoughts on the distribution of earnings, Oxford Economic Papers, 3, pp. 135–46 Rubery, J., J. Earnshaw and B. Burchell (1993). New Forms and Patterns of Employment: The Role of Self-employment in Britain, Baden-Baden, Nomos Verlagsgesellschaft Ruef, M., H. E. Aldrich and N. M. Carter (2003). The structure of organizational founding teams: homophily, strong ties, and isolation among US entrepreneurs, American Sociological Review, 68(2), pp. 195–222 Sahlman, W. A. (1990). The structure and governance of venture-capital organisations, Journal of Financial Economics, 27, pp. 473–521 Sanders, J. M. and V. Nee (1996). Immigrant self-employment: the family as social capital and the value of human capital, American Sociological Review, 61, pp. 231–49

References

529

Sanfey, P. and U. Teksoz (2007). Does transition make you happy? Economics of Transition, 15(4), pp. 707–31 Santarelli, E. (1998). Start-up size and post-entry performance: the case of tourism services in Italy, Applied Economics, 30(2), pp. 157–63 (2000). The duration of new firms in banking: an application of Cox regression analysis, Empirical Economics, 25, pp. 315–25 Santarelli, E. and E. Pesciarelli (1990). The emergence of a vision: the development of Schumpeter’s theory of entrepreneurship, History of Political Economy, 22(4), pp. 677–96 Santarelli, E. and M. Vivarelli (2007). Entrepreneurship and the process of firms’ entry, survival and growth, Industrial & Corporate Change, 16(3), pp. 455–88 Sapienza, H. J., D. De Clercq and W. R. Sandberg (2005). Antecedents of international and domestic learning effort, Journal of Business Venturing, 20(4), pp. 437–57 Sarasvathy, D. K., H. A. Simon and L. Lave (1998). Perceiving and managing business risks: differences between entrepreneurs and bankers, Journal of Economic Behavior & Organization, 33, pp. 207–25 Sarkar, M. B., R. Echambadi, R. Agarwal and B. Sen (2006). The effect of the innovative environment on exit of entrepreneurial firms, Strategic Management Journal, 27(6), pp. 519–39 Saxenian, A. (1994). Regional Advantage: Culture and Competition in Silicon Valley and Route 128, Cambridge MA, Harvard University Press Say, J. B. (1828). Cours Complet d’Economie Politique Practique, Paris SBA: See Small Business Administration Scase, R. and R. Goffee (1982). The Entrepreneurial Middle Class, London, Croom Helm Schade, C. and P. Koellinger (2007). Heuristics, biases, and the behaviour of entrepreneurs, in M. Minniti (ed.), Entrepreneurship: The Engine of Growth, Volume 1, Westport CT, Praeger, pp. 41–63 Schaffner, J. A. (1993). Rising incomes and the shift from self-employment to firm-based production, Economics Letters, 41, pp. 435–40 Scherer, F. M. (1980). Industrial Market Structure and Economic Performance, Chicago, Rand McNally College Publishing (1984). Innovation and Growth: Schumpeterian Perspectives, Cambridge MA, MIT Press (1991). Changing perspectives on the firm size problem, in Z. J. Acs and D. B. Audretsch (eds.), Innovation and Technological Change, Ann Arbor MI, University of Michigan Press, pp. 24–38 Schieman, S., Y. K. Whitestone and K. van Gundy (2006). The nature of work and the stress of higher status, Journal of Health and Social Behavior, 47, pp. 242–57 Schiller, B. R. and P. E. Crewson (1997). Entrepreneurial origins: a longitudinal inquiry, Economic Inquiry, 35, pp. 523–31 Schjoedt, L. and K. G. Shaver (2007). Deciding on an entrepreneurial career: a test of the pull and push hypotheses using the PSED data, Entrepreneurship Theory & Practice, 31(5), pp. 733–52 Schmidt, K. (2003). Convertible securities and venture capital finance, Journal of Finance, 43, pp. 1139–66 Schmidt-Mohr, U. (1997). Rationing versus collateralisation in competitive and monopolistic credit markets with asymmetric information, European Economic Review, 41, pp. 1321–42 Schmitz, J. A. (1989). Imitation, entrepreneurship and long-run growth, Journal of Political Economy, 97, pp. 721–39 Scholten, U. (2000). Rotating savings and credit associations in developed countries: the German– Austrian Bausparkassen, Journal of Comparative Economics, 28(2), pp. 340–63 Schreiner, M. and J. Morduch (2001). Replicating microfinance in the United States: opportunities and challenges, in J. Carr and Z.-Y. Tong (eds.), Replicating Microfinance in the United States, Washington DC, Fannie Mae Foundation, pp. 19–64

530

References

Schreiner, M. and G. Woller (2003). Microenterprise development programs in the United States and in the developing world, World Development, 31, pp. 1567–80 Schubert, R., M. Brown, M. Gyster and H. W. Brachinger (1999). Financial decision-making: are women really more risk-averse? American Economic Review, Papers and Procedings, 89(2), pp. 381–85 Schuetze, H. J. (2000). Taxes, economic conditions and recent trends in male self-employment: a Canada–US comparison, Labour Economics, 7(5), pp. 507–44 (2002). Profiles of tax non-compliance among the self-employed in Canada: 1969 to 1992, Canadian Public Policy, 28(2), pp. 219–38 (2006). Income splitting among the self-employed, Canadian Journal of Economics, 39(4), pp. 1195–1220 Schuetze, H. J. and H. Antecol (2006). Immigration, entrepreneurship and the venture start-up process, in S. C. Parker (ed.), The Life Cycle of Entrepreneurial Ventures, New York, Springer, pp. 107–35 Schuetze, H. J. and D. Bruce (2004). Tax policy and entrepreneurship, Swedish Economic Policy Review, 11(2), pp. 233–65 Schultz, T. P. (1990). Women’s changing participation in the labour force: a world perspective, Economic Development & Cultural Change, 38, pp. 457–88 Schultz, T. W. (1980). Investment in entrepreneurial ability, Scandinavian Journal of Economics, 82, pp. 437–48 Schumpeter, J. A. (1934). The Theory of Economic Development, Cambridge MA, Harvard University Press (1939). Business Cycles, New York, McGraw-Hill (1947). The creative response in economic history, Journal of Economic History, 7, pp. 149–59 Schutjens, V. and E. Wever (2000). Determinants of new firm success, Papers in Regional Science, 79, pp. 135–59 Scott, J. A. (2006). Loan officer turnover and credit availability in small firms, Journal of Small Business Management, 44(4), pp. 544–62 Scott, J. A. and T. C. Smith (1986). The effect of the Bankruptcy Reform Act of 1978 on small business loan pricing, Journal of Financial Economics, 16, pp. 119–40 Secrieru, O. and M. Vigneault (2004). Public venture capital, occupational choice, and entrepreneurship, B. E. Topics in Economic Analysis and Policy, 4, pp. 1–22 Segal, U. and A. Spivak (1989). Firm size and optimal growth rates, European Economic Review, 33, pp. 159–67 Segarra, A. and M. Callejon (2002). New firms’ survival and market turbulence: new evidence from Spain, Review of Industrial Organization, 20, pp. 1–14 Segerstrom, P. (1991). Innovation, imitation and economic growth, Journal of Political Economy, 99, pp. 190–207 Segerstrom, P., T. C. Anant and E. Dinopoulos (1990). A Schumpeterian model of the product life cycle, American Economic Review, 80, pp. 1077–91 Selden, J. (1999). Small and medium enterprises: their role in the economy, Labour Market Trends, 107, pp. 543–50 Sexton, D. L. and N. Bowman (1985). The entrepreneur: a capable executive and more, Journal of Business Venturing, 1, pp. 29–40 Shaffer, S. (2006). Establishment size and local employment growth, Small Business Economics, 26(5), pp. 439–54 Shane, S. (1996). Explaining variation in rates of entrepreneurship in the United States: 1899–1988, Journal of Management, 22, pp. 747–81 (2001). Technology regimes and new firm formation, Management Science, 47(9), pp. 1173–81 (2003). A General Theory of Entrepreneurship: The Individual-Opportunity Nexus, Cheltenham, Edward Elgar

