The Capitalization of Knowledge: A Triple Helix of University-industry-government

The Capitalization of Knowledge The Capitalization of Knowledge A Triple Helix of University–Industry– Government Ed...

Author: Riccardo Viale | Fondazione Rosselli

152 downloads 1071 Views 1MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

The Capitalization of Knowledge

The Capitalization of Knowledge A Triple Helix of University–Industry– Government

Edited by

Riccardo Viale Fondazione Rosselli, Turin, Italy

Henry Etzkowitz Stanford University, H-STAR, the Human-Sciences and Technologies Advanced Research Institute, USA and the University of Edinburgh Business School, Centre for Entrepreneurship Research, UK

Edward Elgar Cheltenham, UK • Northampton, MA, USA

© The Editors and Contributions Severally 2010 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical or photocopying, recording, or otherwise without the prior permission of the publisher. Published by Edward Elgar Publishing Limited The Lypiatts 15 Lansdown Road Cheltenham Glos GL50 2JA UK Edward Elgar Publishing, Inc. William Pratt House 9 Dewey Court Northampton Massachusetts 01060 USA

A catalogue record for this book is available from the British Library Library of Congress Control Number: 2009941141

ISBN 978 1 84844 114 9

02

Printed and bound by MPG Books Group, UK

Contents List of contributors Acknowledgements Abbreviations

vii ix x

Introduction: anti-cyclic triple helix Riccardo Viale and Henry Etzkowitz PART I 1 2

3 4

5

6

1

HOW TO CAPITALIZE KNOWLEDGE

Knowledge-driven capitalization of knowledge Riccardo Viale ‘Only connect’: academic–business research collaborations and the formation of ecologies of innovation Paul A. David and J. Stanley Metcalfe Venture capitalism as a mechanism for knowledge governance Cristiano Antonelli and Morris Teubal How much should society fuel the greed of innovators? On the relations between appropriability, opportunities and rates of innovation Giovanni Dosi, Luigi Marengo and Corrado Pasquali Global bioregions: knowledge domains, capabilities and innovation system networks Philip Cooke Proprietary versus public domain licensing of software and research products Alfonso Gambardella and Bronwyn H. Hall

31

74 98

121

143

167

PART II TRIPLE HELIX IN THE KNOWLEDGE ECONOMY 7

8

A company of their own: entrepreneurial scientists and the capitalization of knowledge Henry Etzkowitz Multi-level perspectives: a comparative analysis of national R&D policies Caroline Lanciano-Morandat and Eric Verdier v

201

218

vi

9

10

11

12

The capitalization of knowledge

The role of boundary organizations in maintaining separation in the triple helix Sally Davenport and Shirley Leitch The knowledge economy: Fritz Machlup’s construction of a synthetic concept Benoît Godin Measuring the knowledge base of an economy in terms of triple-helix relations Loet Leydesdorff, Wilfred Dolfsma and Gerben Van der Panne Knowledge networks: integration mechanisms and performance assessment Matilde Luna and José Luis Velasco

Index

243

261

291

312

335

Contributors Cristiano Antonelli, Dipartimento di Economia S. Cognetti de Martlis Università di Torino, Italy and BRICK (Bureau of Research in Innovation, Complexity and Knowledge) Collegio Carlo Alberto, Italy. Philip Cooke, Centre for Advanced Studies, Cardiff University, Wales, UK. Sally Davenport, Victoria Management School, Victoria University of Wellington, New Zealand. Paul A. David, Department of Economics, Stanford University, CA, USA. Wilfred Dolfsma, School of Economics and Business, University of Groningen, The Netherlands. Giovanni Dosi, Laboratory of Economics and Management, Sant’Anna School of Advanced Studies, Pisa, Italy. Henry Etzkowitz, Stanford University, H-STAR, the Human-Sciences and Technologies Advanced Research Institute, USA and the University of Edinburgh Business School, Centre for Entrepreneurship Research, UK. Alfonso Gambardella, Department of Management, Università Luigi Bocconi, Milano, Italy. Benoît Godin, Institut National de la Recherche Scientifique, Montréal, Québec, Canada. Bronwyn H. Hall, Department of Economics, University of California at Berkeley, Berkeley, CA, USA. Caroline Lanciano-Morandat, Laboratorie d’économie et de sociologie du travail (LEST), CNRS, Université de la Méditerranée et Université de Provence, Aix en Provence, France. Shirley Leitch, Swinburne University of Technology, Melbourne, Australia. vii

viii


Loet Leydesdorff, Amsterdam School of Communications Research, University of Amsterdam, The Netherlands. Matilde Luna, Instituto de Investigaciones Sociales, Universidad Nacional Autónoma de México. Luigi Marengo, Sant’Anna School of Advanced Studies, Pisa, Italy. J. Stanley Metcalfe, ESRC-CRIC, University of Manchester, UK. Corrado Pasquali, Università degli Studi di Teramo, Italy. Morris Teubal, Department of Economics, The Hebrew University, Jerusalem, Israel. Gerben Van der Panne, Economics of Innovation, Delft University of Technology, Delft, The Netherlands. José Luis Velasco, Instituto de Investigaciones Sociales, Universidad Nacional Autónoma de México. Eric Verdier, Laboratorie d’économie et de sociologie du travail (LEST), CNRS, Université de la Méditerranée et Université de Provence, Aix en Provence, France. Riccardo Viale, Fondazione Rosselli, Turin, Italy.

Acknowledgements The publishers wish to thank the following, who have kindly given permission for the use of copyright material. Elsevier for articles by Alfonso Gambardella and Bronwyn H. Hall (2006), ‘Proprietary versus public domain licensing of software and research products’, in Research Policy, 35 (6), 875–92, Loet Leydesdorff, Wilfred Dolfsma and Gerben Van der Panne (2006), ‘Measuring the knowledge base of an economy in terms of triple-helix relations among “technology, organization, and territory”’ in Research Policy, 35 (2), 181–99 and G. Dosi, L. Marengo and C. Pasquali (2006) ‘How much should society fuel the greed of innovators? On the relations between appropriability, opportunities and rates of innovation’, in Research Policy, 35 (8), 1110–21. Every effort has been made to trace all the copyright holders, but if any have been inadvertently overlooked the publishers will be pleased to make the necessary arrangements at the first opportunity. Some of the chapters in this book are completely new versions of the papers presented at the 5th Triple Helix Conference held in Turin in 2005 and organized by Fondazione Rosselli. Other chapters have been written especially for this book or are a new version of already published papers.

ix

Abbreviations ACRI AIM CADs CAFC CEO CNRS CRI DARPA DBF EPO ERISA FDI GDP GM GMO GNF GPL HEI HMS HR ICND ICT INRIA IP IPO IPR JCSG KIS LGPL LSN MIT NACE

Association of Crown Research Institutes (New Zealand) Alternative Investment Market (UK) Complex Adaptive Systems Court of Appeals for the Federal Court (USA) chief executive officer National Centre for Scientific Research (France) Crown Research Institute (New Zealand) Defense Research Projects Agency (USA) dedicated biotechnological firms European Patent Office Employment Retirement Income Security Act (USA) foreign direct investment gross domestic product genetic modification genetically modified organism Genomics Institute of the Novartis Research Foundation (USA) Generalized Public License (USA) higher education institution Harvard Medical School human resources Institute of Childhood and Neglected Diseases (USA) information and communication technology National Institute of Computer Science (France) intellectual property initial public offering intellectual property rights Joint Center for Structural Genomics (USA) knowledge-intensive services lesser Generalized Public License (USA) Life Sciences Network (New Zealand) Massachusetts Institute of Technology Nomenclature générale des Activités économiques dans les Communautés

x

Abbreviations

NASDAQ NIBR NIH NSF NUTS NYU OECD OTC PD PLACE PoP PR PRO R&D RCGM RDI RIO RoP RSNZ SHIPS

SMEs TIP TTO UCSD UCSF VCE WYSIWYG

National Association of Securities Dealers Automated Quotations Novartis Institutes for Biomedical Research (USA) National Institutes of Health (USA) National Science Foundation (USA) Nomenclature des Unités Territoriales Statistiques New York University Organisation for Economic Co-operation and Development over the counter public domain proprietary, local, authoritarian, commissioned, expert professor of practice proprietary research public research organization research and development Royal Commission into Genetic Modification (New Zealand) research, development and innovation regional innovation organizer researcher of practice Royal Society of New Zealand Strategic, founded on Hybrid and interdisciplinary communities, able to stimulate Innovative critique and should be Public and based on Scepticism small and medium-sized enterprises Technology Investment Program (USA) technology transfer officer University of California San Diego University of California San Francisco virtual centre of excellence what you see is what you get (IT)

xi

Introduction: anti-cyclic triple helix Riccardo Viale and Henry Etzkowitz THE TRIPLE HELIX IN ECONOMIC CYCLES The year 2009 may have represented a turning point for research and innovation policy in Western countries, with apparently contradictory effects. Many traditional sources of financing have dried up, although some new ones have emerged, for example as a result of the US stimulus package. Manufacturing companies are cutting their R&D budgets because of the drop in demand. Universities saw their endowments fall by 25 per cent or more because of the collapse in financial markets. Harvard interrupted the construction of its new science campus, while Newcastle University speeded up its building projects in response to the economic crisis. Risk capital is becoming increasingly prudent because of the increased risk of capital loss (according to the International Monetary Fund, the ratio between bank regulatory capital and riskweighted assets increased on average between 0.1 and 0.4 for the main OECD countries during 2009) while sovereign funds, like Norway’s, took advantage of the downturn to increase their investments. According to the National Venture Capital Association, American venture capital shrank from US$7.1 billion in the first quarter of last year to US$4.3 billion in the first quarter of 2009 (New York Times, 13 April 2009). Many of the pension funds, endowments and foundations that invested in venture capital firms have signalled that they are cutting back on the assets class. The slowdown is attributable in part to venture capitalists and their investors taking a wait-and-see approach until the economy improves. The future outlook for R&D looks poor unless a ‘white knight’ comes to its rescue. This help may come from an actor whose role was downplayed in recent years, but that now, particularly in the USA, seems to be in the ascendant again. It is the national and regional government that will have to play the role of the white knight to save the R&D system in Western economies (Etzkowitz and Ranga, 2009). In the previous 20 years the proportion of public financing had gradually fallen in

1

2


percentage terms, while the private sector had become largely dominant (the percentage of Gross Domestic Expenditure in R&D financed by industry now exceeds 64 per cent in OECD countries). In some technological sectors, such as biotechnology, the interaction between academy and industry has become increasingly autonomous from public intervention. University and corporate labs established their own agreements, created their own joint projects and laboratories, exchanged human resources and promoted the birth of spin-off and spin-in companies without relevant help from local and national bodies. Cambridge University biotech initiatives or University of California at San Diego relations with biotech companies are just some of many examples of double-helix models of innovation. In other countries and in other technological sectors the double-helix model didn’t work and needed the support of the public helix. Some European countries, like France, Germany and Italy, saw a positive intervention of public institutions. In France, Sophia Antipolis was set up with national and regional public support. In Italy, support from Piedmont regional government to the Politecnico of Turin allowed the development of an incubator of spin-off companies that incubated more than 100 companies. In sectors such as green technologies, aerospace, security and energy, public intervention to support the academy–industry relationship is unavoidable. Silicon Valley venture capitalists invested heavily in renewable energy technology in the upturn, and then looked to government to provide funding to their firms and rescue their investments once the downturn took hold. In emerging and Third World economies, the role of the public helix in supporting innovation is also unavoidable. In the least developed countries industry is weak, universities are primarily teaching institutions and government is heavily dependent upon international donors to carry out projects. In newly developed countries the universities are developing research and entrepreneurship activities and industry is taking steps to promote research, often in collaboration with the universities, while government plays a creative role in developing a venture capital industry and in offering incentives to industry to support research through tax breaks and grants. The novelty of the current crisis is that the public helix becomes crucial even in countries and in sectors where the visible public role was minimal in the past. The Advanced Technology Program, the US answer to the European Framework Programmes, shrunk to virtual inactivity with zero appropriations under the Bush Administration but has found a second life under the Obama Administration and has been renamed the TIP (the Technology Investment Program). The triple-helix model seems to play an anti-cyclic role in innovation.

Introduction: anti-cyclic triple helix

3

It is a default model that guarantees optimal or quasi-optimal levels of academy–industry interaction through public intervention. It expresses its potential when the interaction is not autonomous, as is now the case in times of crisis, and the collaboration between universities and companies calls for financial support and organizational management. It works as a ‘nudge tool’ (Thaler and Sunstein, 2008), whose aim is to maintain a sufficient flow of innovation through the right incentives and institutional mechanisms for academy–industry collaboration. In this book we will examine various models for the capitalization of knowledge and attempt to discern the features of the new relationship that is emerging between the state, universities and industry. Are they converging in a certain way across different sociopolitical cultures and political institutions? Which of the key groups (scientists, politicians, civil servants, agency officials, industrialists, lobby groups, social movements and organized publics) are emerging as relevant players in the science and technology (S&T) policy arenas? What and how divergent are the strategies that they are pursuing and at what levels in the policy-making process do they take part? What are the ‘appropriate’ policies that respond to these changes? Do they call for a radical paradigm-like shift from previously established research policy? What degrees of freedom and autonomy can universities gain within the new triangular dynamics? Are the new patterns of interaction among those sectors designing a new mode of knowledge production? How are such changes altering the structure and operations of the knowledge-producing organizations inside these sectors?

POLYVALENT KNOWLEDGE: THREATS AND BENEFITS TO ACADEMIC LIFE The triple helix is a model for capitalizing knowledge in order to pursue innovation (Etzkowitz, 2008). Academic communities are fearful that capitalization will diminish the university goal of knowledge production per se. This fear seems to be linked to a traditional image of the division of labour in universities. Curiosity-driven research is separated from technology-driven research. Therefore, if a university focuses on the latter, it handicaps and weakens the former. On the contrary, in our opinion, in many technological fields knowledge production simultaneously encompasses various aspects of research. The theory of polyvalent knowledge (Etzkowitz and Viale, 2009) implies that, contrary to the

4


division of knowledge into divergent spheres – applied, fundamental, technological – or into mode 1 (disciplinary knowledge) and mode 2 (applied knowledge) (Stokes, 1997; Gibbons et al., 1994), a unified approach to knowledge is gradually becoming established. In frontier areas such as nanotechnologies and life sciences, in particular, practical knowledge is often generated in the context of theorizing and fundamental research. And, on the other hand, new scientific questions, ideas and insights often come from the industrial development of a patent and the interaction of basic researchers and industrial labs. The polyvalence of knowledge encourages the multiple roles of academics and their involvement in technology firms, and vice versa for industrial researchers in academic labs. One way of testing the reliability of this theory is to verify whether or not there is any complementarity between scientific and technological activities, measured by the number of publications and patents respectively. In the case of polyvalent knowledge, the same type of knowledge is able to generate both scientific output and technological output. Since the scientific knowledge contained in a publication generates technological applications represented by patents, and technological exploitation generates scientific questions and answers, we should expect to see some complementarity between publishing and patenting. Researchers who take out patents should show greater scientific output and a great capacity to affect the scientific community, measured by the impact factor or citation index. In other words, increasing integration between basic science and technology implies that there is no rivalry between scientific and technological output. The rivalry hypothesis holds that there is a crowding-out effect between publication activities and patenting. The substitution phenomenon between publications and patents stems from the inclusion of marketrelated incentives into the reward structure of scientists (Dasgupta and David, 1985; Stephan and Levin, 1996). Scientists increasingly choose to allocate their time to consulting activities and research agreements with industrial partners. They spend time locating licensees for their patents or working with the licensee to transfer the technology. Time spent doing research may be compromised. These market goals substitute peer-review judgement and favour short-term research trajectories and lower-quality research (David, 1998). Moreover, the lure of economic rewards encourages scientists to seek IP (intellectual property) protection for their research results. They may postpone or neglect publication and therefore public disclosure. Industry funding, commercial goals and contract requirements may lead researchers to increase secrecy with regard to research methodology and results (Blumenthal et al., 1986; Campbell


5

et al., 2002). Both these mechanisms may reduce the quantity and the quality of scientific production. This behaviour supports the thesis of a trade-off between scientific research and industrial applications. On the contrary, a non-rivalry hypothesis between publishing and patenting is based on complementarity between the two activities. The decision of whether or not to patent is made at the end of research and not before the selection of scientific problems (Agrawal and Henderson, 2002). Moreover, relations with the licensee and the difficulties arising from the development of patent innovation can generate new ideas and suggestions that point to new research questions (Mansfield, 1995). In a study, 65 per cent of researchers reported that interaction with industry had positive effects on their research. A scientist said: ‘There is no doubt that working with industry scientists has made me a better researcher. They help me to refine my experiments and sometimes have a different perspective on a problem that sparks my own ideas’ (Siegel et al., 1999). On the other hand, the opposition between basic and technological research seems to have been overcome in many fields. In particular, in the area of key technologies such as nanotechnology, biotechnology, ICT (information and communication technologies), new materials and cognitive technologies, there is continuous interaction between curiositydriven activities and control of the technological consequences of the research results. This is also borne out by the epistemological debate. The Baconian ideal of a science that has its raison d’être in practical application is becoming popular once again after years of oblivion. And the technological application of a scientific hypothesis, for example regarding a causal link between two classes of phenomena, represents an empirical verification. An attempt at technological application can reveal anomalies and incongruities that make it possible to define initial conditions and supplementary hypotheses more clearly. In short, the technological ‘check’ of a hypothesis acts as a ‘positive heuristic’ (Lakatos, 1970) to develop a ‘positive research programme’ and extend the empirical field of the hypothesis. These epistemological reasons are sustained by other social and economic reasons. In many universities, scientists wish to increase the visibility and weight of their scientific work by patenting. Collaboration with business and licensing revenues can bring additional funds for new researchers and new equipment, as well as meeting general research expenses. This in turn makes it possible to carry out new experiments and to produce new publications. In fact Jensen and Thursby (2003) suggest that a changing reward structure may not alter the research agenda of faculty specializing in basic research. Indeed, the theory of polyvalent knowledge suggests that dual goals may enhance the basic research agenda.

6


COMPLEMENTARITY BETWEEN PUBLISHING AND PATENTING The presence of a complementary effect or the substitution of publishing and patenting has been studied empirically in recent years. Agrawal and Henderson (2002) have explored whether at the Departments of Mechanical and Electrical Engineering of MIT patenting acts as a substitute or a complement to the process of fundamental research. Their results suggest that while patent counts are not a good predictor of publication counts, they are a reasonable predictor of the ‘importance’ of a professor’s publications as measured by citations. Professors who patent more write papers that are more highly cited, and thus patenting volume may be correlated with research impact. These results offer some evidence that, at least at the two departments of MIT, patenting is not substituting for more fundamental research, and it might even be an accelerating activity. Stephan et al. (2007) used the Survey of Doctorate Recipients to examine the question of who is patenting in US universities. They found patents to be positively and significantly correlated to the number of publications. When they broke the analysis down into specific fields, they found that the patent–publishing results persisted in the life sciences and in the physical/engineering sciences. The complementarity between publishing and patenting in life sciences has been studied by Azoulay et al. (2005). They examined the individual, contextual and institutional determinants of academic patenting in a panel data set of 3884 academic life scientists. Patenting is often accompanied by a flurry of publication activity in the year preceding the patent application. A flurry of scientific output occurs when a scientist unearths a productive domain of research. If patenting is a by-product of a surge of productivity, it is reasonable to conclude that a patent is often an opportunistic response to the discovery of a promising area. In the past, senior scientists and scientists with the most stellar academic credentials were usually also the most likely to be involved in commercial endeavours. But a feature of the Second Academic Revolution and the birth and diffusion of entrepreneurial universities is that the academic system is evolving in a way that accommodates deviations from traditional scientific norms of openness and communalism (Etzkowitz, 2000). In fact, Azoulay et al.’s (2005) data indicate that many patenting events now take place in the early years of scientists’ careers and the slope of the patent experience curve has become steeper with more recent cohorts of scientists. Patents are becoming legitimate forms of research output in promotion decisions. Azoulay et al. (2005) show that patents and papers encode similar pieces of knowledge and correspond to two types of output

Introduction: anti-cyclic triple helix Non-patenters

7

Patenters

70 60

% of scientists

50 40 30 20 10 0 0

100 200 300 400 500 600 0

100 200 300 400 500 600

Total number of publications

Figure I.1

Distribution of publication count for patenting and non-patenting scientists

that have more in common than previously believed. Figure I.1 shows the complementary of patenting and publishing in Azoulay et al. (2005). It plots the histogram for the distribution of publication counts for our 3884 scientists over the complete sample period, separately for patenting and non-patenting scientists. The study that makes the most extensive analysis of the complementarity between patenting and publishing is by Fabrizio and DiMinin (2008). It uses a broad sample drawn from the population of university inventors across all fields and universities in the USA, with a data set covering 21 years. Table I.1 provides the annual and total summary statistics for the entire sample and by inventor status. A difference of mean test for the number of publications per year for inventors and non-inventors suggests that those researchers holding a patent applied for between 1975 and 1995 generate significantly more publications per year than non-inventors. The inventors in their sample are more prolific in terms of annual publications, on the order of 20–50 per cent more publications than their non-inventor colleagues. The results suggest also that there is not a significant positive relationship between patenting and citations and a faculty member’s publications. Nor was evidence of a negative trade-off between publishing and patenting found in Europe. Van Looy et al. (2004) compared the publishing

8

Table I.1


Patenting and publishing summary statistics for inventors and non-inventors Inventors

Annual pubs Annual pats Total pubs Total pats

Non-inventors

All

Mean

St. dev.