References

531

Shane, S. and M. Foo (1999). New firm survival: institutional explanations for new franchisor mortality, Management Science, 45, pp. 142–59 Shane, S. and T. Stuart (2002). Organisational endowments and the performance of university start-ups, Management Science, 48(1), pp. 154–70 Shane, S. and S. Venkataraman (2000). The promise of entrepreneurship as a field of research, Academy of Management Review, 25, pp. 217–26 Shapero, A. (1975). The displaced uncomfortable entrepreneur, Psychology Today, 9, pp. 83–88 Sharma, M. and M. Zeller (1997). Repayment performance in group-based credit programs in Bangladesh: an empirical analysis, World Development, 25(10), pp. 1731–42 Sharpe, S.A. (1990). Asymmetric information, bank lending, and implicit contracts: a stylised model of customer relationships, Journal of Finance, 45, pp. 1069–87 Shaver, K. G. and L. R. Scott (1991). Person, process, choice: the psychology of new venture creation, Entrepreneurship Theory & Practice, 16, pp. 23–45 Shepherd, D. A. (2003). Learning from business failure: propositions of grief recovery for the selfemployed, Academy of Management Review, 28(2), pp. 318–28 Shepherd, W. G. (1984). ‘Contestability’ vs. competition, American Economic Review, 74, pp. 572–87 Sherer, P. D., N. Rogovsky and N. Wright (1998). What drives employment relationships in taxicab organisations? Linking agency to firm capabilities and strategic opportunities, Organization Science, 9(1), pp. 34–48 Sheshinski, E. and J. Drèze (1976). Demand fluctuations, capacity utilisation, and costs, American Economic Review, 66, pp. 731–42 Shirey, M. R. (2006). Entrepreneurship: Is it in your DNA? Clinical Nurse Specialist, 20(3), 130–33 Shorrocks, A. F. (1988). Wealth holdings and entrepreneurial activity, in D. Kessler and A. Masson (eds.), Modelling the Accumulation and Distribution of Wealth, Oxford, Clarendon Press, pp. 241–56 Shrader, R. and D. S. Siegel (2007). Assessing the relationship between human capital and firm performance: evidence from technology-based new ventures, Entrepreneurship Theory & Practice, 31(6), pp. 893–908 Shutt, J. and J. Sutherland (2003). Encouraging the transition into self-employment, Regional Studies, 37(1), pp. 97–103 Sillamaa, M.-A. and M. R. Veall (2001). The effect of marginal tax rates on taxable income: a panel study of the 1988 tax flattening in Canada, Journal of Public Economics, 80, pp. 341–56 Silva, O. (2007). The jack-of-all-trades entrepreneur: innate talent or acquired skill? Economics Letters, 97(2), pp. 118–23 Simmons, P. (1989). Bad luck and fixed costs in bankruptcy, Economic Journal, 99, pp. 92–107 Simon, H. A. and C. P. Bonini (1958). The size distribution of business firms, American Economic Review, 48, pp. 607–17 Simon, M. and S. M. Houghton (2003). The relationship between overconfidence and the introduction of risky products: evidence from a field study, Academy of Management Journal, 46(2), pp. 139–49 Simon, M., S. M. Houghton and K. Aquino (2000). Cognitive biases, risk perception, and venture formation – implications of interfirm (mis)perceptions for strategic decisions, Journal of Business Venturing, 15, pp. 113–34 Sine, W. D., R. J. David and H. Mitsuhashi (2007). From plan to plant: effects of certification on operational start-up in the emergent independent power sector, Organization Science, 18(4), pp. 578–94 Sinn, H. W. (1995). Social insurance, incentives and risk taking, International Tax and Public Finance, 3, pp. 259–80

532

References

Skuras, D., N. Meccheri, M. B. Moreira, J. Rosell and S. Stathopolou (2005). Entrepreneurial human capital accumulation and the growth of rural businesses, Journal of Rural Studies, 21, pp. 67–79 Sleeper, S. (1998). The role of firm capabilities in the evolution of the laser industry: the making of a high-tech market, Carnegie Mellon University, PhD dissertation Slemrod, J. (2007). Cheating ourselves: the economics of tax evasion, Journal of Economic Perspectives, 21(1), pp. 25–48 Slovin, M B. and M. E. Sushka (1983). A model of the commercial loan rate, Journal of Finance, 38, pp. 1583–96 Sluis, J. van der, M. van Praag and W. Vijverberg (2003). Entrepreneurship, selection and performance: a meta-analysis of the impact of education in industrialised countries, Tinbergen Institute Discussion Paper 04–046/3, Amsterdam (2005). Entrepreneurship, selection and performance: a meta-analysis of the impact of education in less developed countries, World Bank Economic Review, 19, pp. 225–61 Sluis, J. van der, M. van Praag and A. van Witteloostuijn (2004). The returns to education: a comparative study between entrepreneurs and employees, Tinbergen Institute Discussion Paper 04–104/3, Amsterdam Small Business Administration (1986, 1997, 2001). The State of Small Business, Washington DC, Office of Advocacy (1996). White Paper on the Financing of Small Business, Washington DC, Office of Advocacy (2002a). Small business’s share of economic growth, Research Summary, 211, Washington, DC (2002b). The influence of R & D expenditures on new firm formation and economic growth, Research Summary, 222, Washington DC (2003). CHI Research Inc, ‘Small serial innovators: the small firm contribution to technical change’, Haddon Heights, NJ 08035, 2002, 32 pages Smallbone, D., M. Ram, D. Deakins and R. Baldock (2003). Access to finance by ethnic minority businesses in the UK, International Small Business Journal, 21(3), pp. 291–314 Smirnov, V. and A. Wait (2007). Staged financing with a variable return, The B.E. Journal of Theoretical Economics: Contributions, 7(1), Article 5 Smith, A. (1937) The Wealth of Nations, New York, Random House Smith, B. D. and M. J. Stutzer (1989). Credit rationing and government loan programs: a welfare analysis, AREUEA Journal, 17, pp. 177–93 Smith, R. L. and J. K. Smith (2003). Entrepreneurial Finance, Second Edition, New York, John Wiley & Sons Smith, S. (1986). Britain’s Shadow Economy, Oxford, Clarendon Press Smith, S. C. (2002). Village banking and maternal and child health: evidence from Ecuador and Honduras, World Development, 30, pp. 707–23 Sobel, R. S. (2008). Testing Baumol: institutional quality and the productivity of entrepreneurship, Journal of Business Venturing, 23, pp. 641–55 Sobel, R. S. and K. A. King (2008). Does school choice increase the rate of youth entrepreneurship? Economics of Education Review, 27(4), pp. 429–38 Sobel, R. S., J. R. Clark and D. R. Lee (2005). Freedom, barriers to entry, entrepreneurship and economic progress, Manuscript http://www.be.wvu.edu/ec/Papers/SobelClarkLee.pdf Sohl, J. E. (1999). The early-stage equity market in the USA, Venture Capital, 1, pp. 101–20 (2003). The private equity market in the USA: lessons from volatility, Venture Capital, 5(1), pp. 29–46 Sørensen, J. B. (2006). Closure vs. exposure: mechanisms in the intergenerational transmission of self-employment, in M. Ruef and M. Lounsbury (eds.), Research in the Sociology of Organisations, Amsterdam, JAI Press

References

533

(2007). Bureaucracy and entrepreneurship: workplace effects on entrepreneurial entry, Administrative Science Quarterly, 52(3), pp. 387–412 Sørensen, J. B. and T. E. Stuart (2000). Aging, obsolescence, and organizational innovation, Administrative Science Quarterly, 45(1), pp. 81–112 Sorenson, O. and T. Stuart (2001). Syndication networks and the spatial distribution of venture capital investments, American Journal of Sociology, 106, pp. 1546–86 Sowell, T. (1981). Markets and Minorities, New York, Basic Books Spear, R. (1992). The co-operative spirit: some UK evidence, in P. Leighton and A. Felstead (eds.), The New Entrepreneurs: Self-employment and Small Business in Europe, London, Kogan Page, pp. 325–41 Spencer, A. S., B. A. Kirchhoff and C. White (2008). Entrepreneurship, innovation and wealth distribution – the essence of creative destruction, International Small Business Journal, 26(1), pp. 9–26 Spilling, O. R. (1996). Regional variation of new firm formation: the Norwegian case, Entrepreneurship and Regional Development, 8, pp. 217–43 Srinivasan, R., G. L. Lilien and A. Rangaswamy (2008). Survival of high-tech firms: the effects of diversity of product-market portfolios, patents and trademarks, International Journal of Research in Marketing, 25, 119–28 Staber, U. and D. Bogenhold (1993). Self-employment: a study of seventeen OECD countries, Industrial Relations Journal, 24, pp. 126–37 Stabile, M. (2004). Payroll taxes and the decision to be self-employed, International Tax and Public Finance, 11, pp. 31–53 Stam, E. (2007). Why butterflies don’t leave: locational behaviour of entrepreneurial firms, Economic Geography, 83(1), pp. 27–50 Stam, E., D. B.Audretsch and J. Meijaard (2008). Renascent entrepreneurship, Journal of Evolutionary Economics, 18, pp. 493–507 Stanworth, J., D. Purdy, S. Price and N. Zafiris (1998). Franchise versus conventional small business failure rates in the US and UK: more similarities than differences, International Small Business Journal, 16, pp. 56–69 Stearns, T. M., N. M. Carter and P. D. Reynolds (1995). New firm survival: industry, strategy and location, Journal of Business Venturing, 10(1), pp. 23–42 Steen, E. van den (2004). Rational over-optimism (and other biases), American Economic Review, 94(4), pp. 1141–51 Steinmetz, G. and E. O. Wright (1989). The fall and rise of the petty bourgeoisie: changing patterns of self-employment in the postwar United States, American Journal of Sociology, 94, pp. 973–1018 Stel, A. van (2005). COMPENDIA: Harmonizing business ownership data across countries and over time, International Entrepreneurship and Management Journal, 1, pp. 105–23 Stel, A. van and M. Carree (2004). Business ownership and sectoral growth: an empirical analysis of 21 OECD countries, International Small Business Journal, 22, pp. 389–419 Stel, A. van and D. J. Storey (2004). The link between firm births and job creation: is there a Upas Tree Effect? Regional Studies, 38(8), pp. 893–909 Stel, A. van and K. Suddle (2008). The impact of new firm formation on regional development in the Netherlands, Small Business Economics, 30, pp. 31–47 Stel, A. van, M. Carree and A. R. Thurik (2005). The effect of entrepreneurial activity on national economic growth, Small Business Economics, 24, pp. 311–21 Stel, A. van, D. J. Storey and A. R. Thurik (2007). The effect of business regulations on nascent and young business entrepreneurship, Small Business Economics, 28(2–3), pp. 171–86 Stephan, P. (1996). The economics of science, Journal of Economic Literature, 34, pp. 1199–1235

534

References

Steuerle, C. E. and V. Hodgkinson (1999). Meeting social needs: comparing the resources of the independent sector and government, in E. Boris and C. E. Steuerle (eds.), Nonprofits and Government, Washington DC, Urban Institute Press Stewart, W. H. Jr and P. L. Roth (2001). Risk propensity differences between entrepreneurs and managers: a meta-analytic review, Journal of Applied Psychology, 86(1), pp. 145–53 Stiglitz, J. E. (1970). The effects of income, wealth and capital gains taxation on risk taking, Quarterly Journal of Economics, 83, pp. 263–83 (1990). Peer monitoring and credit markets, World Bank Economic Review, 4, pp. 351–66 Stiglitz, J. E. and A. Weiss (1981). Credit rationing in markets with imperfect information, American Economic Review, 71, pp. 393–410 (1983). Incentive effects of terminations: applications to the credit and labour markets, American Economic Review, 73, pp. 912–27 (1992).Asymmetric information in credit markets and its implications for macro-economics, Oxford Economic Papers, 44, pp. 694–724 Stoll, H. R. (1984). Small firms’ access to public equity financing, in P. M. Horvitz and R. R. Pettit (eds.), Small Business Finance: Sources of Financing for Small Business, Part A, Greenwich CT, JAI Press, pp. 177–238 Storey, D. J. (1982). Entrepreneurship and the New Firm, London, Croom Helm (1991). The birth of new firms – does unemployment matter? A review of the evidence, Small Business Economics, 3, pp. 167–78 (1994a). Understanding the Small Business Sector, London, Routledge (1994b). The role of legal status in influencing bank financing and new firm growth, Applied Economics, 26, pp. 129–36 (2000). Six steps to heaven: evaluating the impact of public policies to support small businesses in developed economies, in H. Landstrom and D. L. Sexton (eds.), Handbook of Entrepreneurship, Oxford, Blackwell (2003). Entrepreneurship, small and medium sized enterprises and public policies, in Z. J. Acs and D. B. Audretsch (eds.), Handbook of Entrepreneurship Research: An Interdisciplinary Survey and Introduction, Boston, Kluwer, pp. 473–511 (2006). Evaluating SME policies and programmes: Technical and political dimensions, in M. Casson, B. Yeung, A. Basu and N. Wadeson (eds.), The Oxford Handbook of Entrepreneurship, Oxford University Press, pp. 248–78 Storey, D. J. and A. M. Jones (1987). New firm formation – a labour market approach to industrial entry, Scottish Journal of Political Economy, 34, pp. 37–51 Storey, D. J. and P. Wynarczyk (1996). The survival and non-survival of micro-firms in the UK, Review of Industrial Organization, 11, pp. 211–29 Storey, D. J., K. Keasey, R. Watson and P. Wynarczyk (1987). The Performance of Small Firms, London, Croom Helm Straume, O. R. (2006). Managerial delegation and merger incentives with asymmetric costs, Journal of Institutional & Theoretical Economics, 162(3), pp. 450–69 Strotmann, H. (2007). Entrepreneurial survival, Small Business Economics, 28, pp. 87–104 Stuart, T. E. and W. W. Ding (2006). When do scientists become entrepreneurs? The social structural antecedents of commercial activity in the academic life sciences, American Journal of Sociology, 112(1), pp. 97–144 Stuart, T. E. and O. Sorenson (2003a). The geography of opportunity: spatial heterogeneity in founding rates and the performance of biotechnology firms, Research Policy, 32, pp. 229–53 (2003b). Liquidity events and the geographic distribution of entrepreneurial activity, Administrative Science Quarterly, 48, pp. 175–201 Subramanian, N. (2005). The economics of intrapreneurial innovation, Journal of Economic Behavior & Organization, 58, pp. 487–510 Sutton, J. (1997). Gibrat’s legacy, Journal of Economic Literature, 35, pp. 40–59

References

535

Swanson, D. and L. Webster (1992). Private sector manufacturing in the Czech and Slovak Republics: a survey of firms, Washington DC, The World Bank Tajnikar, M., P. Dosenovic Bonca and L. Zajec (2006). Harvesting in high growth firms, in S. C. Parker (ed.), The Life Cycle of Entrepreneurial Ventures, New York, Springer, pp. 533–82 Takii, K. (2008). Fiscal policy and entrepreneurship, Journal of Economic Behavior & Organization, 65, pp. 592–608 Tamásy, C. (2006). Determinants of regional entrepreneurship dynamics in contemporary Germany: a conceptual and empirical analysis, Regional Studies, 40(4), pp. 365–84 Tamásy, C. and R. Le Heron (2008). The geography of firm formation in New Zealand, Tijdschrift voor Economische en Sociale Geografie, 99(1), pp. 37–52 Taniguchi, H. (2002). Determinants of women’s entry into self-employment, Social Science Quarterly, 83, pp. 875–93 Taris, T. W., S. A. E. Guerts, W. B. Schaufeli, R. W. B. Blonk and S. E. Lagerveld (2008). All day and all of the night: the relative contribution of two dimensions of workaholism to well-being in self-employed workers, Work and Stress, 22(1), pp. 153–65 Taylor, M. P. (1996). Earnings, independence or unemployment: why become self-employed?, Oxford Bulletin of Economics and Statistics, 58, pp. 253–66 (1999). Survival of the fittest? An analysis of self-employment duration in Britain, Economic Journal, 109, pp. C140–C155 (2001). Self-employment and windfall gains in Britain: evidence from panel data, Economica, 68, pp. 539–65 (2004). Self-employment in Britain: when, who and why? Swedish Economic Policy Review, 11(2), pp. 139–73 Taymaz, E. (2005). Are small firms really less productive? Small Business Economics, 25(5), pp. 429–45 Tervo, H. (2006). Regional unemployment, self-employment and family background, Applied Economics, 38, pp. 1055–62 (2008). Self-employment transitions and alternation in Finnish rural and urban labour markets, Papers in Regional Science, 87(1), pp. 55–76 Tervo, H., and H. Niittykangas (1994). The impact of unemployment on new firm formation in Finland, International Small Business Journal, 13, pp. 38–53 Tether, B. S., I. J. Smith and A. T. Thwaites (1997). Smaller enterprises an innovation in the UK: the SPRU Innovations Database revisited, Research Policy, 2, pp. 19–32 Thadden, E.-L. von (2004).Asymmetric information, bank lending and implicit contracts: the winner’s curse, Finance Research Letters, 1, pp. 11–23 Thakor, A. and R. Calloway (1983). Costly information production equilibria in the bank credit market with applications to credit rationing, Journal of Financial & Quantitative Analysis, 18, pp. 229–56 Thornhill, S. and R. Amit (2003). Learning about failure: Bankruptcy, firm age and the resource-based view, Organization Science, 14(5), pp. 497–509 Thornton, J. (1994). Estimating the choice behaviour of self-employed business proprietors: an application to dairy farmers, Southern Economic Journal, 60, pp. 579–95 (1998). The labour supply behaviour of self-employed solo practice physicians, Applied Economics, 30, pp. 85–94 Thornton, P. H. and K. H. Flynn (2003). Entrepreneurship, networks and geographies, in Z. J. Acs and D. B. Audretsch (eds.), Handbook of Entrepreneurship Research, Boston MA, Kluwer, pp. 401–33 Thurik, A. R. (1999). Entrepreneurship, industrial transformation and growth, in G. D. Libecap (ed.), The Sources of Entrepreneurial Activity, Vol. 11, Stamford CT, JAI Press, pp. 29–65 Thurik, A. R., M. A. Carree, A. van Stel and D. B. Audretsch (2008). Does self-employment reduce unemployment? Journal of Business Venturing, 23, pp. 673–86

536

References

Toole, A. A. and D. Czarnitzki (2007). Biomedical academic entrepreneurship through the SBIR program, Journal of Economic Behavior & Organization, 63, pp. 716–38 Torrini, R. (2005). Cross-country differences in self-employment rates: the role of institutions, Labour Economics, 12, pp. 661–83 Townsend, R. (1979). Optimal contracts and competitive markets with costly state verification, Journal of Economic Theory, 21, pp. 417–25 Trester, J. J. (1998). Venture capital contracting under asymmetric information, Journal of Banking and Finance, 22(6), pp. 675–99 Troske, K. R. (1999). Evidence on the employer size–wage premium from worker–establishment matched data, Review of Economics and Statistics, 81(1), pp. 15–26 Tubergen, F. van (2005). Self-employment of immigrants: a cross-national study of 17 Western societies, Social Forces, 84(2), pp. 709–32 Tucker, I. B. (1987). Use of the decomposition technique to test the educational screening hypothesis, Economics of Education Review, 4(4), pp. 321–26 (1988). Entrepreneurs and public-sector employees: the role of achievement motivation and risk in occupational choice, Journal of Economic Education, 19, pp. 259–68 Tushman, M. and P. Anderson (1986). Technological discontinuities and organisational environments, Administrative Science Quarterly, 31, pp. 439–65 Tykvová, T. (2007). What do economists tell us about venture capital contracts? Journal of Economic Surveys, 21(1), pp. 65–89 Ucbasaran, D., P. Westhead and M. Wright (2006). Habitual entrepreneurs, in Casson, M. et al (eds.), The Oxford Handbook of Entrepreneurship, Oxford University Press, pp. 461–83 Ueda, M. (2004). Banks versus venture capital: project evaluation, screening, and expropriation, Journal of Finance, 59, pp. 601–21 Uhlaner, L. and R. Thurik (2007). Postmaterialism influencing total entrepreneurial activity across nations, Journal of Evolutionary Economics, 17, pp. 161–85 Uusitalo, R. (2001). Homo entreprenaurus? Applied Economics, 33, pp. 1631–38 VandenHeuvel, A. and M. Wooden (1997). Self-employed contractors and job satisfaction, Journal of Small Business Management, 35, pp. 11–20 Varian, H. (1990). Monitoring agents with other agents, Journal of Institutional & Theoretical Economics, 146(2), pp. 153–74 Variyam, J. N. and D. S. Kraybill (1992). Empirical evidence on determinants of firm growth, Economics Letters, 38, pp. 31–36 Vercammen, J. (2002). Welfare-improving adverse selection in credit markets, International Economic Review, 43, pp. 1017–33 Verheul, I. and R. Thurik (2001). Start-up capital: ‘Does gender matter?’, Small Business Economics, 16, pp. 329–45 Verheul, I., A. van Stel and R. Thurik (2006). Explaining female and male entrepreneurship at the country level, Entrepreneurship & Regional Development, 18, pp. 151–83 Vesala, T. (2007). Financial matching, asymmetric information and entrepreneurial risk taking, Scandinavian Journal of Economics, 109(3), pp. 469–85 Vijverberg, W. P. M. (1986). Consistent estimates of the wage equation when individuals choose among income-earning activities, Southern Economic Journal, 52, pp. 1028–42 Vinogradov, E. and L. Kolvereid (2007). Cultural background, human capital and self-employment rates among immigrants in Norway, Entrepreneurship & Regional Development, 19(4), pp. 359–76 Vivarelli, M. (2004). Are all the potential entrepreneurs so good? Small Business Economics, 23, pp. 41–49 Vivarelli, M. and D. B. Audretsch (1998). The link between the entry decision and post-entry performance: evidence from Italy, Industrial & Corporate Change, 7, pp. 485–500