Mean

St. dev.

Mean

St. dev.

3.99 0.56 79.93 11.02

5.18 1.55 84.78 16.21

2.24 0 43.71 0

2.96 0 47.72 0

3.12 0.28 62.00 5.57

4.32 1.14 71.30 12.77

output of a sample of researchers in the contract research unit at the Catholic University of Leuven in Belgium with a control sample from the same university. The researchers involved in contract research published more than their colleagues in the control sample. Univalent single-sourced formats are less productive than the polyvalent research groups at the Catholic University of Louvain that ‘have developed a record of applied publications without affecting their basic research publications and, rather than differentiating between applied and basic research publications, it is the combination of basic and applied publications of a specific academic group that consolidates the groups R&D potential’ (Ranga et al., 2003, pp. 301–20). This highly integrated format of knowledge production evolved from two divergent sources: industrial knowledge gained from production experience and scientific knowledge derived from theory and experimentation. In Italy an empirical analysis of the consequences of academic patenting on scientific publishing has been made by Calderini and Franzoni (2004), in a panel of 1323 researchers working in the fields of engineering chemistry and nanotechnologies for new materials over 30 years. As shown in Table I.2, the impact of patents is positive in the quantity of publications. Development activities are likely to generate additional results that are suitable for subsequent publications, although there might be one or two years of lag. Moreover, quality of research measured by the impact factor is likely to increase with the number of patents filed in the period following the publication. Scientific performance increases in the proximity of a patent event. This phenomenon can be explained in two ways. Top-quality scientific output generates knowledge that can be exploited technologically. And technological exploitation is likely to generate questions and problems that produce further insights and, consequently, additional publications. The same kind of results are found by Breschi et al. (2007), in a study done on a sample of 592 Italian academic inventors (see Table I.3).


Table I.2

Results of T-test and of test of proportions (two samples)

T-test

Mean patentholders

Mean nonpatentholders

Observations Total publications Impact factor Total cites received at 2003

133

1190

32.41 1.70 8.21

19.30 1.48 5.17

Test of proportions

Mean patentholders

Mean nonpatentholders

Observations Academic personnel Technician Area

133

1190

Note:

9

0.98 0.02 0.68

0.96 0.38 0.70

P-value Variance, Variance, Stat T Ho: 0 fidd (two tails) patentnonholders patent- in mean holders 133 1026.59 1.88 39.97

1190 897.86 2.82 33.30

4.50 0.01 5.30

0.00* 0.00* 0.00*

Stat Z P-value Ho: 0 fidd (two tails) in mean

0.89 1.35 0.42

0.37 0.187 0.68

* r < 0.05.

TRIPLE HELIX: LABORATORY OF INNOVATION The incorporation of economic development into university missions and the further integration of the knowledge infrastructure into innovation systems take different forms in various countries and regions. Most regions, however, lack innovation systems; rather they are innovation environments in which some elements to encourage innovation are present and others missing. In such situations it is important for some group or organization to play the role of regional innovation organizer (RIO) and bring the various elements of the triple helix together to foster new projects. Momentum starts to grow around concepts such as Silicon Alley in New York and Oresund in Copenhagen/southern Sweden, uniting politicians, business persons and academics. Imagery is also important since there often are not strong market reasons to allocate resources to the development of a region.

10


Table I.3

Publications per year, inventors versus controls, 1975–2003; by field N

Mean

Std

Median

Inventors Chem. eng. & materials tech.** Pharmacology* Biology* Electronic & telecom* All fields

63 83 78 72 296

2.0 2.2 2.5 1.7 2.1

1.75 1.21 2.10 1.04 1.60

1.5 2.0 2.0 1.4 1.8

Controls Chem. eng. & materials tech. Pharmacology Biology Electronics & Telecom All fields

63 83 78 72 296

1.3 1.7 1.8 1.3 1.6

1.10 1.11 1.27 1.18 1.28

1.1 1.6 1.5 1.0 1.3

Note: * – ** Inventor–control distribution difference significant at 0.90 - 0.95 (Kolmogorov– Smirnov test) Source:

Elaborations on EP–INV database and ISI Science Citation Index.

Rather than importing innovation mechanisms that appear to have worked well elsewhere, it is important as an initial step to analyse a local situation in order to determine: (1) (2)

the available resources that can be used to start the incubation process for knowledge-based development; what is missing and how and where those missing resources can be found, either locally or internationally.

To arrive at such a determination and follow-on collaboration means that there must be discussions among the potential actors rather than government saying by itself: this is what should be done. A consensus space, a forum that brings together the different triplehelix actors in a region, is often the source of new ideas and plans for knowledge-based development. From the analysis of the resources in a region, an awareness can be generated of the potential of its knowledge space, the research units, formal and informal, in the science and arts that, in turn, can become the basis for the creation of an innovation space, a mechanism to translate ideas into reality. The invention of the venture capital firm in 1940s New England is one example. These ‘triple-helix


11

spaces’ may be created in any order, with any one of them used as the basis for the development of others (Etzkowitz and Ranga, 2010). Creating new technology-based economic niches has become a third strategy for regional and local development. As the number of niches for science-based technology increases, the opportunity for more players to get involved also increases. Universities not traditionally involved in research are becoming more research-oriented, often with funding from their state and local governments, which increasingly realize that research is important to local economic growth. A firm may start from a business concept looking for a technology to implant within it or a technology seeking a business concept to realize its commercial potential. The entrepreneur propelling the technology may be an amateur or an experienced professional. Whichever the case, the technology comes with a champion who is attempting to realize its commercial potential by forming a firm. Universities, as well as firms, are developing strategic alliances and joint ventures. Karolinska University has recruited schools in the health and helping professions across Sweden into collaborations in order to increase its ‘critical mass’ in research. Groups of universities in Oresund, Uppsala and Stockholm have formed ‘virtual universities’, which are then translated into architectural plans for centres and science parks to link the schools physically. As entrepreneurial academic activities intensify, they may ignite a selfgenerating process of firm-formation, no longer directly tied to a particular university. The growth of industrial conurbations around universities, supported by government research funding, has become the hallmark of a regional innovation system, exemplified by Silicon Valley; the profile of knowledge-based economic development was further raised by the founding of Genentech and other biotechnology companies based on academic research in the 1980s. Once take-off occurs in the USA, only the private sector is usually credited; the role of government, for example, the Defense Research Projects Agency (DARPA), in founding SUN, Silicon Graphics and Cisco is forgotten. The triple helix denotes not only the relationship of university, industry and government, but also the internal transformation within each of these spheres. The transformation of the university from a teaching institution into one that combines teaching with research is still ongoing, not only in the USA, but in many other countries. There is a tension between the two activities, but nevertheless they coexist in a more or less compatible relationship. Although some academic systems still operate on the principle of separating teaching and research, it has generally been found to be both more productive and more cost-effective to combine the two functions, for example by linking research to the PhD training process. Will the same

12


relationship hold for three functions, with the emerging third mission of economic and social development combined with teaching and research? A recent experiment at Newcastle University points the way towards integration of the three academic functions. A project for the redevelopment of the region’s economy as a Science City was largely predicated on building new laboratories for academic units and for firms in the expectation that the opportunity to ‘rub shoulders’ with academics in related fields would be a sufficient attractor. However, a previous smaller-scale project, the Centre for Life, based on the same premise, did not attract a significant number of firms and the allotted space was turned over to academic units. To jump-start Science City, the professor of practice model, based on bringing distinguished practitioners into the university as teachers, has been ‘turned on its head’ to attract researchers of a special kind: PhD scientific entrepreneurs who have started successful firms but may have been pushed aside as the firm grew and hired professional managers. Newcastle University, in collaboration with the Regional Development Agency in Northeast UK, established four professors of practice (PoPs), one in each of the Science City themed areas – a scheme for knowledgebased economic development from advanced research. The PoPs link enterprise to university and are intentionally half-time in each venue so that they retain their industrial involvement at a high level and do not become traditional academics. The PoPs have initiated various projects, ranging from an interdisciplinary centre drawing together the university’s drug discovery expertise, which aims to undertake larger projects and attract higher levels of funding, to a new PhD programme integrating business, engineering and medical disciplines to train future academic and industrial leaders in the medical devices field. The next step in developing the PoP model is to extend it down the academic ladder by creating researchers of practice (RoPs), postdoctoral fellows and lecturers, who will work half-time in an academic unit and half-time in the business development side of the university, e.g. technology transfer office, incubator facility or science park. The RoPs would be expected to involve their students in analysing feasibility of technology transfer projects and in developing business plans with firms in the university’s incubator facility. Each PoP could mentor three or four RoPs, extending the reach of the senior PoPs as they train their junior colleagues. Moreover, the PoP model is relevant to all academic fields with practitioner constituencies, including the arts, humanities and social sciences. Until this happens, entrepreneurial activities will typically be viewed as an adjunct to, rather than an equal partner with, the now traditional missions of teaching and research. In the medium term, the PoP model may be expected to become a


13

forward linear model, as professors spinning out firms reduce to a half-time academic workload, superseding the typical current UK practice of forced choice. As some professors reduce to half-time, additional RoPs may be hired to share their positions. When 25 per cent or one in four academics are PoPs or RoPs, the entrepreneurial academic model will be institutionalized. The RoPs are intended to link academic units with the new business development units of the university. Incubators and tech transfer offices are typically established as administrative arms rather than as extensions of academic units, their natural location once the ‘third mission’ for economic and social development is fully accepted. In the interim, there is a need to bridge internal ‘silos’ as well as external spheres. The university is a flexible and capacious organization. Like the church, its medieval counterpart, it is capable of reconciling apparent contradictions while pursuing multiple goals in tandem. As the university takes up a new role in promoting innovation, its educational and research missions are also transformed. As the university expands its role in the economy, from a provider of human resources to a generator of economic activity, its relationship to industry and government is enhanced. Paradoxically, as the university becomes more influential in society, it is also more subject to influence, with academic autonomy increased in some instances and reduced in others. When bottom-up initiatives that have proved successful, such as the incubator movement in Brazil, are reinforced by top-down policies and programmes, perhaps the most dynamic and fruitful result is achieved. It also means that universities and other knowledge-producing institutions can play a new role in society, not only in training students and conducting research but also in making efforts to see that knowledge is put to use. There is also a convergence between top-down and bottom-up initiatives, which ideally reinforce one another. The flow of influence can go in both directions. If top-down, the local or regional level may adapt the policy or programme to local or regional needs. Bottom-up may also be the source of public pressure for action and the creation of models that can later be generalized top-down or through isomorphic mimesis. The US agricultural innovation system is a classic example of the ‘hybridization’ of top-down and bottom-up approaches, with government at various levels acting as a ‘public entrepreneur’ (Etzkowitz et al. 2001). The result is an interactive model, with intermediate mechanisms that integrate the two traditional starting points of science and technology policy. In contrast to biological evolution, which arises from mutations and natural selection, social evolution occurs through ‘institution formation’ and conscious intervention. The triple helix provides a flexible framework to guide our efforts, working from different starting points to achieve

14


the common goal of knowledge-based economic and social development. Innovation thus becomes an endless transition, a self-organizing process of initiatives among the institutional spheres.

CAPITALIZING KNOWLEDGE AND THE TRIPLE HELIX This book is structured into two parts. Part I deals with the ways, proprietary and not, to obtain an economic return from scientific and technological research. One of its focuses is how the epistemological and cognitive features of knowledge, its generation and utilization, can constrain and shape the way in which it is capitalized. Economic and social factors are necessary to explain the capitalization of knowledge. However, they are not sufficient. Their flaw is to neglect the constraints on the capitalization process deriving from the epistemological structure and cognitive dimension of knowledge. According to Chapter 1 by Riccardo Viale, these are crucial factors in the organizational and institutional changes taking place in the capitalization of knowledge. For example let’s consider the different ontologies and languages present in physics compared to material science or biology. The different use of quantitative measures versus qualitative and pictorial representations, the different types of laws and the role of experiments constrain the organizational ways knowledge can be capitalized. Any attempt to devise the right format for capitalization should consider these aspects. ‘Nudging’ (Thaler and Sunstein, 2008) the capitalization of knowledge means impressing institutions and organizations on the minds of the actors involved in the process of knowledge generation and application. Higher cognitive complexity, as in the case of converging technologies, cannot be coped with by isolated individual minds but calls for increasing division of computational labour. Greater epistemological generality, as in the case of the application of inclusive theories of physics and chemistry, allows better knowledge transfer but requires the involvement of many disciplines to widen the innovation field. Different background knowledge between university and companies hinders the reciprocal understanding that can be improved by better face-to-face interaction. Different cognitive styles in problem-solving, reasoning and decision-making hamper collaboration in research projects, but that can be remedied by the emergence of hybrid roles and organizations. Two ‘nudge’ suggestions made in the chapter are greater proximity between universities and companies and the emergence of a two-faced Janus scientist, a hybrid figure that combines the cognitive styles and values of both academic and industrial scientist.


15

While Viale’s chapter sees the collaboration between businesses and universities as being difficult owing to differences in values and cognitive styles, the thesis proposed by Paul David and Stanley Metcalfe in Chapter 2 proposes an alternative, but not incompatible, account. After the Bayh–Dole law, US universities are under increasing pressure to capitalize their knowledge and strengthen IPR (intellectual property rights). Collaboration with companies is not intermediated by faculty researchers but by employees of university ‘service units’ (e.g. technology transfer office, university research services, sponsored research office, external relations office). The main task of these offices is to sell licences and patents derived from the knowledge produced in university labs. Their main expertise is property right protection and avoidance of legal liability. When they have the chance to collaborate with companies on a joint research project, their risk adversity towards legal liability and concerns over IPR often hinders the agreement. Firms are always complaining about the proprietary approach taken by universities. Interaction has become difficult. As Stuart Feldman, IBM’s vice-president for computer science, explained to the New York Times: ‘Universities have made life increasingly difficult to do research with them because of all the contractual issues around intellectual property . . . We would like the universities to open up again’. Thus, in order to be more useful for companies, universities should become less entrepreneurial in managing IPR, according to IBM. The paradox is obvious and the thesis is counterintuitive. Most innovation policies in the USA and Europe were informed by the opposite equation: more knowledge transfer equals a more entrepreneurial university. Without reorienting the incentive structure of universities towards commercial aims, it seemed difficult to increase the knowledge transfer towards businesses. In short, if it is true that joint collaboration implies more similarity in background knowledge and cognitive styles between academic and industrial researchers, and if it is true that universities should abandon their overly aggressive IPR policy and that there is a growing need for open innovation and science commons in universities, the conclusion is that companies will bear the burden of cultural change. Their researchers should acquire more academic values and styles in order to pursue fruitful joint collaborations with universities. New ways of connecting universities with companies, like academic public spaces and new academic and industrial research roles, like the Janus scientists or professors of practice, must be introduced to support a rewarding collaboration. Older technology companies are especially fearful of universities creating competitors to themselves. They are concerned that by licensing IP to start-ups, including those founded by faculty and students, new firms may be created that will displace their prominence on the commercial

16


landscape. Thus, we have the irony of formerly closed firms espousing ‘open innovation’ and imploring universities to donate their IP rights. Not surprisingly, most universities take the innovation side of the IP debate and increasingly assist new firm formation. Universities are making a long-term bet that equity and employment created in a growth firm will redound to the benefit of the region where they are located and to the university’s endowment. As universities invest in the formation of firms from the IP they generate, they become closely tied to the venture capital phenomenon that, in an earlier era, they helped create. A clear example of ‘knowledge-driven capitalization of knowledge’ is the emergence of venture capitalism. Until the beginning of the twentieth century, technological knowledge was mainly idiosyncratic and tacit. These epistemological features limited the possibility of its tradability but also limited the need for the legal protection of intellectual property. With the widening of the scientific base of innovation, knowledge became more general and explicit. As tacitness decreased, so the tradability of knowledge and the need for IPR grew. Recognition of this epistemological change is essential to understanding the institutional phenomena that brought about the emergence of venture capitalism, the current major institution in the capitalization of knowledge. Cristiano Antonelli and Morris Teubal in Chapter 3 outline the evolutionary dynamics of the financial markets for innovation. Whoever finances innovation must assess two combined sets of risk: that the innovative project could fail and that the result cannot be appropriated by the inventor. The second risk is better faced by equity finance, e.g. corporate bodies, because the investors have the right to claim a share of the profits of successful companies whereas the lenders, e.g. banks, can claim only their credits. The first risk is better faced by banks because their polyarchic decision-making (i.e. great variety of expertise and number of experts less tied to vested interests) results in a higher chance of including outstanding projects, whereas the hierarchical (i.e. less variety of expertise and experts more closely tied to vested interests) decision-making of corporate bodies tends to favour only minor incremental innovation. Venture capitalism seems to be able to combine the advantages of both, that is the screening procedure performed by competent polyarchies, a distinctive feature of banks, and the equity-based provision of finance to new undertakings, a distinctive feature of corporations. Since the early days venture capital firms have specialized in the provision of ‘equity finance’ to new science-based start-up companies, together with business services and management advice. Limited partnerships, which were the leading form of organization for start-ups during the 1960s and 1970s, converged progressively into private stock companies


17

based upon knowledge-intensive property rights shares in the new sciencebased start-up companies. Private investors and financial companies elaborated exit strategies for collecting the value of these new firms after their creation and successful growth. Exit took place mainly through the sale of knowledge-intensive property rights. Initially these were private transactions over the counter. Later a public market emerged, characterized by automatic quoting mechanisms to report the prices and quantities of private transactions. This mechanism, better known as NASDAQ, evolved into a marketplace for selling knowledge-intensive property rights to the public at large. The demand for new knowledge-intensive property rights by investment funds, pension funds and retail investors accelerated the diffusion of NASDAQ with a snowball effect. The growing size of the market enabled it to become an efficient mechanism for identifying the correct value of knowledge-intensive property rights, a key function for the appreciation of the large share of intangible assets in the value of the new science-based companies. Intellectual property rights (IPR) have become the currency of technology deals, the ‘bargaining chips’ in the exchange of technology among different firms. What kind of justification is there for this strong emphasis on IPR? Are IPR the best way to strengthen the capitalization of knowledge? Is the current explosion of patents the right explanation for the innovation rate? Are there other factors, detached from proprietary incentives, capable of driving innovation pathways? For example, in the case of converging technologies, the high interdisciplinarity of their epistemological problem-solving dynamics tends to overcome any attempt to create proprietary monopolies of knowledge. The rate of change of knowledge is very fast; the knowledge useful for innovation is tacit in the minds of scientists; innovative problem-solving stems from the conceptual recombination and theoretical integration of different sources of knowledge; only open discussion, without IPR constraints, among academic and industrial researchers can generate the proper ‘gestalt shift’ that will afford the proper solution. Moreover, each particular body of knowledge drives the dynamics of knowledge change and capitalization of knowledge. The many components (ontic, deontic, epistemological and cognitive) of knowledge shape the directions of technological development towards given innovative products. They address the technological trajectories of technological paradigms described by Giovanni Dosi, Luigi Marengo and Corrado Pasquali in Chapter 4. The authors discuss the classic dilemma of the relation between IPR and innovation: on the one hand, the intellectual property monopolies afforded by patents or copyright raise product prices, while on the other, IPR provide a significant economic incentive for producing new knowledge. The answer to this question is not straightforward.