References

537

Wadeson, N. (2006). Cognitive aspects of entrepreneurship: decision-making and attitudes to risk, in M. Casson, B. Yeung,A. Basu and N. Wadeson (eds.), The Oxford Handbook of Entrepreneurship, Oxford University Press, pp. 91–113 Wadhwani, S. B. (1986). Inflation, bankruptcy, default premia and the stock market, Economic Journal, 96, pp. 120–38 Wagner, J. (1994). The post-entry performance of new small firms in German manufacturing industries, Journal of Industrial Economics, 42, pp. 141–54 (1995). Firm size and job creation in Germany, Small Business Economics, 7, pp. 469–74 (2003a). Testing Lazear’s jack-of-all-trades view of entrepreneurship with German micro data, Applied Economics Letters, 10, pp. 687–89 (2003b). Taking a second chance: entrepreneurial re-starters in Germany, Applied Economics Quarterly, 49, pp. 255–72 (2004). Are young and small firms hothouses for nascent entrepreneurs? Evidence from German micro data, Working Paper No. 301, University of Lüneburg (2006a). Are nascent entrepreneurs ‘Jacks-of-all-trades’? A test of Lazear’s theory of entrepreneurship with German data, Applied Economics, 38, pp. 2415–19 (2006b). Nascent entrepreneurs, in S. C. Parker (ed.), Handbook of Entrepreneurship Research, New York, Springer, pp. 15–37 Wagner, J. and R. Sternberg (2004). Start-up activities, individual characteristics, and the regional milieu, Annals of Regional Science, 38, 219–40 Waldinger, R. (1986). Through the Eye of the Needle: Immigrants and Enterprise in the New York Garment Trade, New York, NYU Press Waldinger, R., H. Aldrich and R. Ward (1990). Ethnic Entrepreneurs: Immigrant Business in Industrial Societies, Sage, London Wales, T. J. (1973). Estimation of a labour supply curve for self-employed business proprietors, International Economic Review, 14, pp. 69–80 Wall, H. J. (2004). Entrepreneurship and the deregulation of banking, Economics Letters, 82, pp. 333–39 Wallstein, S. J. (2000). The effects of government-industry R & D programs on private R & D: the case of the Small Business Innovation Research program, Rand Journal of Economics, 31(1), pp. 82–100 Walstad, W. and M. Kourilsky (1998). Entrepreneurial attitudes and knowledge of black youth, Entrepreneurship Theory & Practice, 23(2), pp. 5–18 Watkins, J. M. and D. S. Watkins (1984). The female entrepreneur: her background and determinants of business choice – some British data, in Frontiers of Entrepreneurship Research, Wellesley MA, Babson College Watson, H. (1984). Credit markets and borrower effort, Southern Economic Journal, 50, pp. 802–13 (1985). Tax evasion and labour markets, Journal of Public Economics, 27, pp. 231–46 Watson, J. (2002). Comparing the performance of male- and female-controlled businesses: relating outputs to inputs, Entrepreneurship Theory & Practice, 26(3), pp. 91–100 (2007). Modelling the relationship between networking and firm performance, Journal of Business Venturing, 22, pp. 852–74 Watson, J. and J. Everett (1993). Defining small business failure, International Small Business Journal, 11(3), pp. 35–48 Watson, J. and S. Robinson (2003). Adjusting for risk in comparing the performances of male- and female-controlled SMEs, Journal of Business Venturing, 18, pp. 773–88 Watson, R. (1990). Employment change, profits and directors’ remuneration in small and closely-held UK companies, Scottish Journal of Political Economy, 37, pp. 259–74 Webb, D. C. (1991). Long-term financial contracts can mitigate the adverse selection problem in project financing, International Economic Review, 32, pp. 305–20

538

References

Weber, M. (1930). The Protestant Ethic and the Spirit of Capitalism, New York, Scribner Weisbrod, B. A. (1988). The Nonprofit Economy, Cambridge MA, Harvard University Press Wellington, A. J. (2001). Health insurance coverage and entrepreneurship, Contemporary Economic Policy, 19, pp. 465–78 (2006). Self-employment: the new solution for balancing family and career? Labour Economics, 13, pp. 357–86 Wennberg, K. J., T. Folta and F. Delmar (2007). A real options model of stepwise entry into selfemployment, Mimeo, Small Business Administration, USA Wennekers, S. and R. Thurik (1999). Linking entrepreneurship and economic growth, Small Business Economics, 13, pp. 27–55 Wennekers, S., A. van Stel, R. Thurik, and P. Reynolds (2005). Nascent entrepreneurship and the level of economic development, Small Business Economics, 24, pp. 293–309 Wenner, M. D. (1995). Group credit: a means to improve information transfer and loan repayment performance, Journal of Development Studies, 32, pp. 263–81 Westhead, P. (1995). Survival and employment growth contrasts between types of owner-managed high-technology firms, Entrepreneurship Theory & Practice, 20, pp. 5–27 Westhead, P. and M. Cowling (1995). Employment change in independent owner-managed hightechnology firms in Great Britain, Small Business Economics, 7, pp. 111–40 Westhead, P. and C. Howarth (2006). Ownership and management issues associated with family firm performance and company objectives, Family Business Review, 19(4), 301–16 Westhead, P. and D. J. Storey (1997). Financial constraints on the growth of high-technology small firms in the United Kingdom, Applied Financial Economics, 7, pp. 197–201 Westhead, P. and M. Wright (1998). Novice, portfolio and serial founders: are they different? Journal of Business Venturing, 13, pp. 173–204 (1999). Contributions of novice, portfolio and serial founders located in rural and urban areas, Regional Studies, 33, pp. 157–73 Wette, H. (1983). Collateral in credit rationing in markets with imperfect information: note, American Economic Review, 73, pp. 442–45 Wetzel, W. E. (1987). The informal venture capital market: aspects of scale and market efficiency, Journal of Business Venturing, 2, pp. 299–313 White, L. J. (1982). The determinants of the relative importance of small business, Review of Economics and Statistics, 64, pp. 42–49 (1984). The role of small business in the US economy, in P. M. Horvitz and R. R. Pettit (eds.), Small Business Finance: Sources of Financing for Small Business, Part A, Greenwich CT, JAI Press, pp. 19–49 White, R. E., S. Thornhill and E. Hampson (2006). Entrepreneurs and evolutionary biology: The relationship between testosterone and new venture creation, Organizational Behaviour & Human Decision Processes, 100, pp. 21–34 (2007). A biosocial model of entrepreneurship: the combined effects of nurture and nature, Journal of Organizational Behaviour, 28, pp. 451–66 Wiggins, S. N. (1995). Entrepreneurial enterprises, endogenous ownership, and the limits to firm size, Economic Inquiry, 33, pp. 54–69 Wiklund, J. (2007). Small-firm growth strategies, in A. Zacharakis and S. Spinelli Jr. (eds.), Entrepreneurship: The Engine of Growth, Volume 2, Westport CT, Praeger, pp. 135–54 Wiklund, J. and D. Shepherd (2003). Aspiring for, and achieving growth: the moderating role of resources and opportunities, Journal of Management Studies, 40(8), pp. 1911–41 (2008). Portfolio entrepreneurship: habitual and novice founders, new entry, and mode of organising, Entrepreneurship Theory & Practice, 32, pp. 701–25 Williams, D. L. (1999). Why do entrepreneurs become franchisees? An empirical analysis of organisational choice, Journal of Business Venturing, 14, pp. 103–24