18


It is important to emphasize that, as far as product innovations are concerned, the most effective mechanisms are secrecy and lead time, while patents are the least effective, with the partial exception of drugs and medical equipment (Levin et al., 1987; Cohen et al., 2000). Moreover, the effects of IPR seem to be deleterious for innovation in the case of strongly cumulative technologies in which each innovation builds on previous ones. To the extent that a given technology is critical for further research, the attribution of broad property rights might hamper further developments. For example, in the case of the Leder and Stewart patent on the genetically engineered mouse that develops cancer, if the patent (the ‘onco-mouse’) protects all the class of products that could be produced (‘all transgenic non-human mammals’) or all the possible uses of a patented invention (a gene sequence), it represents a serious obstacle to research and innovation. On the other hand, Stanford University’s Office of Technology Licensing demonstrated that, by proactively licensing the Cohen–Boyer patent for recombinant DNA at reasonable rates, it helped create a new industry. How to navigate between the Scylla of creating fears due to appearance of ‘free riders’ in the absence of Clear IPR and the Charybdis of over protection stifling innovation is a persisting question. Symmetrically a ‘tragedy of anti-commons’ is likely to arise also when IP regimes give too many subjects the right to exclude others from using fragmented and overlapping pieces of knowledge with no one having the effective privilege of use. In the software industry, extensive portfolios of legal rights are considered means for entry deterrence and for infringement and counter-infringement suits against rivals. When knowledge is so finely subdivided into separate property claims on complementary chunks of knowledge, a large number of costly negotiations might be needed in order to secure critical licences. Finally, the history of innovation highlights many cases in which industry developed strongly with weak IPR regimes. For example, the core technologies of ICT – including transistors, semiconductors, software and telecommunication technologies like the mobile phone – were developed under weak IPR regimes. The organizational effect of the epistemological structure of knowledge is evident in some science-based innovations, such as biotechnology, in particular biopharmaceuticals. The high level of interdisciplinarity between biotechnology and biochemistry, informatics, mathematics, nanotechnology, biophysics, immunology and so on makes the knowledge very unstable. The potential problem-solving resulting from the intersection, hybridization and conceptual recombination of different and connected models and theories is very high. Thus there are many inventive solutions that continuously transform the field. As pointed out by Philip Cooke in Chapter 5, skills are in short supply and requirements change


19

rapidly. The potential inventions are in the scientists’ minds. Therefore the level of useful and crucial tacit knowledge is very high. Knowledge in the scientist’s mind is often taken as a tacit quasi proof-of-concept of an invention waiting to be disclosed. This is why we need proximity and direct interaction between academics and business. Interaction and face-to-face discussion can improve focusing on the right technological solution, making knowledge transfer easier. Thus, at the early stage of knowledge exploration, the cluster of dedicated biotechnological firms (DBFs) and research institutes is geographically localized. Only when the cluster enters the stage of knowledge exploitation does it tend to globalize. Knowledge-driven capitalization of knowledge is evident also in another way. Why did Germany experience so many difficulties and false starts in biotechnology? Because the predominant knowledge and methodology was that of organic chemistry and pharmaceutics in which Germany was the world leader. Germany tried to implant the embryo of a biotechnological industry in an epistemological environment characterized by the knowledge structure of industrial chemistry, the methodological techniques of pharmaceutics and the reasoning and decision-making processes of organic chemists. On the contrary, the UK epistemological environment was much more fertile because it was the birthplace of molecular biology, of which biotechnology is an applied consequence. Nevertheless, there was a significant gap between the discovery of the double helix and the rise of a UK biotech industry. In the interim, the US took world leadership through a plethora of ‘companies of their own’ founded by entrepreneurial scientists at universities that had made an early bet on molecular biology. Consequently, different epistemological path dependencies lead to different ways of capitalizing knowledge. Lastly, the birth of biopharma clusters is a clear example of a triple-helix model of innovation. The holy trinity of research institutes, DBFs and big pharmaceutical companies within the cluster is often triggered by local governments or national agencies. For example, the driver of cluster-building in Washington and Maryland’s Capitol region was, mainly, the National Institutes of Health. From this point of view, biopharma clusters neither conform to the Schumpeterian model nor to Porterian clusters. They tend to be more milieu than market. They need public financial support in order to implement research programmes. The phenomenon of knowledge-driven capitalization of knowledge is evident in the different architectures of capitalization depending on the different conceptual domains and disciplines. In a few cases the proprietary approach may be useful at the first stages of research, while in many other cases it is better only for downstream research. Sometimes for the same kind of knowledge a change happens in the way it is capitalized.

20


The change is linked to cultural and economic factors. In some instances the proprietary domain is perceived to be socially bad (because of hyperfocusing and exaggerated risk-perception of the consequences of the ‘tragedy of anti-commons’) and there are emergent successful cases of open innovation having good economic returns. Therefore scientific communities tend to shift from a proprietary approach to an open one and back again as their interest shifts from working with existing firms to starting new ones. Indeed, the same individual may pursue both courses of action simultaneously with different research lines or even with pieces of the same one. Software research is a case of a knowledge capitalization that has shifted in recent years from the first to the second category in upstream research, maintaining the proprietary domain only for the downstream stages. This is because if the proprietary domain is applied to the upstream basic software knowledge, there is a risk of lack of development and improvement that is achieved mainly through an open source approach. If the public domain is applied to the downstream commercial developments, there is a risk of lack of economic private resources because companies don’t see any real economic incentive to invest. Alfonso Gambardella and Bronwyn H. Hall in Chapter 6 analyse and explain in economic terms the evolution of the proprietary versus public domain in the capitalization of knowledge. The proprietary regime assigns clear property rights and provides powerful incentives at the cost of creating temporary monopolies that will tend to restrict output and raise prices. The public regime does not provide powerful incentives but the dissemination of knowledge is easy and is achieved at low cost. There is an alternative third system, that of ‘collective invention’. This system allows ‘the free exchange and spillover of knowledge via personnel contact and movement, as well as reverse engineering, without resort to intellectual property protection’. Collective invention in the steel and iron industry, in the semiconductor industry, and in the silk industry are some of the historical examples of this third alternative. The production of knowledge is supported by commercial firms that finance it through the sale of end products. The sharing of information is motivated since rewards come from product sales rather than information about incremental innovations. This system works when an industry is advancing and growing rapidly and the innovation areas are geographically localized, but it doesn’t work when an industry is mature or the innovation areas are not geographically localized. In these cases, how can an open science approach to upstream research be supported? Without coordination, scientists don’t perceive the advantages of a public rather than a proprietary approach, namely the utility of a larger stock of public knowledge and the visibility


21

of their research and achievements. Therefore, without coordination they tend to behave egoistically and collective action is hard to sustain (as in Mancur Olson’s famous theory of collective action). A policy device, particularly useful in software research, and that could sustain the right amount of coordination, is the Generalized Public License (GPL), also dubbed the copyleft system: ‘the producer of an open source program requires that all modifications and improvements of the program are subject to the same rules of openness, most notably the source code of all the modifications ought to be made publicly available like the original program’(Gambardella and Hall, this volume) In order to make the GPL function there is need of legal enforcement because the norms and social values of the scientific community, and the reputation effect of their infringement, seem to be insufficient. Part II deals with the growing importance of the triple-helix model in the knowledge economy. The generation of knowledge, in particular knowledge that can be capitalized, seems to be linked to emerging knowledge networks characterized by academy–industry–government interaction. As an institution, the university has changed its mission many times in the last hundred years. From the ivory tower of the German university model in the nineteenth century, focused mainly on basic research and education, the shift has been towards a growing involvement in solving social and economic problems at the beginning of the twentieth century. The growing involvement of universities in social and economic matters reached its acme with the birth and prevalence of the MIT–Stanford model. According to Henry Etzkowitz in Chapter 7 academy–industry relations are radically changing the actors of the knowledge economy. Universities are becoming the core of a new knowledge and creative economy. Paradoxically, it is by holding to the values of basic research that radical innovations with the highest market value are created. Capitalization of knowledge is becoming a central target of research policy in most American and many European universities. Technology transfer officers (TTOs) are acquiring a strong and proactive policy for IPR, the sale of licences and the creation of spin-off enterprises. Dual career is an ever more popular option. Academic scientists in some fields such as life sciences feel increasingly at ease in collaborating with companies. The market culture is not a novelty in the American university. Already during and especially after the 1980s the fall in public funding for research obliged academics to seek resources in a competitive way from companies, foundations and public agencies. The concept of the ‘quasi firm’ was born at that time. Researchers joined to form a group of colleagues with common research and also economic objectives. They competed with other groups for a slice of the funding cake. Universities that organized their funding

22


successfully through ‘quasi firms’, were the most suitable to become entrepreneurial and to capitalize knowledge. The capitalization of knowledge through IPR is losing ground according to the analysis made by Caroline Lanciano-Morandat and Eric Verdier in Chapter 8. National R&D policies can be divided into four categories based on numerous important factors: 1.

the Republic of Science is summarized by the Mertonian ethos and has as its main aim the development of codified knowledge. Its incentive structure is based on peer evaluation that implies disclosure and priority norms; 2. the state as an entrepreneur is based on the convention of a missionoriented public policy aimed at pursuing national priorities in technology and innovation. It uses top-down planning exemplified by the traditional French technology policy successful in big military projects or in aerospace and energy. The incentives shaping individual behaviour are mainly public power over the actors of the scientific and industrial worlds; 3. the state as a regulator promotes the transfer of scientific knowledge to the business world. The objectives of academic research should be shaped by the market expectations. The incentives are focused on the definition of property rights in order to promote the creation of hightech academic startups and the development of contractual relations between universities and firms; 4. the state as facilitator of technological projects is represented by the triple helix of the joint co-production of knowledge by universities, companies and public agencies. The emergence of hybrid organizations, strategic alliances and spin-offs relies on local institutional dynamics. The information generated by invention is ephemeral and rapidly depreciates due to the speed of technological change. This tends to reduce the protective role of contracts and IPR for the capitalization of knowledge. Therefore incentives in science and technology districts are fewer royalties and more capitalization through shares, revenues and stock options arising from participation in new industrial initiatives. Individual competences include the ability to cooperate, to work in networks and to combine different types of knowledge. ‘Janus scientists’ capable of interfacing knowledge and markets and of integrating different conceptual tools pertaining to different disciplines will become increasingly sought after in the labour market. This R&D policy is gaining ground in most Western countries. The triple helix is characterized by the birth of ever new and changing


23

hybrid organizations. One specific example of a hybrid organization is that of boundary organizations that group the representatives of the three helixes and aim to bridge the gap between science and politics. In Chapter 9 Sally Davenport and Shirley Leitch present a case study on the Life Sciences Network (LSN) in New Zealand that acted as the boundary between the position of the scientific community and industry and that of the political parties in the discussion about the moratorium on GMO. LSN increased the chances of pro-GMO arguments being accepted by public opinion because of its supposed neutrality and authoritative status compared to those of member organizations. In this way it increased the chances of achieving the common aims of the representatives of the three helixes. The economic role of the triple helix in the knowledge economy is more than just a way to capitalize knowledge. Knowledge is not economically important only because it can be capitalized. The concept of the knowledge economy is justified by a wider interpretation of the economic impact of knowledge. The economic utility of knowledge is present not only when it becomes an innovation. All economic activities involved in knowledge production and distribution are relevant. The problem is: how can we define what knowledge is and how can we measure its economic value? Chapter 10 by Benoît Godin introduces the pioneering work of Fritz Machlup. He tries to define knowledge with the help of epistemology, cybernetics and information theory. Knowledge is not only an explicit set of theories, hypotheses and empirical statements about the world. It is also the, often implicit and tacit, set of procedures, skills and abilities that allows individuals to interact in the real world. Or in the words of Ryle, and of Polanyi, knowledge is not only represented by ‘know-that’ statements but also by ‘know-how’ abilities (or to put it differently, by ontic and deontic knowledge). If knowledge does not need to be merely explicit, and a true linguistic representation of certified events and tested theories, and if it can also be subjective, conjectural, implicit and tacit, then it can include many expressions of social and economic life: practical, intellectual, entertaining, spiritual, as well as accidentally acquired. From an operational point of view, knowledge should be analysed in two phases: generation and transmission. Therefore R&D, education, media and information services and machines are the four operational elements of knowledge. Knowledge is not only a static concept, that is to say what we know, but a dynamic one, that is what we are learning to know. The first is knowledge as state, or result, while the second means knowledge as process, or activity. How can we measure knowledge? Not by using the Solow approach as a production function. By using this approach Solow formalized early works on growth accounting (breaking down GDP into capital and

24


labour) and equated the residual in his equation with technical change. According to Machlup, the production function is only an abstract construction that correlates input and output, without any causal meaning. The only reliable way to measure knowledge is by national accounting, that is the estimate of costs and sales of knowledge products and services (according to his broad definition). Where the data were not available, as in the case of internal production and the use of knowledge inside a firm, he looked at complementary data, such as occupational classes of the census, differentiating white-collar workers from workers who were not knowledge producers, like blue-collar workers. His policy prescriptions were in favour of basic research and sceptical about the positive influence of the patent system on inventive activity. Basic research is an investment, not a cost. It leads to an increase in economic output and productivity. Too much emphasis on applied research is a danger because it drives out pure research, which is its source of knowledge. Finally, his policy focus on information technologies was very supportive. Information technologies are a source of productivity growth because of improved records, improved decision-making and improved process controls, and are responsible for structural changes in the labour market, encouraging continuing movement from manual to mental and from less to more highly skilled labour. The knowledge economy is difficult to represent. The representation must not focus only on economic growth and knowledge institutions. It should focus also on the knowledge base and on the dynamic distribution of knowledge. To reach this goal, knowledge should not be represented only as a public good but as a mechanism for coordinating society. Machlup was the first to describe knowledge as a coordination mechanism when he qualified it in terms of the labour force. In Chapter 11, Loet Leydesdorff, Wilfred Dolfsma and Gerben Van der Panne try to define a model for measuring the knowledge base of an economy. In their opinion it can be measured as an overlay of communications among differently codified communications. The relevant variables are the territorial economy, its organization and technology. The methodological tools are scientometrics, which measures knowledge flow, and economic geography. Territorial economies are created by the proximity – in terms of relational dimensions – of organizations and technologies. New niches of knowledge production emerge as densities of relations and as a consequence of the self-organization of these interactions. The triple helix is an exemplification of this dynamics. It is the emergence of an overlay from the academy–industry–government interaction. In some cases feedback from the reflective overlay can reshape the network relations from which it emerged.


25

Academy–industry relations aim to establish knowledge networks that function as complex problem-solvers devoted to the generation and diffusion of knowledge. They are Complex Adaptive Systems (CADs), whose emergent dynamics are difficult to predict but whose micromotives driving the individual behaviours can be represented (Viale and Pozzali, 2010). According to Chapter 12 by Matilde Luna and José Luis Velasco, there are four mechanisms for integrating triple-helix actors with different and diverging norms, interests, resources, theories and abilities: trust, translation, negotiation and deliberation. As is highlighted in other chapters of the book, communication and collaboration between the members of the three helixes offset the difficulties posed by a different set of values, interests and skills. It is difficult for industrial scientists to coordinate with academic researchers if their perception of time and money is different: if for one the aim is commercial and for the other it is epistemological, or if one has expertise of the more practical problem-solving kind while the other tends towards more theoretical solutions. If there are too many differences, no trust can be generated and therefore collaboration is difficult. Moreover, often there is linguistic distance, and cognitive styles of reasoning and decision-making are different. This calls for translation between two worlds. This can be provided by players who fulfil a bridging role between academy and industry (e.g. technology transfer officers) or by translation provided by the scientists themselves (e.g. Janus scientists). Without translation it is impossible to find a rational ground for deliberation about the goals, methods, techniques and timescale of the research project. There can be only tiring and long negotiations, often with irrational and unbalanced results. Academy–industry relations can be assessed on two different functional and operative levels. If a knowledge network is capable of generating outputs that satisfy normative, epistemological and pragmatic desiderata, and if these outputs are achieved with the lowest costs (time, technical resources, money, physical effort etc.), they show a positive functional performance. If they become stable and ‘robust’ in their organizational structure and activities, they show a positive operative feature. Under these conditions The Capitalization of Knowledge both advances knowledge and presages a new mode of production beyond industrial society in which the university is co-equal with industry and government.

ACKNOWLEDGEMENTS We wish to thank Laura Angela Gilardi of Fondazione Rosselli for her invaluable help with the secretarial and editorial work. The book could

26


not have been completed without her support. Thanks to Chiara Biano for her editorial processing. We also thank Raimondo Iemma for supplying some of the data for the introduction. We also wish to thank the staff of Fondazione Rosselli and in particular Daniela Italia, Paola Caretta, Rocío Ribelles Zorita, Elisabetta Nay, Carlotta Affatato, Giulia Datta, Elena Bazzanini, Anna Mereu, Giovanni De Rosa, Michele Salcito, Maria Cristina Capetti, Francesca Villa, Fabiana Manca and Laura Alessi for the excellent organization of the conference and the follow-up initiatives.