References

539

Williams, D. R. (2000). Consequences of self-employment for women and men in the United States, Labour Economics, 7, pp. 665–87 (2003). Returns to education and experience in self-employment: evidence from Germany, Journal of Applied Social Science Studies, 123, pp. 139–50 (2004). Youth self-employment: its nature and consequences, Small Business Economics, 23(4), pp. 323–36 Williamson, O. E. (1985). The Economic Institutions of Capitalism, New York, Free Press Williamson, S. D. (1986). Costly monitoring, financial intermediation and equilibrium credit rationing, Journal of Monetary Economics, 18, pp. 159–79 (1987). Costly monitoring, loan contracts and equilibrium credit rationing, Quarterly Journal of Economics, 102, pp. 135–45 Wilson, F., S. Carter, S. Tagg, E. Shaw and W. Lam (2007). Bank loan officers’ perceptions of business owners: the role of gender, British Journal of Management, 18, pp. 154–71 Winborg, J. and H. Landström (2000). Financial bootstrapping in small businesses: examining small business managers’ resource acquisition behaviours, Journal of Business Venturing, 16, pp. 235–54 Winter, S. G. (1984). Schumpeterian competition in alternative technological regimes, Journal of Economic Behavior & Organization, 5, pp. 287–320 Winton, A. and V. Yerramilli (2008). Entrepreneurial finance: banks versus venture capital, Journal of Financial Economics, 88, pp. 51–79 Wit, G. de (1993). Models of self-employment in a competitive market, Journal of Economic Surveys, 7, pp. 367–97 Wit, G. de and F. van Winden (1989). An empirical analysis of self-employment in the Netherlands, Small Business Economics, 1, pp. 263–72 (1990). An empirical analysis of self-employment in the Netherlands, Economics Letters, 32, pp. 97–100 (1991). An M-sector, N-group behavioural model of self-employment, Small Business Economics, 3, pp. 49–66 Wolpin, K. I. (1977). Education and screening, American Economic Review, 67, pp. 949–58 Wong, A. (2002). Angel finance: the other venture capital, Working Paper, University of Chicago Wong, L. and M. Ng (2002). The emergence of small transnational enterprise in Vancouver: the case of Chinese immigrant entrepreneurs, International Journal of Urban and Regional Research, 26(3), pp. 508–30 Wong, P-K., Y. P. Ho and E. Autio (2005). Entrepreneurship, innovation and economic growth: evidence from GEM data, Small Business Economics, 24, pp. 335–50 Wong, Y-C. (1986). Entrepreneurship, marriage, and earnings, Review of Economics and Statistics, 68, pp. 693–99 Woodruff, C. and R. Zenteno (2007). Migration networks and microenterprises in Mexico, Journal of Development Economics, 82, pp. 509–28 Wren, C. (1998). Subsidies for job creation: is small best? Small Business Economics, 10, pp. 273–81 Wren, C. and D. J. Storey (2002). Evaluating the effect of ‘soft’ business support upon small firms’ performance, Oxford Economic Papers, 54, pp. 334–65 Wright, M., K. Robbie and C. Ennew (1997). Venture capitalists and serial entrepreneurs, Journal of Business Venturing, 12, pp. 227–49 Wright, M., P. Westhead and J. Sohl (1998). Editors’ introduction: habitual entrepreneurs and angel investors, Entrepreneurship Theory & Practice, 22, pp. 5–21 Wu, B. and A. M. Knott (2006). Entrepreneurial risk and market entry, Management Science, 52(9), pp. 1315–30 Wu, X. (2006). Communist cadres and market opportunities: entry into self-employment in China, 1978–1996, Social Forces, 85(1), pp. 389–411

540

References

Wu, Z., J. H. Chua and J. J. Chrisman (2007). Effects of family ownership and management on small business equity financing, Journal of Business Venturing, 22, pp. 875–95 Wunnava, P. V. and B. T. Ewing (2000). Union–nonunion gender wage and benefit differentials across establishment sizes, Small Business Economics, 15(1), pp. 47–57 Wydick, B. (1999). Can social cohesion be harnessed to repair market failures? Evidence from group lending in Guatemala, Economic Journal, 109, pp. 463–75 Wynne, J. (2005). Wealth as a determinant of comparative advantage, American Economic Review, 95(1), pp. 226–55 Xu, B. (1998). A re-estimation of the Evans–Jovanovic entrepreneurial choice model, Economics Letters, 58, pp. 91–95 Yamada, G. (1996). Urban informal employment and self-employment in developing countries: theory and evidence, Economic Development and Cultural Change, 44, pp. 289–314 Yamawaki, H. (1990). The effects of business conditions on net entry: evidence from Japan, in P. A. Geroski and J. Schwalbach (eds.), Entry and Market Contestability: An International Comparison, Oxford, Basil Blackwell Yang, D. (2008). International migration, remittances and household investment: evidence from Philippine migrants’ exchange rate shocks, Economic Journal, 118, pp. 591–630 Yoon, I.-J. (1991). The changing significance of ethnic and class resources in immigrant business, International Migration Review, 25, pp. 303–31 (1997). On My Own: Korean Business and Race Relations in America, University of Chicago Press Yuengert, A. M. (1995). Testing hypotheses of immigrant self-employment, Journal of Human Resources, 30, pp. 194–204 Zacharakis, A. and M. Eckermann (2007). Venture capital financing, in A. Zacharakis and S. Spinelli Jr. (eds.), Entrepreneurship: The Engine of Growth, Volume 2, Westport CT, Praager, pp. 109–34 Zazzaro,A. (2005). Should courts enforce credit contracts strictly? Economic Journal, 115, pp. 166–84 Zeller, M. (1998). Determinants of repayment performance in credit groups: the role of program design, intra-group risk pooling and social cohesion, Economic Development & Cultural Change, 46(1), pp. 599–620 Zhang, J., L. X. Zhang, S. Rozelle and S. Boucher (2006). Self-employment with Chinese characteristics: the forgotten engine of rural China’s growth, Contemporary Economic Policy, 24(3), pp. 446–58 Zhang, J., V. Souitaris, P. H. Soh and P. K. Wong (2008). A contingent model of network utilisation in early financing of technology ventures, Entrepreneurship Theory & Practice, 32(4), pp. 593–613 Zhao, H. and S. E. Seibert (2006). The big five personality dimensions and entrepreneurial status: a meta-analytical review, Journal of Applied Psychology, 91(2), pp. 259–71 Zissimopoulos, J. M. and L. A. Karoly (2007). Transitions to self-employment at older ages: the role of wealth, health, health insurance and other factors, Labour Economics, 14, pp. 269–95 Zook, C. and J. Allen (1999). The Facts About Growth, New York, Bain & Co Zucker, L., M. Darby and J. Armstrong (1998). Intellectual human capital and the birth of US biotechnology enterprises, American Economic Review, 88, pp. 290–306

Index

2SE (two-step estimator) 95 2SLS (two-stage least squares) 89 3Com 64 achievement motive 129 Acs, Zoltan 293 active learning 356 additionality 408, 415 Adobe Systems 64 adolescence 389–90, 395 affirmative action schemes 440 age of entrepreneur ageing and retirement 351–5 as determinant of entrepreneurship 113–15 as determinant of nascent entrepreneurship 152 as determinant of venture growth 319 agency costs, equity finance 245 agents 52 agglomeration benefits 149 aggregate wealth accumulation, entrepreneurs’ role in 264–5 Aghion and Bolton (1997) model 69, 216, 272 agreeableness 130 altruism, and non-profit enterprises 59 Amazon 406 ambiguity tolerance 129 AMD 64 ‘artisanal’ development 68–9 Audretsch, David 293 Australia gender earnings gap 190, 191, 192 immigrant self-employment rates 177 nascent entrepreneurship 9 self-employment rates 16, 18 Australian Innovation Investment Fund programme 420 autonomy 110–13 BancoSol 252 Banerjee and Newman (1993) model 68–9, 271–2 bank competition 227, 352 bank-optimal interest rate 214 Bank Rayat 252

bankruptcy 446–8 banks over-optimism 225 screening 217–21 see also Bester’s (1985) screening model; due diligence see also investment banks BANs (business angel networks) 419 Baumol, William 1–2, 3–4 Bayesian learning see Fraser and Greene’s Bayh–Dole Act (1980) 424 benefits, of entrepreneurship to society 1 see also agglomeration benefits; retirement benefits Berger and Udell (1992) 282–3 Bernhardt’s (2000) model 228–9 Bester’s (1985) screening model 230–1 bias see hindsight bias; sample selection bias; status quo bias; survival bias binary choice models, for duration of spells in entrepreneurship 100 Birch, David 293 Birch Report 297 bivariate probit model 94–5 black-owned businesses credit 167–8 gazelles 176 performance 164 bootstrapping 277 Borjas (1986) decomposition 181–2 borrowing constraints 56 Botswana, self-employment rates 22 Branson, Sir Richard 376 bribery 445 Buffett, Warren 126–7 burn rate 243 business advice and assistance 440–2, 460 business angel networks (BANs) 419 business angels 195, 234, 245, 248–50, 423 business cycle, as determinant of entrepreneurship 142–3 business ownership, as measure of entrepreneurship 10

541

542

Index

business plans and new venture survival 153 and venture growth 321, 322 business stealing effect 407 Canada business angels 423 immigrant self-employment 177 job satisfaction 112 loan guarantee schemes 413 nascent entrepreneurship 9, 154 public policy 420 self-employment rates 16, 18 Cantillon, Richard 32 capital gains tax (CGT) 421–2 care homes, use of sedatives 59–60 Carlsson, Chester 64 Carroll et al.’s (2000a) model of labour demand and taxation 294–5, 306–7 category opening products 304 certification, by venture capitalists 240 CGT (capital gains tax) 421–2 Chapter 11 bankruptcy 446 Chapter 7 bankruptcy 446 children, presence of, and propensity to entrepreneurship 188–9 China, self-employment rates 23–4 Choice Dilemma Questionnaire 122 closed-end venture funds 235, 247 co-integrating vectors 97 co-operative finance schemes 257–9 Coase, R. 52 codified knowledge 73 collateral 205–6 as screening instrument 217–18 competing risks models 101 computer usage and business ownership 140 of entrepreneurs 95 conscientiousness 130 contractors, rationale of using (‘make or buy’ decision) 52–7 see also ‘hire or outsource’ model convertible contracts 242–4 corporate governance regulation 450 corporate venture capital 237–8 corporation tax (CT) 423, 436–7 correlations 86 corruption 339, 435, 445, 449 see also rent seeking ‘costly state verification’ argument 209 country effects 329 Cox proportional hazard model 101 CPS Monthly Household Labour Force Survey 17 creative destruction 34–5 credit co-operatives 257–8 credit markets asymmetric information 204 and intrapreneurship 66–7 black-owned businesses 167–8 discrimination against women 196

employer discrimination spillover 167 regulation 451–3 credit rationing 203 and profitability 280–1 Type I 209, 212 see also Bernhardt’s (2000) model Type II 209 evidence relating to 281–4 see also Stiglitz and Weiss’ (1981) model typology 209–11 see also Bester’s (1985) screening model credit scoring 207–8 cross-section binary models of occupational choice 90–2 extensions 92 heteroscedastic probit 96 inclusion of relative incomes 92–4 multiple equation systems 94–5 multiple occupational choices 94 non-binary occupational choices 96 cross-section regression models 87–90 CT (corporation tax) 423, 436–7 Cypress 64 DARA (decreasing absolute risk aversion) 37–8, 276 de Meza and Southey (1996) model 225 de Meza and Webb (1987) over-investment model 223–5, 226 deadweight cost 408 debt finance fragility of models 226–7 new venture survival 394 optimality of 208–9 overview 204–5, 226–8 terminology 209–12 see also convertible contracts Dell, Michael 376 dependent self-employed workers 13 dependent spinoffs 31 determinants of entrepreneurship 155–7 destructive entrepreneurship see unproductive/destructive entrepreneurship determinants of entrepreneurship business cycle 142–3 dependent starts 155–7 equations for 106–7 family background 134–8 health issues 133–4 human capital 44, 113–19 independence and job satisfaction 110–13 industry characteristics 138–9 knowledge spillovers 140–2 marital status 132–3, 187–8 pecuniary incentives 107–10 psychological traits 128–32 over-optimism and over-confidence 123–8 regional factors 147–51 risk attitudes 121–4 social capital 119–21 summary 108, 157 technology 139–40

Index

UK 109 unemployment 143–7 determinants of nascent entrepreneurship 151–5 difference-in-means approaches 86–7 discouraged borrower phenomenon 284 discrimination see ethnic discrimination displacement 408 distressed company phenomenon 283 Dixit, A. 40 double moral hazard 243, 422 see also Keuschnigg-Nielsen double moral hazard model draw 363, 368 due diligence 237 EAS (Enterprise Allowance Scheme) 438–9 econometric methods 86–7 economics, and the study of entrepreneurship 2–5 economics of entrepreneurship, definitions 4 education see education level; entrepreneurship education education level and raising finance 219–20 and startup propensities 154 and success 376–80 and survival 393 as determinant of entrepreneurship 117–19, 178–80 as determinant of venture growth 320 female entrepreneurs 117, 186 see also RTE EF see equity finance efficiency effect 302–3 empirical methods 86–7 employer discrimination against ethnic minorities 165–6 spillover into credit market 167 employment assistance schemes 438–40 employment protection legislation (EPL) 450–1 endogeneity 86 in cross-section regression models 88–90 English Common Law 420, 449 Enterprise Allowance Scheme (EAS) 438–9 enterprise culture 457–8 entrepreneurial ability in Lucas model 44 Lucas model extensions 45 entrepreneurial duration models 100–1 entrepreneurial dynasties 68 entrepreneurial effort 333–4 see also work hours entrepreneurial learning 356–7 entrepreneurial managerial ability, as variable over time 44 entrepreneurial spawning agency cost theories 65–7 definitions 61 determinants of entrepreneurship 155–7 incumbents’ firm size 62–3 incumbents’ organisational limitations 63–5 learning theories 67 see also independent spinoffs

543

entrepreneurs as ‘jacks of all trades’ 45 see also Lazear (2005) jack-of-all-trades (JAT) hypothesis definitions 1, 6, 7 founder characteristics and venture growth 319 inexorable decline in numbers predictions 42–3 role in aggregate wealth accumulation 264–5 unwillingness to cede ownership and control rights 244–5 see also habitual entrepreneurs; nascent entrepreneurs; necessity entrepreneurs; novice entrepreneurs; opportunity entrepreneurs entrepreneurship definitions 6 measures of 7–15 vs paid employment, theory of the firm considerations 52–7 see also theories of entrepreneurship entrepreneurship capital 327, 330 entrepreneurship education 119 see also business advice and assistance entrepreneurship policies 403–4 case against 407–8 case for 405–7 credit market interventions 417–18 policy evaluation 409–10 principles of policy design 408–9 promotion of equity finance 418–19 regulatory policies 419–21 tax policies 421–3 see also innovation policy; Loan Guarantee Schemes EPL (employment protection legislation) 450–1 equilibrium, economic concept of 3 equity finance (EF) as an optimal financial contract 242–4 factors inhibiting use 244–6 public policy promotions 418–19 regulatory policies 419–21 tax policies 421–3 see also business angels equity gap 246 equity rationing 246–7 escape mechanisms 277–8 ethnic cultures 173–4 ethnic differences 174–5 ethnic discrimination in product markets 170–1 in the capital market 166–70 in the labour market 165–6 positive factors 171–4 see also affirmative action schemes ethnic enclaves 172–3, 178 ethnic entrepreneurship in transition economies 22 overview 163–5 ethnic-minority businesses, family finance 250 ethnic role models 174

544

Index

ethnicity and intergenerational entrepreneurship 137–8 as determinant of nascent entrepreneurship 152 Evans and Jovanovic (1989) model 263–71 evolutionary theories of enterprise development 317 exits see venture closure, and failure experience, as determinant of entrepreneurship 115–17 extroversion 130 Fairchild Semiconductor 64, 67 family background as determinant of entrepreneurship 134–8 as determinant of nascent entrepreneurship 151 family-business background, entrepreneurs with 46, 135 family finance 169, 203, 234, 250–1 family labour, use in ethnic groups 173 family-owned firms and profit-non-maximization 57–8 retirement 355 role in the economy 328 see also entrepreneurial dynasties female entrepreneurship age as determinant 114 and venture growth 320 and venture survival 393 education level as determinant 117, 186 financing 195–6 gender earnings gap 190–4 growth and survival rates 194–5 hiring workers 295–6 in transition economies 22 marriage and household production 187–8 moonlighting 341 mothers’ self-employment experience 137 motivations 108 nascent entrepreneurship 9, 155 overview 184–6, 197–8 retirement 354 self-employment rates, UK 18 unemployment as determinant 145 work hours 341 see also affirmative action schemes FIML (Full Information Maximum Likelihood) 95 financing 203–5 female entrepreneurship 195–6 sources of 234–5, 261 see also co-operative finance schemes; debt finance; family finance; non-profit-making lenders; self-finance; trade credit; venture capital Finland age as determinant 115 habitual entrepreneurs 25 income 371 nascent entrepreneurship 9 new firm survival rates 389 first-mover advantage 315 first-order stochastic dominance (FOSD) 78 fixed-effects models 98 flexibility hypothesis 188–9

FOSD (first-order stochastic dominance) 78 France business set-up procedures 443 employment assistance schemes 438 family background of entrepreneurs 134 loan guarantee schemes 413 nascent entrepreneurship 9 new firm survival rates 389 self-employment rates 16, 17–18, 140 franchisees and classification as entrepreneurs 13–14 relative incomes 381 franchising, avoiding discrimination problems 169–71 Frank’s model of entrepreneurs’ labour supply and ageing 351–2, 356, 357–8 Fraser and Greene’s (2006) heteroscedastic probit model 102–3 Fried and Howitt’s (1980) implicit contract theory 283 Full Information Maximum Likelihood (FIML) 95 Gates, Bill 376 gazelles 316, 397 GEM (Global Entrepreneurship Monitor) 7–8 generated regressors 93 Germany employment assistance schemes 438–9 habitual entrepreneurs 25 incomes 371 loan guarantee schemes 413 Mutual Guarantee Schemes 258 nascent entrepreneurship 9 new firm survival rates 389 regional growth 327 self-employment rates 18 Ghatak et al. (2007) model 75 Gibrat’s Law of Proportionate Effect 42, 43, 310, 311, 330 Gifford, S. 32 glass ceiling 193–4, 197 Glass–Steagall Act 244 Global Entrepreneurship Monitor (GEM) 7–8 GLSs (group lending schemes) 258 government failure, risk of 408 Grameen Bank 252–3 grammar schools 118 grandstanding 241 group lending schemes (GLSs) 258 growth see macroeconomic growth; venture growth growth aspirations, as determinant of venture growth 319 habitual entrepreneurs 15, 24–6 see also portfolio entrepreneurs; serial entrepreneurs happiness 112 hazard models for duration of spells in entrepreneurship 100–1 see also moral hazard hazard ratios 101 health, as determinant of entrepreneurship 133–4 health insurance 134