REFERENCES Agrawal, A. and R. Henderson (2002), ‘Putting patents in context: exploring knowledge transfer from MIT’, Management Science, 48 (1), 44–60. Azoulay P., W. Ding and T. Stuart (2005), ‘The determinants of faculty patenting behavior: demographics or opportunities?’, NBER Working Paper 11348. Blumenthal, D., M. Gluck, K. Lewis, M. Stoto and D. Wise (1986), ‘University– industry relations in biotechnology: implications for the university’, Science, 232 (4756), 1361–66. Breschi, S., F. Lissoni and F. Montobbio (2007), ‘The scientific productivity of academic inventors: new evidence from Italian data’, Economics of Innovation and New Technology, 16 (2), 101–18. Calderini, M. and C. Franzoni (2004), ‘Is academic patenting detrimental to high quality research? An empirical analysis of the relationship between scientific careers and patent applications’, Università Bocconi, CESPRI Working Paper 162. Campbell, E.G., B.R. Clarridge, M. Gokhale, L. Birenbaum, S. Hilgartner, N.A. Holtzman and D. Blumenthal (2002), ‘Data withholding in academic genetics’, JAMA, 287 (4), 473–80. Cohen, W., R.R. Nelson and J. Walsh (2000), ‘Protecting their intellectual assets: appropriability conditions and why US manufacturing firms patent or not’, NBER Discussion Paper 7552. Dasgupta, P. and P. David (1985), ‘Information disclosure and the economics of science and technology’, CEPR Discussion Papers 73. David, P.A. (1998), ‘Common agency contracting and the emergence of “open science” institutions’, The American Economic Review, 88 (2), 15–21. Etzkowitz, H. (2000), ‘Bridging the gap: the evolution of industry–university links in the United States’, in L. Branscomb et al. (eds), Industrializing Knowledge, Cambridge: MIT Press, pp. 203–33. Etzkowitz, H. (2008), The Triple Helix: University–Industry–Government Innovation in Action, London: Routledge. Etzkowitz, H. and M. Ranga (2009), ‘A transKeynesian vision of government’s role in innovation: picking winners revisited’, Science and Public Policy, 36 (10), 799–808. Etzkowitz, H. and M. Ranga (2010), ‘From spheres to spaces: a triple helix system


27

for knowledge-based regional development’, http://www.triplehelix8.org/, last accessed 22 June 2010. Etzkowitz, H. and R. Viale (2009), ‘The third academic revolution: polyvalent knowledge and the future of the university’, Critical Sociology, 34: 4 July 2010. Etzkowitz, H., M. Gulbrandsen and J. Levitt (2001), Public Venture Capital, 2nd edn, New York: Aspen Kluwer. Fabrizio, K.R. and A. DiMinin (2008), ‘Commercializing the laboratory: faculty patenting and the open science environment’, Research Policy, 37 (5), 914–31. Gibbons, M., C. Limoges, H. Nowotny, S. Schwartzman, P. Scott and M. Trow (1994), The New Production of Knowledge: The Dynamics of Science and Research in Contemporary Societies, Sage: London. Jensen, R. and M. Thursby (2003), ‘The academic effects of patentable research’, paper presented at the NBER Higher Education Meeting, 2 May, Cambridge, MA. Lakatos, I. (1970), ‘Falsification and the methodology of scientific research programmes’, in I. Lakatos and A. Musgrave (eds), Criticism and the Growth of Knowledge, Cambridge: Cambridge University Press, pp. 51–8. Levin, R., A. Klevorick, R.R. Nelson and S. Winter (1987), ‘Appropriating the returns from industrial R&D’, Brookings Papers on Economic Activity, 18 (3), 783–832. Mansfield, E. (1995), ‘Academic research underlying industrial innovation: sources, characteristics, and financing’, Review of Economics and Statistics, 77 (1), 55–65. Ranga, L.M., K. Debackere and N. von Tunzelman (2003), ‘Entrepreneurial universities and the dynamics of academic knowledge production: a case of basic vs. applied research in Belgium’, Scientometrics, 58 (2), 301–20. Siegel, D., D. Waldman and A. Link (1999), ‘Assessing the impact of organizational practices on the productivity of university technology transfer offices: an exploratory study’, NBER Working Paper 7256. Stephan, P.E. and S.G. Levin (1996), ‘Property rights and entrepreneurship in science’, Small Business Economics, 8 (3), pp. 177–88. Stephan, P.E., S. Gurmu, A.J. Sumell and G. Black (2007), ‘Who’s patenting in the university? Evidence from the survey of doctorate recipients’, Economics of Innovation and New Technology, 16 (2), 71–99. Stokes, D.E. (1997), Pasteur’s Quadrant: Basic Science and Technological Innovation, Washington, DC: Brookings Institution Press. Thaler, R.H. and C.R. Sunstein (2008), Nudge. Improving Decisions About Health, Wealth, and Happiness, New Haven, CT: Yale University Press. Van Looy, B., L.M. Ranga, J. Callaert, K. Debackere and E. Zimmerman (2004), ‘Balancing entrepreneurial and scientific activities: feasible or mission impossible? An examination of the performance of university research groups at K.U. Leuven, Belgium’, Research Policy, 33 (3), 443–54. Viale, R. and A. Pozzali (2010), ‘Complex adaptive systems and the evolutionary triple helix’, Critical Sociology, 34.

PART I

How to capitalize knowledge

1.

Knowledge-driven capitalization of knowledge Riccardo Viale

INTRODUCTION Capitalization of knowledge happens when knowledge generates an economic added value. The generation of economic value can be said to be ‘direct’ when one sells the knowledge for some financial, material or behavioural good. The generation of economic value is considered ‘indirect’ when it allows the production of some material or service goods that are sold on the market. The direct mode comprises the sale of personal know-how, such as in the case of a plumber or of a sports instructor. It also comprises the sale of intellectual property, as in the case of patents, copyrights or teaching. The indirect mode comprises the ways in which organizational, declarative and procedural knowledge is embodied in goods or services. The economic return in both cases can be financial (e.g. cash), material (e.g. the exchange of consumer goods) or behavioural (e.g. the exchange of personal services). In ancient times, the direct and indirect capitalization of knowledge was based mainly on procedural knowledge. Artisans, craftsmen, doctors and engineers sold their know-how in direct or indirect ways within a market or outside of it. Up to the First Industrial Revolution, the knowledge that could be capitalized remained mainly procedural. Few were the inventors that sold their designs and blueprints for the construction of military or civil machines and mechanisms. There were some exceptions, as in the case of Leonardo da Vinci and several of his inventions, but, since technological knowledge remained essentially tacit, it drove a capitalization based primarily on the direct collaboration and involvement of the inventors in the construction of machines and in the direct training of apprentices. In the time between the First and Second Industrial Revolutions, there was a progressive change in the type of knowledge that could be capitalized. The ‘law of diminishing returns’, as it manifested itself in the economic exploitation of invention, pushed companies and inventors, lacking a scientific base, to look for the causal explanation of innovations (Mokyr, 2002a, 2002b). For example, Andrew Carnegie, Eastman Kodak, DuPont, 31

32


AT&T, General Electric, Standard Oil, Alcoa and many others understood the importance of scientific research for innovation (Rosenberg and Mowery, 1998). Moreover, the revolution in organic chemistry in Germany shifted industrial attention towards the fertility of collaboration between universities and companies. Searching for a scientific base for inventions meant developing specific parts of declarative knowledge. Depending on the different disciplines, knowledge could be more or less formalized and could contain more or fewer tacit features. In any case, from the Second Industrial Revolution onwards, the capitalization of technological knowledge began to change: a growing part of knowledge became protected by intellectual property rights (IPR); patents and copyrights were sold to companies; institutional links between academic and industrial laboratories grew; companies began to invest in R&D laboratories; universities amplified the range and share of applied and technological disciplines and courses; and governments enacted laws to protect academic IPR and introduced incentives for academy–industry collaboration. New institutions and new organizations were founded with the aim of strengthening the capitalization of knowledge. The purpose of this chapter is to show that one of the important determinants of the new forms of the capitalization of knowledge is its epistemological structure and cognitive processing. The thesis of this chapter is that the complexity of the declarative part of knowledge and the three tacit dimensions of knowledge – competence, background and cognitive rules (Pozzali and Viale, 2007) – have a great impact on research behaviours and, consequently, on the ways of capitalizing knowledge. This behavioural impact drives academy–industry relations towards greater face-to-face interactions and has led to the development of a new academic role, that of the ‘Janus scientist’1. The need for stronger and more extensive face-to-face interaction is manifested through the phenomenon of the close proximity between universities and companies and through the creation of hybrid organizations of R&D. The emergence of the new academic role of the Janus scientist, one who is able to interface both with the academic and industrial dimensions of research, reveals itself through the introduction of new institutional rules and incentives quite different from traditional academic ones.

EPISTEMOLOGICAL AND COGNITIVE CONSTRAINTS IN KNOWLEDGE TRANSFER Scientific knowledge is variegated according to different fields and disciplines. The use of formal versus natural language, conceptual complexity

Knowledge-driven capitalization of knowledge

33

versus simplicity, and explicit versus tacit features of knowledge vary a great deal from theoretical physics to entomology (to remain within the natural sciences). Different epistemological structures depend mainly on the ontology of the relative empirical domain. For example, in particle physics the ontology of particles allows the use of numbers and of natural laws written in mathematical language. On the contrary, in entomology the ontology of insects allows us to establish empirical generalizations expressed in natural language. Different epistemological structures mean different ways of thinking, reasoning and problem-solving. And this cognitive dimension influences behavioural and organizational reality. To better illustrate the role of epistemological structure, I shall introduce several elementary epistemological concepts. Knowledge can be subdivided into the categories ontic and deontic. Ontic knowledge analyses how the world is, whereas deontic knowledge is focused on how it can be changed. These two forms of knowledge can be represented according to two main modes: the analytical mode deals with the linguistic forms that we use to express knowledge; the cognitive mode deals with the psychological ways of representing and processing knowledge. Two main epistemological features of knowledge influence the organizational means of knowledge generation and transfer. The first is the rate of generality. The more general the knowledge is, the easier it is to transfer and apply it to subjects different from those envisioned by the inventor. The second is complexity. The more conceptually and computationally complex the knowledge is, the more there will be a concomitant organizational division of work in problem-solving and reasoning. 1

Analytical Mode of Ontic Knowledge

Analytical ontic knowledge is divided into two main types, descriptive and explanatory. Descriptive The first type comprises all the assertions describing a particular event according to given space-time coordinates. These assertions have many names, such as ‘elementary propositions’ or ‘base assertions’. They correspond to the perceptual experience of an empirical event by a human epistemic agent at a given time.2 A descriptive assertion has a predicative field limited to the perceived event at a given time. The event is exceptional because its time-space coordinates are unique and not reproducible. Moreover, this uniqueness is made stronger by the irreproducibility of the perception of the agent. Even if the same event were reproducible, the perception of it would be different because of the continuous changes in

34


perceptual ability. Perception is related to cortical top-down influences corresponding to schemes, expectations, frames and other conceptual structures that change constantly. The perception of an event causes a change in the conceptual categorization to which the event belongs. This change can modify the perception of a similar event that happens afterwards.3 Therefore a singular descriptive assertion can correspond only to the timespace particular perception of a given epistemic agent and cannot have any general meaning. For example, the observational data in an experiment can be transcribed in a laboratory diary. An observational assertion has only a historical meaning because it cannot be generalized. Therefore a process technique described by a succession of descriptive assertions made by an epistemic agent cannot be transferred and replicated precisely by a different agent. It will lose part of its meaning and, consequently, replication will be difficult. Inventions, before and during the First Industrial Revolution, were mainly represented as a set of idiosyncratic descriptive assertions made by the inventor. Understanding the assertions and replicating the data were only possible for the inventor. Therefore technology transfer was quite impossible at that time. Moreover, the predicative field of an invention was narrow and fixed. It applied only to the events described in the assertions. There was no possibility of enlarging the semantic meaning of the assertions, that is, of enlarging the field of the application of the invention in order to produce further innovations. As a result, the law of diminishing returns manifested itself very quickly and effectively. Soon the economic exploitation of the invention reached its acme, and the diminishing returns followed. Only a knowledge that was based not on descriptive assertions but on explanatory ones could provide the opportunity to enlarge and expand an invention, to generate corollary innovations and, thus, to invalidate the law of diminishing returns. This economic motive, among others, pushed inventors and, mainly, entrepreneurs to look for the explanatory basis of an invention, that is, to pursue collaborations with university labs and to establish internal research and development labs (Mowery and Rosenberg, 1989; Rosenberg and Birdzell, 1986). Explanatory These assertions, contrary to descriptive ones, have a predicative field that is wide and unfixed. They apply to past and future events and, in some cases (e.g. theories), to events that are not considered by the discoverer. They can therefore allow the prediction of novel facts. These goals are achieved because of the syntactic and semantic complexity and flexibility of explanatory assertions. Universal or probabilistic assertions, such as the inductive generalization of singular observations (e.g. ‘all crows are


35

black’ or ‘a large percentage of crows are black’) are the closest to descriptive assertions. They have little complexity and their application outside the predicative field is null. In fact, their explanatory and predictive power is narrow, and the phenomenon is explained in terms of the input–output relations of a ‘black box’ (Viale, 2008). In contrast, theories and models tend to represent inner parts of a phenomenon. Usually, hypothetical entities are introduced that have no direct empirical meaning. Theoretical entities are then linked indirectly to observations through bridge principles or connecting statements. Models and metaphors often serve as heuristic devices used to reason more easily about the theory. The complexity, semantic richness and plasticity of a theory allow it to have wider applications than empirical generalizations. Moreover, theories and models tend not to explain a phenomenon in a ‘black box’ way, but to represent the inner mechanisms that connect input to output. Knowing the inner causal mechanisms allows for better management of variables that can change the output. Therefore they offer better technological usage. Inductive generalizations were the typical assertions made by individual inventors during the First Industrial Revolution. Compared to descriptive assertions, they represent progress because they lend themselves to greater generalization. They thus avoid being highly idiosyncratic and, in principle, can be transferred to other situations. Nevertheless, inductive generalizations are narrow in their epistemological meaning and don’t allow further enlargement of the invention. This carries with it the inevitable consequence of their inability to generate other innovations. Therefore inductive generalizations are fixed in the law of diminishing returns. In contrast, theories attempting to give causal explanations of an invention presented the ability to invalidate the law of diminishing returns. They opened the black box of the invention and allowed researchers to better manipulate the variables involved in order to produce different outputs, or rather, different inventions. A better understanding of inventions through the discovery of their scientific theoretical bases began to be pursued during and after the Second Industrial Revolution. To better exemplify the relations between descriptive assertions, empirical generalizations and theories in technological innovation, I now describe a historical case (Viale, 2008, pp. 23–5; Rosenberg and Birdzell, 1986). At the end of the eighteenth century and the beginning of the nineteenth, the growth in urban populations, as people moved out of the country to search for work in the city, posed increasing problems for the provisioning of food. Long distances, adverse weather conditions and interruptions in supplies as a result of social and political unrest meant that food was often rotten by the time it reached its destination. The authorities urgently

36


needed to find a way to preserve food. In 1795, at the height of the French Revolution, Nicholas Appert, a French confectioner who had been testing various methods of preserving edibles using champagne bottles, found a solution. He placed the bottles containing the food in boiling water for a certain length of time, ensuring that the seal was airtight. This stopped the food inside the bottle from fermenting and spoiling. This apparently commonplace discovery would be of fundamental importance in years to come and earned Appert countless honours, including a major award from the Napoleonic Society for the Encouragement of Industry, which was particularly interested in new victualling techniques for the army. For many years, the developments generated by the original invention were of limited application, such as the use of tin-coated steel containers introduced in 1810. When Appert developed his method, he was not aware of the physical, chemical and biological processes that prevented deterioration once the food had been heated. His invention was a typical example of know-how introduced through ‘trial and error’. The extension of the invention into process innovation was therefore confined to descriptive knowledge and to an empirical generalization. It was possible to test new containers or to try to establish a link between the change in temperature, the length of time the container was placed in the hot water and the effects on the various bottled foods, and to then draw up specific rules for the preservation of food. However, this was a random, time-consuming approach, involving endless possible combinations of factors and lacking any real capacity to establish a solid basis for the standardization of the invention. Had it been a patent, it would have been a circumscribed innovation, whose returns would have been high for a limited initial period and would then have gradually decreased in the absence of developments and expansions of the invention itself.4 The scientific explanation came some time later, in 1873. Louis Pasteur discovered the function of bacteria in certain types of biological activity, such as in the fermentation and deterioration of food. Microorganisms are the agents that make it difficult to preserve fresh food, and heat has the power to destroy them. Once the scientific explanation was known, chemists, biochemists and bacteriologists were able to study the effects of the multiple factors involved in food spoilage: ‘food composition, storage combinations, the specific microorganisms, their concentration and sensitivity to temperature, oxygen levels, nutritional elements available and the presence or absence of growth inhibitors’ (Rosenberg and Birdzell, 1986; Italian translation 1988, pp. 300–301). These findings and many others expanded the scope of the innovation beyond its original confines. The method was applied to varieties of fruit, vegetables and, later, meats that could be heated. The most suitable type of container was identified, and the effects of canning on the food’s


37

flavour, texture, colour and nutritional value were characterized. As often happens when the scientific bases of an invention are identified, the innovation generated a cascade effect of developments in related fields, such as insulating materials, conserving-agent chemistry, and genetics and agriculture for the selection and cultivation of varieties of fruit and vegetables better suited to preservation processes. Why is it that the scientific explanation for an invention can expand innovation capacity? To answer this question, reference can again be made to Appert’s invention and Pasteur’s explanation. When a scientific explanation is produced for a phenomenon, two results are obtained. First, a causal relationship is established at a more general level. Second, once a causal agent for a phenomenon has been identified, its empirical characteristics can be analysed. As far as the first result is concerned, the microbic explanation furnished by Pasteur does not apply simply to the specific phenomenon of putrefaction in fruit and vegetables; bacteria have a far more general ability to act on other foods and to cause pathologies in people and animals. The greater generality of the causal explanation compared with the ‘local’ explanation – the relationship between heat and the preservation of food – means that the innovation can be applied on a wider scale. The use of heat to destroy microbes means that it is possible to preserve not only fruit and vegetables, but meat as well, to sterilize milk and water, to prepare surgical instruments before an operation, to protect the human body from bacterial infection (by raising body temperature) and so on. All this knowledge about the role of heat in relation to microbes can be applied in the development of new products and new innovative processes, from tinned meat to autoclaves to sterilized scalpels. As to the second result, once a causal agent has been identified, it can be characterized and, in the case of Pasteur’s microbes, other methods can be developed to neutralize or utilize them. An analysis of a causal agent begins by identifying all possible varieties. Frequently, as in the case of microbes, the natural category identified by the scientific discovery comprises a huge range of entities. And each microbe – bacterium, fungus, yeast, and so on – presents different characteristics depending on the natural environment. Some of these properties can be harnessed for innovative applications. Yeasts and bacilli, for example, are used to produce many kinds of food, from bread and beer to wine and yoghurt; bacteria are used in waste disposal. And this has led scientists, with the advent of biotechnology, to attempt to transform the genetic code of microorganisms in order to exploit their metabolic reactions for useful purposes. Returning to our starting point, the preservation of food, once the agent responsible for putrefaction had been identified, it also became possible to extend the range of methods used to neutralize it. It was eventually discovered that microbes could also be destroyed with

38


ultraviolet light, with alcohol or with other substances, which, for a variety of reasons, would subsequently be termed disinfectants. So to answer our opening question, the scientific explanation for an invention expands the development potential of the original innovation because it ‘reduces’ the ontological level of the causes and extends the predicative reach of the explanation. Put simply, if the phenomenon to be explained is a ‘black box’, the explanation identifies the causal mechanisms inside the box (‘reduction’ of the ontological level) that are common to other black boxes (extension of the ‘predicative reach’ of the explanation). Consequently, it is possible to develop many other applications or micro-inventions, some of which may constitute a product or process innovation. Innovation capacity cannot be expanded, however, when the knowledge that generates the invention is simply an empirical generalization describing the local relationship between antecedent and causal consequent (in the example of food preservation, the relationship between heat and the absence of deterioration). In this case, knowledge merely describes the external reality of the black box, that is, the relationship between input (heat) and output (absence of deterioration); it does not extend to the internal causal mechanisms and the processes that generate the causal relationship. It is of specific, local value, and may be applied to other contexts or manipulated to generate other applications to only a very limited degree. The knowledge inherent in Appert’s invention, which can be described as an empirical generalization, is regarded as genuine scientific knowledge by some authors (Mokyr, 2002a, Italian translation 2004). We do not want to get involved here in the ongoing epistemological dispute over what constitutes scientific knowledge (Viale, 1991): accidental generalizations of ‘local’ value only (e.g. the statement ‘the pebbles in this box are black’), empirical generalizations of ‘universal’ value (e.g. Appert’s invention) and causal nomological universals (e.g. a theory such as Pasteur’s discovery). The point to stress is that although an empirical generalization is ‘useful’ in generating technological innovation (useful in the sense adopted by Mokyr, 2002b, p. 25, derived from Kuznets, 1965, pp. 84–7), it does not possess the generality and ontological depth that permit the potential of the innovation to be easily expanded in the way that Pasteur’s discovery produced multiple innovative effects. In conclusion, after Pasteur’s discovery of the scientific basis of Appert’s invention, a situation of ‘increasing economic returns’ developed, driven by the gradual expansion of the potential of the innovation and a causal concatenation of micro-inventions and innovations in related areas. This could be described as a recursive cascade phenomenon, or as a ‘dual’ system (Kauffman, 1995), where the explanation of the causal mechanism for putrefaction gave rise to a tree


39

structure of other scientific problems whose solution would generate new technological applications and innovations, and also raise new problems in need of a solution. The example of Appert’s invention versus Pasteur’s discovery is also revealed in another phenomenon. The discovery of the scientific basis of an invention allows the horizontal enlargement of the invention into areas different from the original one (e.g. from alimentation to hygiene and health). The interdisciplinarity of inventive activities has grown progressively from the Second Industrial Revolution until now, with the recent birth of the converging technology programme (National Science Foundation, 2002). The new technologies are often the result of the expansion of a theory outside its original borders. This phenomenon implies the participation of different disciplines and specializations in order to be able to understand, grasp and exploit the application of the theory. The strong interdisciplinarity of current inventions implies a great division of expert labour and increased collaboration among different experts in various disciplines. Thus, only complex organizations supporting many different experts can cope with the demands entailed in the strong interdisciplinarity of current inventive activity. 2

Cognitive Mode of Ontic Knowledge

The cognitive approach to science (Giere, 1988; Viale, 1991; Carruthers et al., 2002; Johnson-Laird, 2008) considers scientific activity as a dynamic and interactive process between mental representation and processing on the one hand, and external representation in some media by some language on the other. According to this approach, scientific knowledge has two dimensions: the mental representations of a natural or social phenomenon; and its linguistic external representation. The first dimension includes the mental models stemming from perceptive and memory input and from their cognitive processing. This cognitive processing is mainly inductive, deductive or abductive. The models are realized by a set of rules: heuristics, explicit rules and algorithms. The cognitive processing and progressive shaping of the mental representations of a natural phenomenon utilize external representations in natural or formal language. The continuous interaction between the internal mental representation and the external linguistic one induces the scientist to generate two products: the mental model of the phenomenon and its external propositional representation. What is the nature of the representation of knowledge in the mind? It seems to be different in the case of declarative (ontic) knowledge than in the case of procedural (deontic) knowledge. The first is represented by

40


networks, while the second is represented by production systems. The ACT-R (Adaptive Control of Thought–Rational) networks of Anderson (1983, 1996) include images of objects and corresponding spatial configurations and relationships; temporal information, such as relationships involving the sequencing of actions, events and the order in which items appear; and information about statistical regularities in the environment. As with semantic networks (Collins and Quillian, 1969) or schemas (Barsalou, 2000), there is a mechanism for retrieving information and a structure for storing it. In all network models, a node represents a piece of information that can be activated by external stimuli, such as sensations, or by internal stimuli, such as memories or thought processes. Given each node’s receptivity to stimulation from neighbouring nodes, activation can easily spread from one node to another. Of course, as more nodes are activated and the spread of activation reaches greater distances from the initial source of the activation, the activation weakens. In other words, when a concept or a set of concepts that constitutes a theory contains a wide and dense hierarchy of interconnected nodes, the connection of one node to a distant node will be difficult to detect. It will, therefore, be difficult to pay the same attention to all the consequences of a given assertion. For example (Sternberg, 2009), as the conceptual category of a predicate (e.g. ‘animal’) becomes more hierarchically remote from the category of the subject of the statement (e.g. ‘robin’), people generally take longer to verify a true statement (e.g. ‘a robin is an animal’) in comparison with a statement that implies a less hierarchically remote category (e.g. ‘a robin is a bird’). Moreover, since working memory can process only a limited amount of information (according to Miller’s magical number 5 ± 2 items) a singular mind cannot compute a large amount of structured information or too many complex concepts, such as those contained in theories. These cognitive aspects explain various features of knowledge production and capitalization in science and technology. First, the great importance given to external representation in natural and formal language and the institutional value of publication satisfy two goals: because of the natural limitations of memory, these serve as memory devices, useful for allowing the cognitive processing of perceptive and memory input; because of the complexity of concepts and the need for different minds working within the same subject, these are social devices, useful for allowing the communication of knowledge and the interaction and collaboration among peers. Second, before and during the First Industrial Revolution, the computational effort of inventors was made apparent primarily in their perceptual ability in detecting relevant features of the functioning of machines and prototypes and in elaborating mental figurative concepts or models


41

that were depicted externally in diagrams, designs, figures, flow charts, drafts, sketches and other representations. The single mind of an inventor could cope with this computational burden. Interaction with other subjects consisted mainly of that with artisans and workers in order to prepare and tune the parts of a machine or of that with apprentices involving knowledge transfer. Few theoretical concepts were needed, and cognitive activity was focused on procedural knowledge (i.e. practical know-how represented, mentally, by production systems) and simple declarative knowledge (i.e. simple schemes that generalize physical phenomena, like the Appert scheme involving the relation between heat and food preservation). The situation changes dramatically after the Second Industrial Revolution with the growing role of scientific research, particularly in the life sciences. Conceptual categories increase in number; concepts become wider with many semantic nodes; and there are increasing overlaps among different concepts. One mind alone cannot cope with this increased complexity, so a growing selective pressure to share the computational burden between different minds arose. The inadequacy of a single mind to manage this conceptual complexity brought about the emergence of the division of expert labour, or in other words, the birth of specializations, collective organizations of research and different roles and areas of expertise.5 Third, knowledge complexity and limited cognition explain the emergence of many institutional phenomena in scientific and technological research, such as the importance of publication, the birth of disciplines and specializations, the division of labour and the growth in the size of organizations. What were the effects of these emergent phenomena on the capitalization of knowledge? While the inventor in the First Industrial Revolution could capitalize his knowledge by ‘selling his mind’ and the incomplete knowledge represented in the patent or in the draft, since the Second Industrial Revolution, many minds now share different pieces of knowledge that can fill the gaps in knowledge contained in the patent or publication. This is particularly true in technological fields, where the science push dimension is strong. In an emerging technology like biotechnology, nanotechnology, ICT, or in new materials and in the next converging technology (National Science Foundation, 2002), the complexity of knowledge and its differentiation lead to interdisciplinary organizations and collaborations and to the creation of hybrid organizations. Knowledge contained in a formal document, be it patent, publication or working paper, is not the full representation of the knowledge contained in the invention. There are tacit aspects of the invention that are crucial to its transfer and reproduction and are linked to the particular conceptual interpretation and understanding of the invention occasioned by the peculiar background knowledge and cognitive rules of the inventors (Balconi et

42


al., 2007; Pozzali and Viale, 2007). Therefore the only way to allow transfer is to create hybrid organizations that put together, face-to-face, the varied expertise of inventors with that of entrepreneurs and industrial researchers aiming to capitalize knowledge through successful innovations.

BACKGROUND KNOWLEDGE AND COGNITIVE RULES An important part of knowledge is not related to the representation of the physical and human world but to the ways in which we can interact with it. Deontic knowledge corresponds to the universe of norms. Rules, prescriptions, permissions, technical norms, customs, moral principles and ideal rules (von Wright, 1963) are the main categories of norms. Various categories of norms are implied in research and technological innovation. Customs or social norms represent the background knowledge that guides and gives meaning to the behaviour of scientists and industrial researchers. Some norms are moral principles and values that correspond to a professional deontology or academic ethos (Merton, 1973). They represent norms for moral actions and are slightly different from ideal rules (Moore, 1922), which are a way of characterizing a model of goodness and virtue (as in the Greek meaning of arête), or in this case, of what it means to be a good researcher. Prescriptions and regulations are the legal norms established by public authorities that constrain research activity and opportunities for the capitalization of knowledge. Rules are mainly identifiable in reasoning cognitive rules applied in solving problems, drawing inferences, making computations and so forth. Lastly, technical norms are those methodological norms and techniques that characterize the research methodology and procedures in generating and reproducing a given innovation. From this point of view, it is possible to assert that a scientific theory or a technological prototype is a mixture of ontic knowledge (propositions and mental models) and deontic knowledge (values, principles, methodologies, techniques, practical rules and so on). Deontic knowledge has been examined analytically as involving a logic of action by some authors (e.g. von Wright, 1963; but see also the work of Davidson, Chisholm and Kenny).6 An analytic approach has been applied primarily to the representation of legal and moral norms. For the purposes of this chapter, the analytic mode of deontic knowledge doesn’t appear relevant. First, it is difficult to apply a truth-functional logic to norms whose ontological existence is not clear. Second, unlike ontic knowledge, where the knowledge relevant for technological innovation is, to a certain extent, expressed in some language and transcribed in some media, deontic


43

knowledge relevant for technological innovation is mainly present at a socio-psychological level. As we shall see, norms are greatly involved in shaping the behaviours responsible for knowledge capitalization. Moral principles and social values are part of the background knowledge that influences the social behaviour of scientists and entrepreneurs, as well as the modalities of their interaction and collaboration. They play an important role in determining different styles of thinking, problem-solving, reasoning and decisionmaking between academic and industrial researchers (Viale, 2009) that can also have an effect on shaping the institutions and organizations for capitalizing knowledge. Before analysing background knowledge and cognitive rules, I wish to focus briefly on technical norms. According to von Wright (1963), technical norms correspond to the means of reaching a given goal. Analytically, they can be represented as conditional assertions of the elliptical form ‘if p, then q’ where the antecedent p is characterized by the goal and the consequent q by the action that should be taken to reach the goal. They represent the bridge between ontic and deontic knowledge. In fact, the antecedent is characterized not only explicitly by the goal but also implicitly by the empirical initial conditions and knowledge that allow the selection of the proper action. In other words, a technical norm would be better represented in the following way: if (p & a), then q where a represents the empirical initial conditions and theoretical knowledge for action. From this analytical representation of technical norms we infer certain features: (1) the more a corresponds to complex theoretical knowledge, the more computationally complex the application of the norm will be; (2) the more a corresponds to descriptive assertions, the more difficult it will be to generalize the understanding and application of the norm; and (3) the more the relevant knowledge contained in a is characterized by tacit features, the more difficult it will be to generalize to others the understanding and application of the norm. Technical norms corresponding to the procedures and techniques needed to generate an invention can manifest these features. Inventions from the First Industrial Revolution, such as Appert’s, presented technical norms characterized by descriptive assertions and tacit knowledge (mainly of the competential type). Thus knowledge transfer was very difficult, and there was no need for the division of expert labour. Inventions after the Second Industrial Revolution, however, involved a growing share of theoretical knowledge and a decrease in competential tacit knowledge; therefore the transfer of knowledge could be, in theory, easier. In any case, it required a greater amount of expertise that was possible only with a complex division of expert labour. This was particularly necessary in disciplines such as physics and chemistry, where the particular

44


ontology and mathematical language to represent the phenomena allow the generation of complex theoretical structures. From a cognitive point of view, technical norms correspond to ‘pragmatic schemes’ (Cheng and Holyoak, 1985, 1989) that have the form of production systems composed of condition–action rules (corresponding to conditional assertions in logic and to production rules in artificial intelligence). Pragmatic schemes are a set of abstract and context-dependent rules corresponding to actions and goals relevant from a pragmatic point of view. According to the analytical formulation of von Wright (1963), the main cognitive rules in pragmatic schemes are that of permission and obligation. More generally, a schema (an evolution of the ‘semantic network’ of Collins and Quillian, 1969) is a structured representation that captures the information that typically applies to a situation or event (Barsalou, 2000). Schemas establish a set of relations that link properties. For example, the schema for a birthday party might include guests, gifts, a cake and so on. The structure of a birthday party is that the guests give gifts to the birthday celebrant, everyone eats cake and so on. The pragmatic schema links information about the world with the goal to be attained according to this information. A pragmatic schema can serve as a cognitive theory for most deontic knowledge relevant in innovation. It can represent values and principles characterizing background knowledge. Social norms ruling research behaviour, moral principles characterizing the ethos of the academic community, pragmatic goals driving the decision-making of industrial researchers and social values given to variables such as time, risk, money and property can be represented by pragmatic schemes. These schemes also seem to influence the application of cognitive rules, such as those used in deduction, induction, causality, decision-making and so forth. The topic is controversial. The dependence of cognitive rules on pragmatic schemes is not justified by theories supporting an autonomous syntactic mental logic. According to these theories (Beth and Piaget, 1961; Braine, 1978; Rumain, et al., 1983), the mind contains a natural deductive logic (which for Piaget offers the propositional calculus) that allows the inference of some things and not others. For example, the human mind is able to apply modus ponens but not modus tollens. In the same way, we could also presuppose the existence of a natural probability calculus, a causal reasoning rule and a risk assessment rule, among others. Many empirical studies and several good theories give alternative explanations that neglect the existence of mental logic and of other syntactic rules (for the pragmatic scheme theories, see Cheng and Holyoak, 1985, 1989; Cheng and Nisbett, 1993; for the mental models theory, see Johnson-Laird, 1983, 2008; for the conceptual semantic theory see Jackendoff, 2007). The first point is that there


45

are many rules that are not applied when the format is abstract but are applied when the format is pragmatic – that is, when it is linked to everyday experience. For example, the solution of the ‘selection task problem’, namely, the successful application of modus tollens, is possible only when the questions are not abstract but are linked to problems of everyday life (Politzer, 1986; Politzer and Nguyen-Xuan, 1992). The second point is that most of the time rules are implicitly learned through pragmatic experience (Reber, 1993; Cleeremans, 1995; Cleeremans et al., 1998). The phenomenon of implicit learning seems so strong that it occurs even when the cognitive faculties are compromised. From recent studies (Grossman et al., 2003) conducted with Alzheimer patients, it appears that they are able to learn rules implicitly but not explicitly. Lastly, the rules that are learnt explicitly in a class or that are part of the inferential repertoire of experts are often not applied in everyday life or in tests based on intuition (see the experiments with statisticians of Tversky and Kahneman, 1971). At the same time, pragmatic experience and the meaning that people give to social and natural events is driven by background knowledge (Searle, 1995, 2008; Smith and Kossylin, 2007). The values, principles and categories of background knowledge, stored in memory, allow us to interpret reality, to make inferences and to act, that is, to have a pragmatic experience. Therefore background knowledge affects implicit learning and the application of cognitive rules through the pragmatic and semantic dimension of reasoning and decision-making7. What seems likely is that the relationships within schemas and among different schemas allow us to make inferences, that is, they correspond to implicit cognitive rules. For example, let us consider our schema for glass. It specifies that if an object made of glass falls onto a hard surface, the object may break. This is an example of causal inference. Similar schemas can allow you to make inductive, deductive or analogical inferences, to solve problems and to take decisions (Markman and Gentner, 2001; Ross, 1996). In conclusion, the schema theory seems to be a good candidate to explain the dependence of cognitive rules on background knowledge. If this is the case, we can expect that different cognitive rules should correspond to different background knowledge, characterizing, in this way, different cognitive styles. Nisbett (2003) has shown that the relation between background knowledge and cognitive rules supports the differences of thinking and reasoning between Americans and East Asians. These differences can explain the difficulties in reciprocal understanding and cooperation between people of different cultures. If this is the situation in industrial and academic research, we can expect obstacles to collaboration and the transfer of knowledge, and the consequent emergence of institutions and organizations dedicated to overcoming these obstacles to the capitalization of knowledge.

46


THE EFFECTS OF DIFFERENT DEONTIC KNOWLEDGE ON ACADEMY–INDUSTRY RELATIONS Usually, the obstacles to the collaboration between universities and companies are analysed by comparing entrepreneurs and managers with academic scientists (plus the academic technology transfer officers, as in the case of Siegel et al., 1999). In my opinion, this choice is correct in the case of the transfer of patents and in licensing technology, because here the relationship is between an academic scientist and an entrepreneur or manager, often mediated by an academic technology transfer officer (TTO). The situation is different in the collaboration between a university and industrial labs in order to achieve a common goal, such as the development of a prototype, the invention of a new technology, the solution to an industrial problem and so on. In these cases, interaction occurs mainly between academic and industrial researchers. Entrepreneurs, managers and TTOs might only play the role of establishing and facilitating the relationship. Since academy–industry relations are not simply reducible to patents and licences (Agrawal and Henderson, 2002), but prioritize joint research collaboration, I prefer to focus on academic and industrial researcher behaviours. Previous studies on obstacles between universities and companies analysed only superficial economic, legal and organizational aspects and focused mainly on the transfer of patents and licences (Nooteboom et al., 2007; Siegel et al., 1999). Since research collaboration implies a complex phenomenon of linguistic and cognitive coordination and adjustment among members of the research group, I think that a deeper cognitive investigation into this dimension might offer some interesting answers to the academy–industry problem. The main hypothesis is that there can be different cognitive styles in thinking, problem-solving, reasoning and decision-making that can hamper collaboration between academic and industrial researchers. These different cognitive styles are linked and mostly determined by different sets of values and norms that are part of background knowledge (as we have seen above). Different background knowledge is also responsible for poor linguistic coordination, misunderstanding and for impeding the successful psychological interaction of the group. The general hypotheses that will be inferred in the following represent a research programme of empirical tests to control the effects of cognitive styles on different scientific and technological domains and geographical contexts (for a more complete analysis see Viale, 2009).


1

47

Background Knowledge

What is the difference in background knowledge as between the university and industrial labs, and how can this influence cognitive styles? Studies in the sociology of science have focused on the values and principles that drive scientific and industrial research. Academic research seems to be governed by a set of norms and values that are close to the Mertonian ethos (Merton, 1973). Qualities such as communitarianism, scepticism, originality, disinterestedness, universalism and so on were proposed by Robert Merton as the social norms of the scientific community. He justified the proposal theoretically. Other authors, such as Mitroff (1974), criticized the Mertonian ethos on an empirical basis. He discovered that scientists often follow Mertonian norms, but that there are, nevertheless, cases in which scientists seem to follow the opposite of these norms. More recent studies (Broesterhuizen and Rip, 1984) confirm most of Merton’s norms. These studies assert that research should be Strategic, founded on Hybrid and interdisciplinary communities, able to stimulate Innovative critique, and should be Public and based on Scepticism (the acronym SHIPS). Other recent studies (Siegel et al., 1999; Viale, 2001) confirm the presence of social norms that are reminiscent of the Mertonian ethos. Scientists believe in the pursuit of knowledge per se, in the innovative role of critique, in the universal dimension of the scientific enterprise and in science as a public good. They believe in scientific method based on empirical testing, the comparison of hypotheses, enhanced problem-solving and truth as a representation of the world (Viale, 2001, pp. 216–19). But the simple fact that scientists have these beliefs doesn’t prove that they act accordingly. Beliefs can be put on hold by contingent interests and opportunistic reasons. They can also represent the invented image that scientists wish to show to society. They can vary from one discipline and specialization to another. Nevertheless, the presence of these beliefs seems to characterize the cultural identity of academic scientists. They constitute part of their background knowledge and can, therefore, influence the implicit cognitive rules for reasoning and decisionmaking. On the contrary, industrial researchers are driven by norms that are quite different from academic ones. They can be summarized by the acronym PLACE (Ziman, 1987): Proprietary, local, authoritarian, commissioned and expert. Research is commissioned by the company, which has ownership of the results, which can’t be diffused, and which are valid locally to improve the competitiveness of the company. The researchers are subjected to the authoritarian decisions of the company and develop a particular expertise valid locally. PLACE is a set of norms and values that characterizes the cultural identity of industrial researchers. These norms

48


constitute part of their background knowledge and may influence the inferential processes of reasoning and decision-making. In summary, the state of the art of studies on social norms in academic and industrial research seems insufficient and empirically obsolete. A new empirical study of norms contained in background knowledge is essential. This study should control the main features characterizing the cultural identity of academic and industrial researchers as established by previous studies. These main features can be summarized as follows. ●

●

●

●

Criticism versus dogmatism: academic researchers follow the norm of systematic critique, scepticism and falsificatory control of knowledge produced by colleagues; industrial researchers aim at maintaining knowledge that works in solving technological problems. Interest versus indifference: academic researchers are not impelled in their activity primarily by economic interest but by epistemological goals; industrial researchers are motivated mainly by economic ends such as technological competitiveness, commercial primacy and capital gain. Universalism versus localism: academic researchers believe in a universal audience of peers and in universal criteria of judgement that can establish their reputation; industrial researchers think locally in terms of both the audience and the criteria of judgement and social promotion. Communitarianism versus exclusivism: academic researchers believe in the open and public dimension of the pool of knowledge to which they must contribute in order for it to increase; industrial researchers believe in the private and proprietary features of knowledge.