Index

Heckman’s method for correcting for selection bias 88 heteroscedastic probit, for cross-section binary models of occupational choice 96 high-tech firms and entrepreneurial spawning 67 and firm size 58 hindsight bias 8 ‘hire or outsource’ model 79–80 see also ‘make or buy’ decision hobby businesses 7, 10 Holmes–Schmitz model 46 Holmstrom and Milgrom (1994) 52, 54, 80 homeworkers 13 homophily 119–20 household income, and new venture survival 394 human capital as determinant of entrepreneurship 44, 113–19 as explaining gender earnings gap 191 female entrepreneurs 186 see also Parker (2005) human capital based model identification issues 93 identifying instruments 89 immigrant entrepreneurship determinants of 178–80 mobility traps 375 participation 165 immigrants, propensities to entrepreneurship 176–7 immigration, effects on entrepreneurship 180–1 imperfect competition 3 imperfect information 3 incentive effects, and work hours 350 income effect, and entrepreneurial work hours 344 income splitting 432 income tax (IT) 423 empirical evidence 433–6 theoretical issues 429–33 incomes and profits 363 definitions 363–4 determinants findings 378–82 theory and methods 375–8 evidence of relative incomes 369–72 inequality and volatility 372–5 measurement problems 364–9 non-response to survey questions 368 under-reporting 365–8 see also household income, and new venture survival; pecuniary incentives incorporated businesses 11 incorporation 206 see also corporation tax independence, as incentive to entrepreneurship 110–13 independent spinoffs 31 see also entrepreneurial spawning India, dearth of research 23 Indonesia micro-finance schemes 252

545

self-employment rates 22 industry characteristics as determinant of entrepreneurship 138–9 female entrepreneurs 186 hiring workers 296 inflation 459 information costs, equity finance 245–6 information spillovers, regional 149 inheritance and business formation 274 as security for a loan 251 wealth transmission 265 inheritance tax (IHT) 437 inherited traits 131 hiring workers 296 see also intergenerational entrepreneurship inherited wealth model 68–9 innovation, modelling 2 innovation policy 423–5 innovation ratios 421 innovative rebelliousness 130 inside collateral 205 institutional entrepreneurs 454 Intel 63, 64 interest-free loans 251 interest rates, as determinant of entrepreneurship 143 intergenerational entrepreneurship 134–8 see also inherited traits intergenerational models, education choices 118 intergenerational transmission, as regional effect 149 internal locus of control 129 international trade, tariff reduction to reduce unproductive entrepreneurship 335 intertemporal substitution 350 intrapreneurs 62–3 see also dependent spinoffs intrapreneurship 31 intuition 130 Inverse Mills Ratio 88, 93 investment banks 227 IPR (intellectual property rights) and entrepreneurial spawning 65–6, 156 and public policy 406, 449–50 IQ (intelligence quotient) 131 IR35 legislation 12 IT (income tax) 423 empirical evidence 433–6 theoretical issues 429–33 Italy bankruptcy provisions 446 business set-up procedures 443 female entrepreneurship 185 Mutual Guarantee Schemes 258 nascent entrepreneurship 9 self-employment rates 14, 16, 18 IV (Instrumental Variables), in cross-section regression models 89–90 ‘jacks of all trades’, entrepreneurs as 45 see also Lazear (2005) jack-of-all-trades (JAT) hypothesis

546

Index

Japan nascent entrepreneurship 9 self-employment rates 14, 16, 18, 140 Jefferson (1997) model 294 job creation by entrepreneurs evidence 295–6 hiring others 293–4 job quality 300, 408 theoretical analysis 294–5 by new firms 299–300 by small firms 297–300 job insecurity 374 job satisfaction 111–13 smaller firms 158 ‘job-shopping’ theory see Miller’s Jovanovic’s (1982) model of industry selection 41, 312–14 judicial efficiency 340, 449 Kanbur, S. M. 34, 36, 39 Keuschnigg-Nielsen double moral hazard model 425–6 Kihlstrom and Laffont (1979) model 49–51, 121–4 Kirzner, I. 32 Klepper, Steven 64, 311 (1996) theory of venture growth 314–15 Knight, Frank H. 32–3, 47 knowledge-based theories of macroeconomic growth 72–4 knowledge spillovers 72–3 and public policy 406–7 as determinant of entrepreneurship 140–2 regional 149–50 Kodak 64 Labour Sponsored Venture Capital Corporations (LSVCCs) 420 Landier (2004) serial entrepreneurship model 74–5, 81–2 language proficiency, host country 174 latent entrepreneurs 113 latent entrepreneurship 26 Laussel and Le Breton model 47 Lazear (2005) jack-of-all-trades (JAT) hypothesis 71, 116, 151, 381 learning see education level; entrepreneurial learning; entrepreneurship education lender discrimination, against ethnic minorities 166–70 Lewis, W. Arthur 22 liabilities of newness and smallness 389–90, 395 liability of obsolescence 390 life-cycle models of entrepreneurial participation 350–1 limited liability 206–7, 218 limited liability constraint 269 Loan Guarantee Schemes (LGS) 282, 459 evaluation of 415–17 organisation 412–13 theoretical perspectives 413–15

logit model 91 loss offsetting 430 lottery participation 122 LSI Logic 64 Lucas (1978) model 41–3 criticisms 43–4, 77 variants and extensions 44–9 macroeconomic growth 324–30 theories 67–8 knowledge-based 72–4 technology-based 70–2 wealth-based 68–70 ‘make or buy’ decision 52–7 see also ‘hire or outsource’ model managerial ability 45 Manove and Padilla (1999) model 225 marginal tax rates (MTR) 434 marital status as determinant of entrepreneurship 132–3, 187–8 as determinant of nascent entrepreneurship 151 Marshall effect 396 mean-preserving spread (MPS) 38, 78 Mexico, self-employment rates 16, 18 MGSs (Mutual Guarantee Schemes) 258–9 micro-finance promise 255 micro-finance schemes 251–7 Miller’s (1984) ‘job-shopping’ theory 114 Miner Sentence Completion Scale 122 minority-owned banks 170 MNL (multinomial logit) model 94 mobility trap 375 mode of entry 46 monitoring, by venture capitalists 239–40 moonlighting 341–3 moral hazard and Newman’s (2007) model 50–1, 272 and public policy 421–2 see also double moral hazard MPS (mean-preserving spread) 38, 78 MTR (marginal tax rates) 434 multi-variate models 86 multinomial logit (MNL) model 94 multiplier effect 396 nascent entrepreneurs (NEs) 8–10 characteristics 151–2 moonlighting 341 venture development paths 152–5 National Semiconductor 64 NDC (Non-Distribution Constraint), competitive advantage from 59–60 necessity entrepreneurs 7, 111 necessity spinoffs 26 need for achievement, as determinant of entrepreneurship 129 Nepal, self-employment rates 22 NEs see nascent entrepreneurs neuroticism 130 new firm formation, and economic growth 325–30 new firms, job creation 299–300

Index

new venture creation, as measure of entrepreneurship 7–10 Newman’s (2007) moral hazard model 50–1, 272 niche purveyor strategy, and new venture survival 395 non-profit-making lenders 257 non-stationarity 86 non-stationary processes 97 Norway habitual entrepreneurs 24 nascent entrepreneurs 154 self-employment rates 16, 18 novice entrepreneurs 15, 24 NP (not-for-profit) sector 58 see also social enterprises occupational choice 36 hysteresis (path-dependence) 40 occupational choice models with heterogeneous ability 41–9 with heterogeneous risk attitudes 49–51 with homogeneous agents definitions of risk aversion and risk 37–8 dynamic models 39–41 static models 38–9 occupational following 137 OECD Labour Force Statistics 11 OLS (ordinary least squares) estimation 87 openness to experience 130 opportunity entrepreneurs 7–8 opportunity spinoffs 26 optimal level of entrepreneurship 329, 330, 460 optimisation, economic concept of 2–3 optimism see over-optimism outside collateral 205 over-investment 211–12, 223 see also de Meza and Webb (1987) over-investment model over-optimism 225 advantages of 127–8 as determinant of entrepreneurship 124–6 misconstrued as risk-taking 123 overconfidence advantages of 127 conflation with over-optimism 125 paid employment, definitions 31 panel-data models 98–100 Panel Study of Entrepreneurial Dynamics (PSED) 8 Pareto optimality 209 Parker (2005) human capital based model 76, 82–3, 150 Parker and Rougier’s model of the entrepreneurial retirement decision 353, 354, 359–60 Parker and van Praag (2006) model 89–90 passive learning 356 patent races 71, 74 Paulson et al. (2006) structural model 286 payroll taxes (PT) 429 see also income tax pecking-order hypothesis 245

547

pecuniary incentives vs non-pecuniary 107–9 see also relative incomes/earnings, as incentive to entrepreneurship Philippines, self-employment rates 22 Poland income 371, 374 self-employment rates 19, 20 political instability 458–9 pooled regression models 98 portfolio entrepreneurs 15, 24, 26 positive assortative matching 253 postmaterialism, and determinants of entrepreneurship 110 Prince’s Trust 442 principals 52 private equity premium puzzle 263, 266–8 probit model 91 procedural utility 111 procurement schemes 440 product quality strategy, and new venture survival 395 professionalisation, provided by venture capitalists 240 profits 363 see also incomes property rights and credit misallocation 340 and rent seeking 336–7, 338 promoting entrepreneurship 448–50 ‘prosperity-pull’ effects 143–4 PSED (Panel Study of Entrepreneurial Dynamics) 8 public insurance 430 public policy 403–4, 459–62 fostering an enterprise culture 457–8 role in promoting entrepreneurship 21 see also affirmative action schemes; business advice and assistance; employment assistance schemes; entrepreneurship policies; innovation policy; political instability; procurement schemes; welfare state random-effects estimators 98 random-effects models 98–9 rational expectations hypothesis 3 ‘recession-push’ effects 143, 145–6 redlining 211, 215–16 reduced form approach 91 reduced form probit 93 reference group neglect 125 regional employment, and new firm formation 325–6 regional factors, as determinants of entrepreneurship 147–51 regression analysis 86 regression fallacy 297 regulation 442, 454–5, 461 costs of compliance 453–4 credit markets 451–3 of entry 442–5 of exit 446–8 see also employment protection legislation regulatory capture 420