To the different backgrounds I should also add the different contingent features of the contexts of decision-making (I refer here to the decisionmaking context of research managers who are heads of a research unit or of a research group) that become operational norms. The main features are related to time, results and funding. In a pure academic context,8 the time allowed for conducting research is usually loose. There are certain temporal requirements when one is working with funds coming from a public source (particularly in the case of public contracts). However, in a contract with a public agency or government department, the deadline is usually not as strict as that of a private contract, and the requested results not quite as well defined nor as specific to a particular product (e.g. a prototype or a new molecule or theorem). Thus time constraints don’t weigh as heavily on the reasoning and decision-making processes of the researchers. In contrast, when


49

an academic researcher works with an industrial contract, the time constraints are similar to those of the corporate researcher. Moreover, in a fixed given time, a precise set of results must be produced and presented to the company. According to private law, the clauses of a contract with a company can be very punitive for the researcher and the university if the signed expected requirements are not complied with. In any case, the consequences of suboptimal results from an academic working with a company are less punitive than for a corporate researcher. For the latter, the time pressure is heavier because the results, in a direct or semi-direct way, are linked to the commercial survival of the company. Suboptimal behaviour increases the risks to their career and job security. As a result, the great expectation of the fast production of positive concrete results presses on them more heavily. Different environmental pressures may generate a different adaptive psychology of time and a different adaptive ontology of what the result of the research might be. In the case of academic research, time might be less discounted. That is, future events tend not to be as underestimated as they may be in industrial research. The corporate researcher might fall into the bias of time discounting and myopia because of the overestimation of short-term results. Even the ontology of an academic researcher in respect of the final products of the research might be different from that of a corporate researcher. While the former is interested in a declarative ontology that aims at the expression of the result in linguistic form (e.g. a report, a publication, a speech and so on), the latter aims at an object ontology. The results for the latter should be linked in a direct or indirect way to the creation of an object (e.g. a new molecule, a new machine, a new material, or a new process to produce them or a patent that describes the process of producing them). The third, different, operational norm concerns financial possibilities. In this case, the problem does not concern the quantity of funding. Funding for academic research is usually lower for each unit of research (or, better, for each researcher) than for industrial research. The crucial problem is the psychological weight of the funds and how much the funds constrain and affect the reasoning and decision-making processes of the researchers. In other words (all other things being equal), this involves the amount of money at disposal and the level to which cognitive processes and, in particular, attention processes refer to a sort of value-for-money judgement in deciding how to act. It is still a topic to be investigated, but from this point of view, it seems that the psychological weight of money on academic researchers is less than that on industrial researchers. Money is perceived to have less value and, therefore, influences decision-making less. The reasons for this different mental representation and evaluation

50


may derive from: (1) the way in which funding is communicated and the ways it can constitute a decision frame (with more frequency and relevance within the company because it is linked to important decisions concerning the annual budget); (2) the symbolic representation of money (with much greater emphasis in the company, whose raison d’etre is the commercial success of its products and increased earnings); (3) the social identity of the researchers is linked more or less strongly to the monetary levels of the wage (with greater importance on the monetary level as an indicator of a successful career in a private company than in the university). The different psychological weight of money has been analysed by many authors, and in particular by Thaler (1999). To summarize, operational norms can be schematized in loose time versus pressing time; undefined results versus well-defined results; and financial lightness versus financial heaviness. How can the values in background knowledge and operational norms influence the implicit cognitive rules of reasoning and decision-making, and how can they be an obstacle to collaboration between industrial and academic researchers? Many aspects of cognition are important in research activity. We can say that every aspect is involved, from motor activity to perception, to memory, to attention, to reasoning, to decision-making and so on. My aim, however, is to focus on the cognitive obstacles to reciprocal communication, understanding, joint decision-making and coordination between academic and corporate researchers, and how these might hinder their collaboration. I shall analyse briefly three dimensions of interaction: language, group and inference (i.e. the cognitive rules in thinking, problem-solving, reasoning and decision-making). 2

Language

My focus is on the pragmatic aspects of language and communication. To collaborate on a common project means to communicate, mainly by natural language. To collaborate means to exchange information in order to coordinate one’s own actions with those of others in the pursuit of a common aim. This means ‘using language’, as the title of Clark’s book (1996) suggests, in order to reach the established common goal. Any linguistic act is at the same time an individual and a social act. It is individual because it is the individual who by motor and cognitive activity articulates the sounds that correspond to words and phrases, and who receives and interprets these sounds. Or, in Goffman’s (1981) terminology on linguistic roles, it is the subject that ‘vocalizes’, ‘formulates’, and ‘means’, and it is


51

another subject that ‘attends the vocalization, identifies the utterances and understands the meaning’ (Clark, 1996, p. 21). It is social because every linguistic act of a speaker has the aim of communicating something to one or more addressees (even in the case of private settings where we talk to ourselves, since here we ourselves play the role of an addressee). In order to achieve this goal, there should be coordination between the speaker’s meaning and the addressee’s understanding of the communication. However, meaning and understanding are based on shared knowledge, beliefs and suppositions, namely, on shared background knowledge. Therefore the first important point is that it is impossible for two or more actors in a conversation to coordinate meaning and understanding without reference to common background knowledge. ‘A common background is the foundation for all joint actions, and that makes it essential to the creation of the speaker’s meaning and addressee’s understanding as well’ (Clark, 1996, p. 14). A common background is shared by the members of the same cultural community. A second important point is that the coordination between meaning and understanding is more effective when the same physical environment is shared (e.g. the same room at a university or the same bench in a park) and the vehicle of communication is the richest possible. The environment represents a communicative frame that can influence meaning and understanding. Moreover, gestures and facial expressions are rich in nonlinguistic information and are also very important aids in coordination. From this point of view, face-to-face conversation is considered the basic and most powerful setting for communication. The third point is that the more simple and direct the coordination, the more effective the communication. There are different ways of complicating communication. The roles of speaking and listening (see above regarding linguistic roles) can be decoupled. The use of spokesmen, ghost writers and translators is an example of decoupling. A spokeswoman for a minister is only a vocalizer, while the formulation vocalized is the ghost writer’s, and the meaning is the minister’s. Obviously, in this case, the coordination of meaning and understanding becomes more difficult (and even more so because it is an institutional setting with many addressees). The non-verbal communication of the spokeswoman might be inconsistent with the meaning of the minister, and the ghost writer might not be able to formulate this meaning correctly. Moreover, in many types of discourse, such as plays, story telling, media news and reading, there is more than one layer of action. The first layer is that of the real conversation. The second layer concerns the hypothetical domain that is created by the speaker (when he is describing an event). Through recursion there can be further layers as well. For example, a play requires three layers: the first

52


is the real-world interaction among the actors, the second is the fictional role of the actors; and the third is the communication with the audience. In face-to-face conversation there is only one layer and no decoupling. The roles of vocalizing, formulating and producing meaning are performed by the same person. The domain of action identifies itself with the conversation; coordination is direct without intermediaries. Thus face-toface conversation is the most effective way of coordinating meaning and understanding, resulting in only minor distortions of meaning and fewer misunderstandings. Academic and industrial researchers are members of different cultural communities and, therefore, have different background knowledge. In the collaboration between academic and industrial researchers, coordination between meanings and understandings can be difficult if background knowledge is different. When this is the case, as we have seen before, the result of the various linguistic settings will probably be the distortion of meaning and an increase in misunderstanding. When fundamental values are different (as in SHIPS versus PLACE), and also when the operational norms of loose time versus pressing time, undefined product versus well-defined product and financial lightness versus financial heaviness are different, it is impossible to transfer knowledge without losing or distorting shares of meaning. Moreover, difficulty in coordination will increase in settings that utilize intermediaries between the academic inventor and the potential industrial user (‘mediated settings’ in Clark, 1996, p. 5). These are cases in which an intermediate technology transfer agent tries to transfer knowledge from the university to corporate labs. In this case, there is a decoupling of speech. The academic researcher is the one who formulates and gives meaning to the linguistic message (also in a written format), while the technology transfer (TT) agent is merely a vocalizer. As a result, there may be frequent distortion of the original meaning, in particular when the knowledge contains a large share of tacit knowledge. This distortion is strengthened by the likely difference in background knowledge between the TT agent and that of the other two actors in the transfer. TT agents are members of a different cultural community (if they are professional, from a private TT company) or come from different sub-communities inside the university (if they are members of a TT office). Usually, they are neither active academic researchers nor corporate researchers. Finally, the transfer of technology can also be accompanied by the complexity of having more than one domain of action. For example, if the relation between an academic and an industrial researcher is not face-to-face, but is instead mediated, there is an emergent second layer of discourse. This is the layer of the story told by the intermediary about the original process and the techniques needed to generate the technology invented by the academic


53

researchers. The story can also be communicated with the help of a written setting, for example, a patent or publication. All three points show that common background knowledge is essential for reciprocal understanding and that face-to-face communication is a prerequisite for minimizing the distortion of meaning and the misunderstandings that can undermine the effectiveness of knowledge transfer. 3

Group

The second dimension of analysis is that of the group. When two or more persons collaborate to solve a common problem, they elicit interesting emergent phenomena. In theory, a group can be a powerful problemsolver (Hinsz et al., 1997). But in order to be so, members of the group must share information, models, values and cognitive processes (ibid.). It is likely that heterogeneity of skill and knowledge is very useful for detecting solutions more easily. Some authors have analysed the role of heterogeneity in cognitive tasks (e.g. the solution of a mathematical problem) and the generation of ideas (e.g. the production of a new logo), and have found a positive correlation between it and the successful completion of these tasks (Jackson, 1992). In theory, this result seems very likely, since finding a solution entails looking at the problem from different points of view. Different perspectives allow the phenomenon of entrenched mental set to be overcome; that is, the fixation on a strategy that normally works well in solving many problems but that does not work well in solving this particular problem (Sternberg, 2009). However, the type of diversity that works concerns primarily cognitive skills or personality traits (Jackson, 1992). In contrast, when diversity is based on values, social categories and professional identity, it can hinder the problem-solving ability of the group. This type of heterogeneity generates the categorization of differences and similarities between the self and others, and results in the emergent phenomenon of the conflict/distance between ‘ingroup’ and ‘outgroup’ (Van Knippenberg and Schippers, 2007). The relational conflict/distance of ingroup versus outgroup is the most social expression of the negative impact of diversity of background knowledge on group problem-solving. As was demonstrated by Manz and Neck (1995), without a common background knowledge, there can be no sharing of goals, of the social meaning of the work, of the criteria to assess and to correct the ongoing activity, of foresight on the results nor on the impact of the results and so on. As described by the theory of ‘teamthink’ (Manz and Neck, 1995), the establishment of an effective group in problemsolving relies on the common sharing of values, beliefs, expectations and, a priori, on the physical and social world. For example, academic and

54


industrial researchers present different approaches concerning disciplinary identity. Academics have a strong faith in the ‘disciplinary matrix’ (Kuhn, 1962) composed of the members of a discipline with their particular set of disciplinary knowledge and methods. In contrast, industrial researchers tend to be opportunistic both in using knowledge and in choosing peers. They don’t feel the need to be a member of an invisible disciplinary college of peers and instead choose à la carte which peers are helpful to them and what knowledge is useful to attain the goal of the research. This asymmetry between academic and corporate researchers is an obstacle to the proper functioning of ‘teamthink’. The epistemological and social referents are different, and communication here can resemble a dialogue between deaf and mutes. Lastly, there is the linguistic dimension. As we have seen above, without common background knowledge, the coordination of meaning and understanding among the members of the group (i.e. the fundamental basis of collaboration) is impossible. Moreover, without common background knowledge, the pragmatic context of communication (Grice, 1989; Sperber and Wilson, 1986) doesn’t allow the generation of correct automatic and non-automatic inferences between speaker and addressee. Foe example the addressee will not be able to generate proper ‘implicatures’ (Grice, 1989) to fill in the lack of information and the elliptical features of the discourse. 4

Cognitive Rules

Finally, different background knowledge influences inference, that is, the cognitive rules in thinking, problem-solving, reasoning and decisionmaking activity. Different implicit cognitive rules mean asymmetry, asynchrony and dissonance in the cognitive coordination among the members of the research group. This generates obstacles to the transfer of knowledge, to the application of academic expertise and knowledge to the industrial goal, and to the development of an initial prototype or technological idea towards a commercial end. I shall discuss this subject only briefly; it is fully developed in Viale (2009). There are two systems of thinking that affect the way we reason, decide and solve problems. The first is the associative system, which involves mental operations based on observed similarities and temporal contiguities (Sloman, 1996). It can lead to speedy responses that are highly sensitive to patterns and to general tendency. This system corresponds to system 1 of Kahneman (2003), and represents the intuitive dimension of thinking. The second is the rule-based system, which involves manipulations based on the relations among symbols (Sloman, 1996). This usually requires the use of deliberate, slow procedures to reach conclusions. Through this system,


55

we carefully analyse relevant features of the available data based on rules stored in memory. This corresponds to system 2 of Kahneman (2003). The intuitive and analytical systems can produce different results in reasoning and decision-making. The intuitive system is responsible for the biases and errors of everyday-life reasoning, whereas the analytical system allows us to reason according to the canons of rationality. The prevalence of one of these two systems in the cognitive activity of academic and industrial researchers will depend on contingent factors, such as the need to finish the work quickly, and on the diverse styles of thinking. I hypothesize that the operational norms of pressing time and well-defined results and the social norms of dogmatism and localism will support a propensity towards the activity of the intuitive system. In contrast, the operational norms of loose time and undefined results, and the social norms of criticism and universalism will support the activity of the analytical system. The role of time in the activation of the two systems is evident. Industrial researchers are used to following time limits and to giving value to time. Thus this operational norm influences the speed of reasoning and decision-making and the activation of the intuitive system. The contrary happens in academic labs. The operational norm concerning results seems less evident. Those operating without the constraints of well-defined results have the ability to indulge in slow and attentive ways of analysing the features of the variables and in applying rule-based reasoning. Those who must produce an accomplished work can’t stop to analyse the details and must proceed quickly to the final results. The social norm of criticism is more evident. The tendency to check, monitor and to criticize results produced by other scientists strengthens the analytical attitude in reasoning. Any form of control requires a slow and precise analysis of the logical coherence, methodological fitness and empirical foundations of a study. On the contrary, in corporate labs the aim is to use highquality knowledge for practical results and not to increase the knowledge pool by overcoming previous hypotheses through control and critique. Finally, the social norm of universalism versus localism is less evident. Scientists believe in a universal dimension to their activity. The rules of the scientific community should be clear and comprehensible to their peers. Furthermore, the scientific method, reasoning style and methodological techniques can’t be understood and followed by only a small and local subset of scientists, but must be explicit to all in order to allow the diffusion of their work to the entire community. Thus universality tends to strengthen the analytical system of mind. The contrary happens where there is no need for the explicitness of rules and where evaluation is made locally by peers according to the success of the final product. The other dimension concerns problem-solving. At the end of the 1950s,

56


Herbert Simon and his colleagues analysed the effect of professional knowledge on problem representation. They discovered the phenomenon of ‘selective perception’ (Dearborn and Simon, 1958), that is, the relation between different professional roles and different problem representations. In the case of industrial and academic scientists, I presume that selective perception will be effective not only in relation to professional and disciplinary roles but also as regards social values and operational norms. These norms and values might characterize the problem representation and, therefore, might influence reasoning and decision-making. For example, in representing the problem of the failure of a research programme, industrial researchers might point more to variables like cost and time, whereas academic scientists might point more towards an insufficiently critical attitude and too local an approach. In any case, it might prove interesting to analyse the time spent by academic and industrial researchers in problem representation. The hypothesis is that time pressure together with an intuitive system of thinking, might cause the industrial researchers to dedicate less time to problem representation than academic researchers. Time pressure can affect the entire problem-solving cycle, which includes problem identification, definition of a problem, constructing a strategy for problem-solving, organizing information about a problem, allocation of resources, monitoring problem-solving, and evaluating problem-solving (Sternberg, 2009). In particular, it might be interesting to analyse the effect of pressing versus loose time on the monitoring and evaluation phases of the cycle. An increase in time pressures could diminish the time devoted to these phases. Dogmatism can accelerate the time spent on monitoring and evaluation, whereas criticism might lead to better and deeper monitoring and evaluation of the problem solution. Finally, time pressure might also have an effect on incubation. In order to allow old associations resulting from negative transfer to weaken, one needs to put a problem aside for a while without consciously thinking about it. One allows for the possibility that the problem will be processed subconsciously in order to find a solution. There are several possible mechanisms for enhancing the beneficial effects of incubation (Sternberg, 2009). Incubation needs time. Thus the pressing-time norm of industrial research may hinder success in problem-solving. The third dimension concerns reasoning. Reasoning is the process of drawing conclusions from principles and from evidence. In reasoning, we move from what is already known to infer a new conclusion or to evaluate a proposed conclusion. There are many features of reasoning that differentiate academic from corporate scientists. Probabilistic reasoning is aimed at updating an hypothesis according


57

to new empirical evidence. Kahneman and Tversky (1973) and Tversky and Kahneman (1980, 1982a, 1982b) have proven experimentally that we forget prior probability and give excessive weight to new data. According to Bar Hillel (1980), we give more weight to new data because we consider them more relevant compared to old data. Relevance in this case might mean that more affective or emotional weight is given to the data and, consequently, stronger attentional processes focused on them. An opposite conservative phenomenon happens when old data are more relevant. In this case we tend to ignore new data. In the case of industrial researchers, it can be hypothesized that time pressure, financial weight and welldefined results tend to give more relevance to new data. New experiments are costly and should be an important step towards the conclusion of the work. They are, therefore, more relevant and privileged by the mechanisms of attention. In contrast, academic scientists, without the pressures of cost, time and the swift conclusion of the project, can have a more balanced perception of the relevance of both old and new data. In deductive reasoning and, in particular, hypothesis-testing, Wason (1960) and Johnson-Laird (1983) proved that, in formal tests, people tend to commit confirmation bias. New studies analysing the emotional and affective dimension of hypothesis-testing have found that when an individual is emotionally involved in a thesis, he/she will tend to commit confirmation bias. The involvement can be varied: economic (when one has invested money in developing an idea), social (when social position is linked to the success of a project), organizational (when a leader holding a thesis is always right) or biographical (when one has spent many years of one’s life in developing a theory). The emotional content of a theory causes a sort of regret phenomenon that can push the individual to avoid falsification of the theory. From this point of view, it is likely that financial heaviness and dogmatism, together with other social and organizational factors, will induce industrial researchers to commit confirmation bias more easily. Research is costly but fundamental to the commercial survival of a company. Therefore researchers’ work should be successful and the results well defined in order for them to retain or to improve their position. Moreover, industrial researchers don’t follow the academic norm of criticism that prescribes the falsificationist approach towards scientific knowledge. This is contrary to the situation of academic scientists, who tend to be critics, and who are not (and should not be) obliged to be successful in their research. It is, therefore, likely that they are less prone to confirmation bias. Causal reasoning, according to Mackie (1974), is based on a ‘causal field’, that is, the set of relevant variables able to cause an effect. It is well known that each individual expert presented with the same event will

58


support a particular causal explanation. Often, once the expert has identified one of the suspected causes of a phenomenon, he/she stops searching for additional alternative causes. This phenomenon is called ‘discounting error’. From this point of view, the hypothesis posits that the different operational norms and social values of academic and corporate research may produce different discounting errors. Financial heaviness, pressing time and well-defined results compared to financial lightness, slow time and ill-defined results may limit different causal fields in the entire project. For example, the corporate scientist can consider time as a crucial causal variable for the success of the project, whereas the academic researcher is unconcerned with it. At the same time, the academic researcher can consider the value of universal scientific excellence of the results to be crucial, whereas the industrial researcher is unconcerned with it. The fourth dimension deals with decision-making. Decision-making involves evaluating opportunities and selecting one choice over another. There are many effects and biases connected to decision-making. I shall focus on certain aspects of decision-making that can differentiate academic from industrial researchers. The first deals with risk. According to ‘prospect theory’ (Kahneman and Tversky, 1979; Tversky and Kahneman, 1992), risk propensity is stronger in situations of loss and weaker in situations of gain. A loss of $5 causes a negative utility bigger than the positive utility caused by the gain of $5. Therefore people react to a loss with risky choices aimed at recovering the loss. Two other conditions that increase risk propensity are overconfidence (Fischhoff et al., 1977; Kahneman and Tversky, 1996) and illusion of control (Langer, 1973). People often tend to overestimate the accuracy of their judgements and the probability of the success of their performance. Both the perception of loss and overconfidence occur when there is competition; decisions are charged with economic meaning and have economic effects. The operational norm of financial heaviness and pressing time, and the social value of exclusivity and the interests of the industrial researcher can increase the economic value of choices and intensify the perception of competitiveness. This, consequently, can increase risk propensity. In contrast, the social values of communitarianism and indifference, and the operational norms of financial lightness and the slow time of academic scientists may create an environment that doesn’t induce a perception of loss or overconfidence. Thus behaviour tends to be more risk-averse. A second feature of decision-making is connected to regret and loss aversion. We saw before that, according to prospect theory, an individual doesn’t like to lose, and reacts with increased risk propensity. Loss aversion is based on the regret that loss produces in the individual.


59

This regret is responsible for many effects. One of the most important is ‘irrational escalation’ (Stanovich, 1999) in all kinds of investments (not only economic, but also political and affective). When one is involved in the investment of money in order to reach a goal, such as the building of a new missile prototype or the creation of a new molecule to cure AIDS, one has to consider the possibility of failure. One should monitor the various steps of the programme and, especially when funding ends, one must coldly analyse the project’s chances for success. In this case, one must consider the monies invested in the project as sunk cost, forget them and proceed rationally. People tend, however, to become affectively attached to their project (Nozick, 1990; Stanovich, 1999). They feel strong regret in admitting failure and the loss of money, and tend to continue investment in an irrational escalation of wasteful spending in an attempt to attain the established goal. This psychological mechanism is also linked to prospect theory and risk propensity under conditions of loss. Irrational escalation is stronger when there is a stronger emphasis on the economic importance of the project. This is the typical situation of a private company, which links the success of its technological projects to its commercial survival. Industrial researchers have the perception that their job and the possibility of promotion are linked to the success of their technological projects. Therefore they are likely to succumb more easily to irrational escalation than academic researchers, who have the operational norm of financial lightness and the social norm of indifference, and whose career is only loosely linked to the success of research projects. The third aspect of decision-making has to do with an irrational bias called ‘myopia’ (Elster, 1979) or temporal discounting. People tend to strongly devalue long-term gains over time. They prefer small, immediate gains to big gains projected in the future. Usually, this behaviour is associated with overconfidence and the illusion of control. Those who discount time prefer the present, because they imagine themselves able to control output and results beyond any chance estimations. In the case of industrial researchers, and of entrepreneurial culture in general, the need to have results at once, to find fast solutions to problems and to assure shareholders and the market that the company is stable and growing seems to align with the propensity towards time discounting. Future results don’t matter. What it is important is the ‘now’ and the ability to have new competitive products in order to survive commercially. Financial heaviness, pressing time and well-defined results may be responsible for the tendency to give more weight to the attainment of fast and complete results at once, even at the risk of making products that in the future will be defective, obsolete and easily overcome by competing products. In the case of academic scientists, temporal discounting might be less strong. In fact, the three

60


operational norms of financial lightness, loose time and undefined results, together with criticism and universalism, may help immunize them from myopic behaviours. Criticism is important because it pushes scientists to be not easily satisfied with quick and unripe results that can be easily falsified by their peers. Universalism is important because academics want to pursue results that are not valid locally but that can be recognized and accepted by the entire community and that can increase their scientific reputation. In the academic community, it is well known that reputation is built through a lengthy process and can be destroyed very quickly.

‘NUDGE’9 SUGGESTIONS TO THE TRIPLE HELIX: THE JANUS SCIENTIST AND PROXIMITY The capitalization of knowledge is usually analysed by recourse to external socioeconomic factors. An example is the way in which the model of the triple helix is proposed. The main determinants of the interaction between university, industry and government in supporting innovation and of the emergence of hybrid organizations, entrepreneurial universities, dual academic careers and so forth (Etzkowitz, 2008) are economic (mainly industrial competitiveness and academic fundraising) and political (mainly regional primacy). Economic and political forces are able to shape organizations and to change institutional norms. In contrast, the thesis of this chapter is that we can’t explain and predict the organizational and institutional development of the capitalization of knowledge without considering the internal dynamics driven by the epistemological and cognitive features of knowledge. Various authors have pinpointed the importance of the features of knowledge and cognition in shaping organizations. For Jim March and Herbert Simon (1993), ‘bounded rationality’ is the conceptual key to understanding the emergence of the organization, the division of labour and of routines. When the human mind cannot process the amount of information that it faces in complex problem-solving, it needs to share this burden with other minds. Different complementary roles in problem-solving emerge. These roles include a great amount of routine, that is, reasoning and decision-making realized in automatic or quasi-automatic ways. Moreover, according to Douglass North (2005) an organization is characterized by the structure of institutional values and norms. The norms and values, or in other words background knowledge, are responsible for shaping the organization and for pushing the actors to act and interact in particular ways. If we follow those authors who wish to explain, predict and also intervene in the organization, we should consider, primarily, variables such


61

as complexity of information, limited cognition and the background knowledge of the actors. It is pointless to try to design organizations and social actions through top-down economic and political planning without considering the microstructure of motivations, norms, cognitive resources and knowledge. ‘Nudge’ (Thaler and Sunstein, 2008) is a thesis that starts from these simple observations. When a policy-maker, a social planner and an economic planner want to reach certain collective goals, they must single out the right institutional tools capable of nudging the individual actors to behave coherently according to the planned aim. In order to nudge the actors effectively, one must be able to consider their cognitive limitations and motivations, and the environmental complexity in which they are placed. If a policy-maker wants to devise successful policy recipes, he/she should reason as a cognitive rule ergonomist; that is, he/she should ‘extract’ the rules from the knowledge of the minds of the actors interacting within a given initial environment. In this chapter, I have analysed the effects of the epistemological and cognitive features of knowledge on the capitalization of knowledge. In particular, I have hypothesized that some intrinsic features of knowledge can have effects on how knowledge can be generated, transferred and developed in order to achieve commercial aims. These effects, in turn, constrain the organizational and institutional forms aimed at capitalizing knowledge. The following is a summary of the main features of the knowledge relevant for capitalization. 1

Generality versus Singularity

When knowledge is composed of descriptive assertions (i.e. elementary propositions or base assertions), it refers to singular empirical events without any claim of generality. As was the case with the descriptive assertions of eighteenth-century inventors, the predicative field was limited to the empirical experience of inventors themselves. The reason, however, is not only epistemological but cognitive as well. In fact, the conceptual categorization of an empirical event changes with the experience. Thus we have a different mental representation of the same object at different times. In any case, descriptive assertions have no explanatory power and can’t allow the enlargement of the field of innovation. The effect of singularity on knowledge was a capitalization that was weakened by the law of diminishing returns. Exploitation was rapid, and only slow and small incremental innovations were generated from the original invention. Research was conducted mainly by individuals, outside the university, with the participation of apprentices. The short-range nature of the work and other institutional and economic factors (Rosenberg and Birdzell,

62


1986; Mowery and Rosenberg, 1989; Mokyr, 2002a, 2002b) pushed industrial companies to try to widen the scientific base of inventions in order to increase generality in knowledge. As we saw in the Appert versus Pasteur case, a theory explaining the causal mechanisms of food preservation allowed the improvement of the same innovation and, moreover, its application outside the original innovative field. The effect of general explanatory knowledge was the development of a capitalization that overcomes the constraints of the law of diminishing returns. Research needs to be conducted in laboratories, inside or in collaboration with a university, concurrent with the birth and development of new applied specializations (i.e. applications of a general theory to find solutions to practical problems and to invent commercial products). Moreover, general theories often apply across different natural and disciplinary domains. For example, DNA theory applies to agriculture, the environment, human health, animal health and so on. Information theory and technology can also be applied in many different areas, from genomics to industrial robotics. The generality of the application of a theory requires interdisciplinary training and research organizations that are able to single out promising areas of innovation and to generate the proper corresponding technological knowledge. This implies an interdisciplinary division of labour that can be afforded only by research universities and by the largest of companies. 2

Complexity versus Simplicity

Analytically, simple knowledge is categorized by a syntactic structure composed of few assertions with few terms whose meaning is conceptually evident (because they are empirically and directly linked to external objects that have a well-defined categorization). A descriptive assertion, such as ‘this crow is black’, or an empirical generalization, such as ‘all crows are black’, is an example of simple knowledge. These analytical features correspond to cognitive ones. The semantic network representing this knowledge is composed of a few interrelated nodes. Complexity, on the other hand, is analytically represented by a great number of assertions, containing many terms whose meaning is conceptually obscure (as, e.g., when there are theoretical terms that have indirect empirical meanings that derive from long linguistic chains, such as in the case of quarks or black holes). Quantum mechanics and string theory are examples of complex knowledge, mainly from the point of view of the meaning of the terms. Linnaeus’s natural taxonomy and Mendeleev’s periodic table of elements are examples mainly from the point of view of the number of assertions they contain. Analytical complexity implies computational overloading. The cognitive representation of a theory or of several theories


63

might correspond to an intricate semantic network with many small, interrelated and distant nodes. For an individual mind, it is usually impossible to have a complete representation of a complex theory, let alone several theories. The cognitive network will represent the conceptual structure of the theory only partially. Usually, some mental model of a theory will play the role of a heuristic device in reasoning and problem-solving. The model serves as a pictorial analogy of the theory and, therefore, does not ensure the completeness or consistency of the problem-solving results. It is evident from what I have previously stated that knowledge simplicity doesn’t require organization in knowledge generation. An individual mind can represent, process and compute knowledge and its consequences. The more complex knowledge becomes, the greater organizational division of labour is needed to cope with it. Only a network of interacting minds can have a complete representation, process the relevant information and compute the deductive consequences of complex knowledge. An organization should be shaped to give room to theoretical scientists working on concepts, to experimental scientists working on bridge laws between theoretical concepts and natural phenomena, and to applied scientists working on technological applications. Statisticians, mathematicians and experimental technicians will also play an important role. ‘Big Science’ projects such as the Los Alamos nuclear bomb, human genome mapping, nuclear fusion and particle physics are examples of this collective problem-solving. When complexity is also linked to generality, the division of labour will be reflected in the interdisciplinarity of the experts involved in collective problem-solving. Most companies will not be endowed with this level of expertise and will, therefore, always rely more on academic support in applying knowledge to commercial aims. Consequently, increasing complexity and generality means a growing ‘industrial’ role for universities. The Obama programme for green technologies might be a future example of the generation and capitalization of complex and general knowledge that will see universities playing a central ‘industrial’ role. 3

Explicitness versus Tacitness

To capitalize knowledge, one should be able to sell it or use it to produce economic added value. In both cases, knowledge must be completely understandable and reproducible by both the inventor and by others. In the latter case, knowledge must not lose part of its meaning in transfer. When knowledge was mainly composed of descriptive assertions and technical norms, it was highly idiosyncratic. Descriptive assertions corresponded to the perceptual and conceptual apparatus of the inventor. Technical norms were represented by competential know-how. Thus knowledge was

64


mainly tacit, and its transfer through linguistic media almost impossible. The organizational centre of capitalization was the inventor’s laboratory, where he/she attempted to transfer knowledge to apprentices through faceto-face teaching and by doing and interacting. Selling patents was pointless without ‘transfer by head’ or proper apprenticeship. According to some authors with the growth of science-based innovation the situation changed substantially. In life sciences, for example, ontic knowledge is composed of explanatory assertions, mainly theories and models. Technical norms are less represented by competential know-how than by explicit condition– action rules. Thus the degree of tacitness seems, at first sight, to be less. Ontic knowledge explaining an invention might be represented, explicitly, by general theories and models, and the process for reproducing the invention would be little characterized by know-how. A patent might be sold because it would allow complete knowledge transfer. Academic labs and companies might interact at a distance, and there would be no need for university–industry proximity. The explicitness of technological knowledge would soon become complete with the ICT revolution (Cowan et al., 2000), that would be able even to automatize know-how. As I have shown in previous articles (Balconi et al., 2007; Pozzali and Viale, 2007), this optimistic representation of the disappearance of tacit knowledge is an error. It considers tacitness only at the level of competential know-how and does not account for the other two aspects of tacitness, namely, background knowledge and cognitive rules. Background knowledge not only includes social norms and values but also principles and categories that give meaning to actions and events. Cognitive rules serve to apply reason to the data and to find solutions to problems. Both tend to be individually variable. The knowledge represented in a patent is, obviously, elliptical from this point of view. A patent can’t explicitly contain background knowledge and cognitive rules used to reason and interpret information contained in it. These irreducible tacit aspects of knowledge oblige technology generators and users to interact directly in order to stimulate a convergent calibration of the conceptual and cognitive tools needed to reason and interpret knowledge. This entails a stimulus towards proximity between university and company and the creation of hybrid organizations between them to jointly develop knowledge towards commercial aims. 4

Shared Background Knowledge versus Unshared Background Knowledge

Norms and values used for action, together with principles and concepts used for understanding, constitute background knowledge. Beyond knowledge transfer, shared background knowledge is necessary for linguistic


65

communication and for effective collaboration in a group. The linguistic dimension has never been analysed in knowledge capitalization. Its importance is evident both in patent transfer and in research collaboration. As I stated above, academic and industrial researchers are members of different cultural communities and, therefore, have different background knowledge. In the collaboration between academic and industrial researchers, the coordination between meanings and understandings can be difficult if background knowledge is different. When this is the case, as I have seen above, the effect on the various linguistic settings will probably be the distortion of meaning and the creation of misunderstandings. Moreover, difficulties in coordination will increase in settings that utilize intermediaries (such as the members of the TT office of a university or of the TT agent of a private company or government) between the academic inventor and the potential industrial user (‘mediated settings’ in Clark, 1996, p. 5). In this case, there is decoupling of speaking. The academic researcher formulates and gives meaning to the linguistic message (also in a written setting), while the TT agent is merely a vocalizer of the message. Thus there may be frequent distortion of the original meaning, in particular when the knowledge contains a great share of tacit knowledge. This distortion is strengthened by the likely differences in background knowledge between the TT agent and the other actors in the transfer. Finally, in the transfer of technology, the complexity of having more than one domain of action can also arise. For example, if the relation between an academic and industrial researcher is not face-to-face but is instead mediated by an intermediary, there is an emergent second layer of discourse. This is the layer of the story that is told by the intermediary about the original process and techniques used to generate the technology invented by the academic researchers. All three points show that common background knowledge is essential for reciprocal understanding and that face-to-face communication is a prerequisite for minimizing the distortion of meaning and the misunderstandings that can undermine effective knowledge transfer. Organizations of knowledge capitalization must, therefore, emphasize the feature of proximity between knowledge producers and users, and support the creation of public spaces for meeting and cultural exchange between members of universities and companies. Moreover, universities primarily, but companies also, should promote the emergence of a new professional figure, a researcher capable of cultural mediation between academic and industrial background knowledge. 5

Shared Cognitive Style versus Unshared Cognitive Style

Analytically, cognitive rules for inference are part of deontic knowledge. Cognitively, they can be considered the emergent results of pragmatic

66


knowledge. In any case, they are influenced by norms and values contained in background knowledge, as was shown by Nisbett (2003) in his study on American and East Asian ways of thinking. The hypothesis of different cognitive rules generated by different background knowledge seems likely but must still be confirmed empirically (Viale et al., forthcoming). I shall now look at some examples of these differences (analysed in the pilot study of Fondazione Rosselli, 2008). Time perception and the operational norm of the loose time versus pressing time differentiate business-oriented academics from entrepreneurial researchers. For the latter, time is pressing, and it is important to find concrete results quickly and not waste money. Their responses show a clear temporal discounting. The business participants charge academics with looking too far ahead and not caring enough about the practical needs of the present. The short-term logic of the industrial researchers seems to follow the Latin saying Primum vivere deinde philosophare (‘First live, then philosophize’). For them, it is better to concentrate their efforts on the application of existing models in order to obtain certain results. The academic has the opposite impetus, that is, to explore boundaries and uncertain knowledge. The different temporal perceptions are linked to risk assessment. The need to obtain fast results for the survival of the company increases the risk perception of the money spent on R&D projects. In contrast, even if the academic participants are not pure but business oriented, they don’t exhibit the temporal discounting phenomenon, and for them risk is perceived in connection with scientific reputation inside the academic community (the social norm of universalism). What is risky to the academic researchers is the possibility of failing to gain scientific recognition (vestiges of academic values). Academic researchers also are more inclined towards communitarianism than exclusivity (vestiges of academic values). They believe that knowledge should be open and public and not used as exclusive private property to be monopolized. For all participants, misunderstandings concerning time and risk are the main obstacles to collaboration. University members accuse company members of being too short-sighted and overly prudent in the development of new ideas; entrepreneurial participants charge university members with being too high-minded and overly far-sighted in innovation proposals. This creates organizational dissonance in planning the milestones of the projects and in setting the amount of time needed for the various aspects of research. Differences in cognitive rules are a strong factor in creating dissonance among researchers. The likely solution to this dissonance is the emergence in universities of a new research figure trained in close contact with industrial labs. This person should have the academic skills of his/her pure scientist colleagues and, at the same time, knowledge of industrial cognitive styles and values. Obviously, hybrid organizations


67

can also play an important role, acting as a type of ‘gym’ in which to train towards the convergence between cognitive styles and values. In conclusion, when the main hypotheses of this chapter will be empirically controlled we shall know what the main epistemological and cognitive determinants in capitalizing knowledge are. This will give us clues on how to ‘nudge’ (Thaler and Sunstein, 2008) the main stakeholders in academy–industry relations to act to improve collaboration and knowledge transfer. In my opinion, ‘nudging’ should be the main strategy of public institutions and policy-makers wishing to strengthen the triple-helix model of innovation. For example, if results confirm the link between social values, operational norms and cognitive style, it might be difficult to overcome the distances between pure academic scientists and entrepreneurial researchers. What would be more reasonable would be to support the emergence of a new kind of researcher. Together with the pure academic researcher and the applied researcher, universities must promote a mestizo, a hybrid figure that, like a two-faced Janus (Viale, 2009), is able to activate mentally two inconsistent sets of values and operational norms, that is, the academic and the entrepreneurial. These Janus scientists would not believe the norms, but would accept them as if they believed them (Cohen, 1992). They would act as the cultural mediators and translators between the two worlds. They should be members of the same department as the pure and applied scientists, and should collaborate with them as well as with the industrial scientists. A reciprocal figure such as this is difficult to introduce into a company unless it is very large and financially well endowed. In fact, commercial competitiveness is the main condition for company survival. Time and risk are leading factors for competitiveness. And cognitive style is strongly linked to them. This creates a certain rigidity when faced with the possibility of change. Thus adaptation should favour the side of university, where there is more potential elasticity in shaping behaviour. Moreover, the two-faced Janus figure is different from that involved in a TT office. It is a figure that should collaborate directly in research activity with corporate scientists, whereas a member of a TT office has the function of establishing the bridge between academics and the company. The first allows R&D collaboration, whereas the second facilitates technology transfer. Empirical confirmation of the emergence of these figures can be found in the trajectories of the development of strongly science-based sectors such as biotechnologies, which have followed a totally different path in America than in Europe (Orsenigo, 2001). While the American system is characterized by a strong proximity between the industrial sector and the world of research, with the universities in the first line in

68


internalizing and in taking on many functions typical of the business world, in Europe universities have been much more reluctant to take on a similar propulsive role. A second ‘nudge’ suggestion that may emerge from this chapter, and in particular from the growing generality and complexity of the knowledge involved in innovation, is the importance of face-to-face interaction and proximity between universities and companies. The need for proximity has been underlined in recent studies (Arundel and Geuna, 2004; for an explanation according to the theory of complexity see Viale and Pozzali, 2010). Virtual clusters and meta districts can’t play the same role in innovation. Proximity and face-to-face interactions are not only important for minimizing the tacitness bottleneck in technology transfer, but face-to-face interaction is also fundamental to collaboration because of the linguistic and pragmatic effect on understanding (see above). It also improves the degree of trust, as has been proved by neuroeconomics (Camerer et al., 2005). Proximity can also increase the respective permeability of social values and operational norms. From this point of view, universities might promote the birth of ‘open spaces’ of discussion and comparison where academicians and business members might develop a kind of learning by interacting. Lastly, proximity allows better interaction between companies and the varied academic areas of expertise and knowledge resources. Indeed, only the university has the potentiality to cope with the growing complexity and interdisciplinarity of the new ways of generating innovation. Emergent and convergent technologies require a growing division of expert labour that includes the overall knowledge chain, from pure and basic research to development. Only a company that can interact and rely on the material and immaterial facilities of a research university can find proper commercial solutions in the age of hybrid technological innovation.