548

Index

Regulatory Flexibility Act 454, 461 relationship lending 207 relative incomes/earnings as incentive to entrepreneurship 109–10 inclusion in cross-section binary models of occupational choice 92–4 rent seeking 333 see also unproductive/destructive entrepreneurship replacement effect 303 retirement, and working hours 351–5 retirement benefits 457 ‘revealed preference’ principle 6, 26, 86 risk distinction from ‘uncertainty’ 33 technical definitions 77–8 risk attitudes as determinants of entrepreneurship 121–4 risk aversion 37 as explaining gender earnings gap 192–3 in static models 38–9 technical definitions 77–8 see also private equity premium puzzle risk capital 237 risk-loving preference 37 risk neutrality 37 Romer, P. M. 72 Roscas (Rotating Savings and Credit Associations) 169–70, 258 Rosen’s (1981) theory of ‘superstars’ 373 Rotter Scale 129 RTE (return to an extra year of education) 376–80, 382 Russia challenges to promoting entrepreneurship 21 entrepreneurial effort 339 income 371, 374 self-employment rates 19–22 sample attrition 99 sample selection bias 86 in cross-section regression models 87–8 Sarbanes–Oxley (SOX) Act 450 saving rates, of entrepreneurs 265 Say, Jean-Baptiste 33–4 SBIR (Small Business Innovation Research) programme 423–4 school voucher programmes 118 Schumpeter, Josef 6, 34–5, 107, 300 screening by banks 217–21 see also Bester’s (1985) screening model; due diligence screening by employers 376 Segal and Spivak’s (1989) model 315–16 self-efficacy 130 self-employment as measure of entrepreneurship 10–15 preference for 11–12, 19 rates in developed countries 15–19 rates in developing countries 22–4 rates in transition economies 19–22 self-finance 203, 224 and new venture success 155

and new venture survival 394 preference for 276 senescence 390 separation of ownership and control 219 serial entrepreneurs 15, 24 and venture capitalists 241 serial entrepreneurship 25–6 see also Landier (2004) serial entrepreneurship model shadow wages 345 shallow pockets 247 Shockley Labs 64 Sierra Semiconductor 64 signalling 127, 149, 206 Silicon Valley compared to Massachusetts 141 regional factors 149 self-employment rates 26 see also determinants of entrepreneurship, dependent starts single risk hazard models 101 size distribution fallacy 297 skewed returns, preference for 267–8, 285–6 Small Business Technology Transfer (STTR) programme 424 small firms and profit-non-maximization 57–8 innovation 300–1 evidence 303–5 theoretical arguments 301–3 job creation 297–300 job satisfaction 158 likelihood of employees to become entrepreneurs 62 Small Firms Loan Guarantee Scheme 410 SME policies 403–4 SMEs (small and medium sized firms) as measure of entrepreneurship 10 importance in the economy 18–19 see also small firms Smith, Adam 124 social capital as determinant of venture growth 319 as explaining gender earnings gap 192 social enterprises competitive advantage 58–61 definitions 58 social entrepreneurs cheerfulness 61 see also social enterprises social mobility 374–5 social security contributions 432 sojourner theory 176, 178–9 sole proprietorships 11 SOSD (second-order stochastic dominance) 38, 78 Spain, self-employment rates 18 spinoffs public policy 425 see also dependent spinoffs; independent spinoffs; necessity spinoffs; opportunity spinoffs

Index

spouses advantages of having 132 affect on retirement age 353 and new venture survival 394 education level 381 occupational choices of 95 stages of growth models 317 start-up capital, as explaining gender earnings gap 191 state-dependent choices 99 statistical discrimination 166–7 status quo bias 130 stickiness of commercial loan rates 282–3 Stiglitz and Weiss’ (1981) model 212–16, 229–30 counter-rebuttals 221–3 rebuttals 216–21 Storey, David 316 stress and anxiety 112 strong ties 119 ‘structural’ occupational choice models 91 STTR (Small Business Technology Transfer) programme 424 substitution effect, and entrepreneurial work hours 344 supernormal profits 3 superstars 373 survival 385 see also venture survival survival bias 8 Sweden, habitual entrepreneurs 24–5 Swedish model welfare state 456 switching between entrepreneurship and paid employment 40 Switzerland, self-employment rates 18 tacit knowledge 73 target income 350 tax compliance enforcement (TCE) 434–5 tax deductibility of not-for-profit organizations 60–1 tax systems, and measures of self-employment 12 taxation 459 and venture growth 324 entrepreneurship policies 421–3 evasion and income under-reporting 364–8, 432, 433, 434, 459–60 on interest income 224–5 promoting entrepreneurial saving 265 see also capital gains tax; corporation tax; income tax taxicab companies 52, 53 TC (trade credit) 259–61 TCMP (Taxpayer Compliance Measurement Program) 364–5 TEA (Total Entrepreneurial Activity) index 7 team starts 9 and venture growth 321 ‘theory of the firm’ considerations 56 technical elasticity of substitution 42 technological change and Lucas model 43–4 as determinant of entrepreneurship 139–40 impact of 71–2

549

technology-based theories of macroeconomic growth 70–2 technology strategy, and new venture survival 395 ten-percenters 316 testosterone, and risk tolerance 130 TFP (total factor productivity) 139–40 theories of entrepreneurship 31–2 early theories 32, 35–6 arbitrage and the bearing of risk and uncertainty 32–3 co-ordination of factors of production 33–4 creative destruction 34–5 leadership and motivation 35 personal or psychological traits 35 existence of an entrepreneurial option 51 non-profit-maximising ventures 57–61 theory of the firm considerations 52–7 macroeconomic growth theories 67–8 knowledge-based 72–4 technology-based 70–2 wealth-based 68–70 modern theories 36, 76–7 multiple equilibrium models 74–6 see also occupational choice models third age entrepreneurship 355 tiering 453 time-series models 96–8 tolerance of ambiguity 129 top coding 369 Total Entrepreneurial Activity (TEA) 7 trade credit (TC) 259–61 trade unions 457 transition economies, self-employment rates 19–22 ‘troublesome’ variables 88 trucking companies 52, 53, 54 turbulence 324 two-equation simultaneous equation model, reduced form 101–2 UK bankruptcy regulation 446, 448 business advice and assistance 441 business angels 248–50 Enterprise Allowance Scheme (EAS) 438–9 enterprise culture 457–8 ethnic discrimination in the capital market 166–7, 169 habitual entrepreneurs 24 income inequality 373–4 income under-reporting 365 Loan Guarantee Schemes 413, 415–17 relative incomes 371 self-employment rates 18, 147 venture growth 316 uncertainty, distinction from ‘risk’ 33 uncertainty avoidance 129 uncompensated wage elasticity 345 under-investment 211 equity finance 247–8 underbidding, by employers 66

550

Index

unemployment, as determinant of entrepreneurship 143–7 unobserved heterogeneity 86 unpaid family workers, and classification as entrepreneurs 14 unproductive/destructive entrepreneurship 333 conceptual issues 334–8 evidence 338–40 urban areas 150 USA age as determinant 114 bankruptcy regulation 446 business angels 248–50 credit market interventions 417–18 ethnic discrimination in the capital market 167–9 ethnic discrimination in the product market 170 habitual entrepreneurs 25 health insurance 187 job creation by new firms 299 Loan Guarantee Schemes 413, 415–17 new firms survival rates 389 public policy 404, 408 Regulatory Flexibility Act 454, 461 Sarbanes–Oxley (SOX) Act 450 self-employment rates 17–18 venture capital market 238–9 Venice, state ownership of galleys 430 venture capital advantages 239–42 definitions 235 drawbacks 244–8 market size 238–9 organisational structure 235–8 see also equity finance venture closure, and failure 386–8 venture cycle 235 venture growth and survival 385, 393 definitional and measurement issues 317–18 evidence 316–17 evidence about determinants 319–24 overview 310 theories 311–16 venture survival definitions 385–6

empirical determinants 391–6 overview 396–7 rates of 388–9 theoretical determinants 389–91 vesting 236 vulture capital 245 weak ties 119 wealth and venture investment decisions 280 and venture survival 279 as determinant of nascent entrepreneurship 152 relationship with entrepreneurship 269, 284–5 Aghion and Bolton (1997) model 69, 216, 272 alternative interpretations 275–9 Banerjee and Newman (1993) model 68–9, 271–2 empirical evidence 273–5 Evans and Jovanovic (1989) model 263–71 Newman’s (2007) moral hazard model 50–1, 272 wealth-based theories of macroeconomic growth 68–70 wealth diversification 268–9 wealth inequality, entrepreneurs’ role in 265 wealth taxes 437 welfare state size 455–6 social security transfers 457 windfalls 273–4, 276 window dressing activities 243 women see female entrepreneurship work–family conflict 112–13 work hours 112, 185, 341–3 ageing and retirement 351–5 explanations for 343–51 see also flexibility hypothesis worker co-operatives, and classification as entrepreneurs 14 worker monitoring 54 Xerox 63–4 Yunus, Muhammad 252 Zilog 64