NOTES 1. Janus is a two-faced god popular in the Greek and Roman tradition. One face looks to the past (or to tradition) and the other looks toward the future (or to innovation). 2. From a formal point of view, the descriptive assertion may be expressed in the following way: (E x, t) (a S b) This means that an event x exists in a given time t such that if x is perceived by the agent a, then it has the features b. Contrary to the pure analytical interpretation, this formulation is epistemic; that is, it includes the epistemic actor a who is responsible for perceiving the feature b of event x.


69

3. In theory, this variability could be overcome by artificial epistemic agents without plasticity in conceptual categorization. Some artificial intelligence systems used at the industrial level have these features. 4. These cases may generate ‘compound techniques’ (Mokyr, 2002b, p. 35) based on new combinations of known techniques (e.g. the use of a metal container instead of a glass bottle) without knowledge of the causal mechanisms behind the phenomenon. This type of innovation is short-lived, however, and soon exhausts its economic potential. 5. The organizational impact of the cognitive features of scientific knowledge has been singled out in some studies on scientific disciplines and specializations. For example, the different organizations of experimental physicists compared to organic chemists or biologists is explained mainly by the different complexity of knowledge (Shinn, 1983). 6. Deontic logic (the name was proposed by Broad to von Wright) uses two operators: O for obligation and P for permission. The pretence of building a logic based on these two operators as prefixes to names of acts A, B, C, and so on, which is similar to propositional logic, has been strongly criticized by many, among them von Wright himself (1963). 7. It is not clear if the process is linear or circular and recursive. In this case, cognitive rules might become part of background knowledge, and this could change its role in pragmatic experience and in reasoning and decision-making processes. 8. The analysis refers mainly to the academic environment of the universities of continental Europe. 9. Nudge (Thaler and Sunstein, 2008) is the title of a book and a metaphor characterizing a way to gently push social actors towards certain collective aims according to their cognitive characteristics.

REFERENCES Agrawal, A. and R. Henderson (2002), ‘Putting patents in context: exploring knowledge transfer from MIT’, Management Science, 48 (1), 44–60. Anderson, J.R. (1983), The Architecture of Cognition, Cambridge, MA: Harvard University Press. Anderson, J.R. (1996), ‘ACT: A simple theory of complex cognition’, American Psychologist, 51, 355–65. Arundel, A. and A. Geuna (2004), ‘Proximity and the use of public science by innovative European firms’, Economic Innovation and New Technologies, 13 (6), 559–80. Balconi, M., A. Pozzali and R. Viale (2007), ‘The “codification debate” revisited: a conceptual framework to analyse the role of tacit knowledge in economics’, Industrial and Corporate Change, 16 (5), 823–49. Bar Hillel, M. (1980), ‘The base-rate fallacy in probabilistic judgements’, Acta Psychologica, 44, 211–33. Barsalou, L.W. (2000), ‘Concepts: structure’, in A.E. Kazdin (ed.), Encyclopedia of Psychology, Vol. 2, Washington, DC: American Psychological Association, pp. 245–8. Beth, E. and J. Piaget (1961), Etudes d’Epistemologie Genetique, XIV: Epistemologie Mathematique et Psichologie, Paris: PUF. Braine, M.D.S. (1978), ‘On the relation between the natural logic of reasoning and standard logic’, Psychological Review, 85, 1–21. Broesterhuizen, E. and A. Rip (1984), ‘No Place for Cudos’, EASST Newsletter, 3.

70


Camerer, C., G. Loewenstein and D. Prelec (2005), ‘Neuroeconomics: how neuroscience can inform economics’, Journal of Economic Literature, 43 (1), 9–64. Carruthers, P., S. Stich and M. Siegal (eds) (2002), The Cognitive Basis of Science, Cambridge: Cambridge University Press. Cheng, P.W. and K.J. Holyoak (1985), ‘Pragmatic versus syntactic aproaches to training deductive reasoning’, Cognitive Psychology, 17, 391–416. Cheng, P.W. and K.J. Holyoak (1989), ‘On the natural selection of reasoning theories’, Cognition, 33, 285–313. Cheng, P.W. and R. Nisbett (1993), ‘Pragmatic constraints on causal deduction’, in R. Nisbett (ed.), Rules for Reasoning, Hillsdale, NJ: Erlbaum, pp. 207–27. Clark, H.H. (1996), Using Language, Cambridge: Cambridge University Press. Cleeremans, A. (1995), ‘Implicit learning in the presence of multiple cues’, in Proceedings of the 17th Annual Conference of the Cognitive Science Society, Hillsdale, NJ: Erlbaum, pp. 298–303. Cleeremans, A., A. Destrebecqz and M. Boyer (1998), ‘Implicit learning: news from the Front’, Trends in Cognitive Science, 2, 406–16. Cohen, J. (1992), An Essay on Belief and Acceptance, Oxford: Blackwell. Collins, A.M. and M.R. Quillian (1969), ‘Retrieval time from semantic memory’, Journal of Verbal Learning and Verbal Behaviour, 8, 240–48. Cowan, R., P.A. David and D. Foray (2000), ‘The explicit economics of knowledge codification and tacitness’, Industrial and Corporate Change, 9, 211–53. Dearborn, D.C. and H.A. Simon (1958), ‘Selective perception: a note on the departmental identifications of executives’, Sociometry, 21 (June), 140–44. Elster, J. (1979), Studies in Rationality and Irrationality, Cambridge: Cambridge University Press. Etzkowitz, H. (2008). The Triple Helix: University–Industry–Government Innovation in Action, London: Routledge. Fischhoff, B., P. Slovic and S. Lichtenstein (1977), ‘Knowing with certainty: the appropriateness of extreme confidence’, Journal of Experimental Psychology, 3, 552–64. Fondazione Rosselli (2008). ‘Different cognitive styles between academic and industrial researchers: a pilot study’, www.fondazionerosselli.it. Giere, R.N. (1988), Explaining Science, Chicago, IL: Chicago University Press. Goffman, E. (1981), Forms of Talk, Philadelphia, PA: University of Pennsylvania Press. Grice, H.P. (1989), Studies in the Way of Words, Cambridge, MA: Harvard University Press. Grossman, M., E.E. Smith, P. Koenig, G. Glosser, J. Rhee, and K. Dennis (2003), ‘Categorization of object descriptions in Alzheimer’s disease and frontal temporal dementia: limitation in rule-based processing’, Cognitive Affective and Behavioural Neuroscience, 3 (2), 120–32. Hinsz, V., S. Tindale and D. Vollrath (1997), ‘The emerging conceptualization of groups as information processors’, Psychological Bulletin, 96, 43–64. Jackendoff, R. (2007), Language, Consciousness, Culture, Cambridge, MA: MIT Press. Jackson, S. (1992), ‘Team composition in organizational settings: issues in managing an increasingly diverse work force’, in S. Worchel, W. Wood and J. Simpson (eds), Group Process and Productivity, Newbury Park, CA: Sage, pp. 138–73. Johnson-Laird, P. (1983), Mental Models, Cambridge: Cambridge University Press.


71

Johnson-Laird, P. (2008), How We Reason, Oxford: Oxford University Press. Kahneman, D. (2003), ‘Maps of Bounded Rationality’, in T. Frangsmyr (ed.), Nobel Prizes 2002, Stockholm: Almquist & Wiksell, pp. 449–89. Kahneman, D. and A. Tversky (1973), ‘On the psychology of prediction’, Psychological Review, 80, 237–51. Kahneman, D. and A. Tversky (1979), ‘Prospect theory: an analysis of decision under risk’, Econometrica, 47, 263–91. Kahneman, D. and A. Tversky (1996), ‘On the reality of cognitive illusions’, Psychological Review, 103, 582–91. Kauffman, S. (1995), At Home in the Universe: The Search for the Laws of SelfOrganization and Complexity, New York: Oxford University Press. Kuhn, T. (1962), The Structure of Scientific Revolutions, Chicago, IL: University of Chicago Press. Kuznets, S. (1965), Economic Growth and Structure, New York: W.W. Norton. Langer, E. (1973), ‘Reduction of psychological stress in surgical patients’, Journal of Experimental Social Psychology, 11, 155–65. Mackie, J. (1974), The Cement of the Universe. A Study on Causation, Oxford: Oxford University Press. Manz, C.C. and C.P. Neck (1995), ‘Teamthink. Beyond the groupthink syndrome in self-managing work teams’, Journal of Managerial Psychology, 10, 7–15. March, J. and H.A. Simon (1993), Organizations, 2nd edn, Cambridge, MA: Blackwell Publishers. Markman, A.B. and D. Gentner (2001), ’Thinking’, Annual Review of Psychology, 52, 223–47. Merton, R. (1973), The Sociology of Science. Theoretical and Empirical Investigations, Chicago, IL: University of Chicago Press. Mitroff, I.I. (1974), The Subject Side of Science, Amsterdam: Elsevier. Mokyr, J. (2002a), The Gifts of Athena: Historical Origins of the Knowledge Economy (Italian translation 2004, I doni di Atena, Bologna: Il Mulino), Princeton, NJ: Princeton University Press. Mokyr, J. (2002b), ‘Innovations in an historical perspective: tales of technology and evolution’, in B. Steil, G. Victor and R. Nelson (eds), Technological Innovation and Economic Performance, Princeton, NJ: Princeton University Press, pp. 23–46. Moore, G.E. (1922), ‘The nature of moral philosophy’, in Philosophical Studies, London: Routledge & Kegan Paul. Mowery, D. and N. Rosenberg (1989), Technology and the Pursuit of Economic Growth, Cambridge: Cambridge University Press. National Science Foundation (2002), Converging Technologies for Improving Human Performance, Washington, DC: National Science Foundation. Nisbett, R.E. (2003), The Geography of Thought, New York: The Free Press. Nooteboom, B., W. Van Haverbeke, G. Duysters, V. Gilsing and A. Van den Oord (2007), ‘Optimal cognitive distance and absorptive capacity’, Research Policy, 36, 1016–34. North, D.C. (2005), Understanding the Process of Economic Change (Italian translation 2006, Capire il processo di cambiamento economico, Bologna: Il Mulino), Princeton, NJ: Princeton University Press. Nozick, R. (1990), ‘Newcomb’s problem and two principles of choice’, in P.K. Moser (ed.), Rationality in Action. Contemporary Approach, New York: Cambridge University Press, pp. 207–35.

72


Orsenigo, L. (2001), ‘The (failed) development of a biotechnology cluster: the case of Lombardy’, Small Business Economics, 17 (1–2), 77–92. Politzer, G. (1986), ‘Laws of language use and formal logic’, Journal of Psycholinguistic Research, 15, 47–92. Politzer, G. and A. Nguyen-Xuan (1992), ‘Reasoning about promises and warnings: Darwinian algorithms, mental models, relevance judgements or pragmatic schemas?’, Quarterly Journal of Experimental Psychology, 44, 401–21. Pozzali, A. and R. Viale (2007), ‘Cognition, types of “tacit knowledge” and technology transfer’, in R. Topol and B. Walliser (eds), Cognitive Economics: New Trends, Oxford: Elsevier, pp. 205–24. Reber, A.S. (1993), Implicit Learning and Tacit Knowledge. An Essay on the Cognitive Unconscious, Oxford: Oxford University Press. Rosenberg, N. and L.E. Birdzell (1986), How The West Grew Rich. The Economic Transformation of the Economic World (Italian translation 1988, Come l’Occidente è diventato ricco, Bologna: Il Mulino), New York: Basic Books. Rosenberg, N. and D. Mowery (1998), Paths of Innovation. Technological Change in 20th-Century America (Italian translation 2001, Il Secolo dell’Innovazione, Milano: EGEA), Cambridge: Cambridge University Press. Ross, B.H. (1996), ‘Category learning as problem solving’, in D.L. Medin (ed.), The Psychology of Learning and Motivation: Advances in Research and Theory, Vol 35, San Diego, CA: Academic Press, pp. 165–92. Rumain, B., J. Connell and M.D.S. Braine (1983), ‘Conversational comprehension processes are responsible for fallacies in children as well as in adults: it is not the biconditional’, Developmental Psychology, 19, 471–81. Searle, J. (1995), The Construction of Social Reality, New York: Free Press. Searle, J. (2008), Philosophy in a New Century, Cambridge: Cambridge University Press. Shinn, T. (1983), ‘Scientific disciplines and organizational specificity: the social and cognitive configuration of laboratory activities’, Sociology of Science, 4, 239–64. Siegel, D., D. Waldman and A. Link (1999), ‘Assessing the impact of organizational practices on the productivity of university technology transfer offices: an exploratory study’, NBER Working Paper 7256. Sloman, S.A. (1996), ‘The empirical case for two systems of reasoning’, Psychological Bulletin, 119, 3–22. Smith, E.E. and S.M. Kossylin (2007), Cognitive Psychology. Mind and Brain, Upper Saddle River, NJ: Prentice Hall. Sperber, D. and D. Wilson (1986), Relevance. Communication and Cognition, Oxford: Oxford University Press. Stanovich, K. (1999), Who Is Rational?: Studies of Individual Differences in Reasoning, Mahwah, NJ: Erlbaum. Sternberg, R.J. (2009), Cognitive Psychology, Belmont, CA: Wadsworth. Thaler, R. (1999), ‘Mental accounting matters’, Journal of Behavioural Decision Making, 12, 183–206. Thaler, R.H. and C.R. Sunstein (2008), Nudge. Improving Decisions About Health, Wealth, and Happiness, New Haven, CT: Yale University Press. Tversky, A. and D. Kahneman (1971), ‘Belief in the law of small numbers’, Psychological Bulletin, 76, 105–90. Tversky, A. and D. Kahneman (1980), ‘Causal schemas in judgements under uncertainty’, in M. Fishbein (ed.), Progress in Social Psychology, Hillsdale, NJ: Erlbaum, pp. 49–72.


73

Tversky, A. and D. Kahneman (1982a), ‘Judgements of and by representativeness’, in D. Kahneman, P. Slovic and A. Tversky (eds), Judgement under Uncertainty: Heuristics and Biases, Cambridge: Cambridge University Press, pp. 84–98. Tversky, A. and D. Kahneman (1982b), ‘Evidential impact of base rate’, in D. Kahneman, P. Slovic and A. Tversky (eds), Judgement under Uncertainty: Heuristics and Biases, Cambridge: Cambridge University Press, pp. 153–60. Tversky, A. and D. Kahneman (1992), ‘Advances in prospect theory: cumulative representation of uncertainty’, Journal of Risk and Uncertainty, 5, 547–67. Van Knippenberg, D. and M.C. Schippers (2007), ‘Work group diversity’, Annual Review of Psychology, 58, 515–41. Viale, R. (1991), Metodo e società nella scienza, Milano: Franco Angeli. Viale, R. (2001), ‘Reasons and reasoning: what comes first’, in R. Boudon, P. Demeulenaere and R. Viale, R. (eds), L’explication des normes sociales, Paris: PUF, pp. 215–36. Viale R. (2008), ‘Origini storiche dell’innovazione permanente’, in R. Viale (a cura di), La cultura dell’innovazione, Milano: Editrice Il Sole 24 Ore, pp. 19–70. Viale, R. (2009), ‘Different cognitive styles between academic and industrial researchers: an agenda of empirical research’, New York: Columbia University, http://www.italianacademy.columbia.edu/publications_working.html#0809. Viale, R. and A. Pozzali (2010), ‘Complex adaptive systems and the evolutionary triple helix’, Critical Sociology, 34. Viale, R., F. Del Missier, R. Rumiati and C. Franzoni (forthcoming), Different Cognitive Styles between Academic and Industrial Researchers: An Empirical Study. von Wright, G.H. (1963), Norms and Action. A Logical Enquiry, London: Routledge. Wason, P.C. (1960), ‘On the failure to eliminate hypotheses in a conceptual task’, Quarterly Journal of Experimental Psychology, 12, 129–40. Ziman, J.M. (1987), ‘L’individuo in una professione collettivizzata’, Sociologia e Ricerca Sociale, 24, 9–30.

2.

‘Only connect’: academic–business research collaborations and the formation of ecologies of innovation Paul A. David and J. Stanley Metcalfe

UNIVERSITY PATENTING FOR TECHNOLOGY TRANSFER – MIRACULOUS OR MISTAKEN MOVEMENT? The international movement to emulate the US institutional reforms of the early 1980s that gave universities and publicly funded technology research organizations the right (rather than a privilege granted by a sponsoring agency) to own and derive income from the commercialization of IP (intellectual property) based on their researchers’ inventions has developed remarkable momentum since its inception at the end of the 1990s (see, e.g., the survey in Mowery and Sampat, 2005). The process of change and adaptation that was thereby set in motion among the EU member states has not yielded the dramatic effects on innovation and employment growth in Europe that had been promised by those who enthusiastically prescribed a dose of ‘the Bayh–Dole solution’ for the region’s sluggish economies. But such expectations were at best unrealistic, and in too many instances stemmed from contemporary European observers’ mistaken suppositions regarding the sources of the revival of productivity growth and the ‘information technology’ investment boom in the American economy during the late 1990s; and more widely shared misapprehensions regarding the fundamental factors that were responsible for the rising frequency with which patents applications filed at the USPTO during the 1980s and 1990s were citing scientific papers by academic authors.1 The movement to promote ‘technology transfers’ from universities to industry through the medium of patent licensing was fueled by a widespread supposition that European academic research was dangerously disconnected from the processes of private sector innovation. This belief rested largely on the observation at the turn of the century that the 74

Academic–business research collaborations

75

regions’ universities were not extensively involved as corporate entities in filing applications for patents, and negotiating the terms on which the inventions could be commercially exploited (whether by being ‘worked’ or not) via business licensees. The obvious contrast was that drawn with the contemporary scene in the USA during the frenzied era of the dotcom and biogenetics boom, where research universities’ patenting and the licensing of technology to venture-capital-fueled startups were growing rapidly. Whatever the accuracy of European perceptions about the realities of events taking place on the far side of the Atlantic Ocean, it has become clear that there was a serious misconception of the realities of university–industry technology transfers closer to home. Several recent studies have revealed, however belatedly, that much of the university research leading to patents in Europe does not show up readily in the statistics, because private firms rather than the universities themselves apply for the patent.2 The impression that university professors in the physical sciences and engineering were not engaged in patent-worthy inventive activities whose results were of interest to industrial firms was firmly dispelled for the case of Italy by a study of the identities of inventors named in patents issued by the European Patent Office (EPO) during 1978–99. Balconi et al. (2004, Table 3) found that for many research areas the Italian academic inventors of those patents formed quite a sizeable share of all the professors working in those fields on the faculties of Italy’s universities and polytechnics at the close of that period.3 In 11 of the 20 research fields studied, 13.9 percent or more of the professors working in the field were identified as the inventors of EPO patents issued for inventions in the corresponding field; in the case of quite a few specialty areas, such as mechanical and chemical bioengineering, and industrial and materials chemistry, the corresponding proportions were much higher – ranging from one-third to one-half. The transfer of the ownership of those patents to industrial firms was the norm in Italy, as was the case elsewhere in Western Europe during this era.4 According to Crespi et al. (2006), about 80 percent of the EPO patents with at least one academic inventor are not owned by the university, which means that no statistical indication of a university’s involvement in the technology’s creation would be found by studying the patent office records. Thus the appearance of a lack of ‘university patents’ in Europe must be understood to be a lack of university-owned patents, and not necessarily indicative of any dearth of university-invented patents. Once the data are corrected to take into account the different ownership structure in Europe and in the USA, very simple calculations made by Crespi et al. (2006) indicate that the European academic system seems to perform considerably better than was formerly believed to be the case: indeed, the

76


patenting output of European universities lags behind only one among the US universities – and in that exception the difference was quite marginal. If there are grounds for suspecting that it may not really have been necessary for Europe to embrace the Bayh–Dole regime’s approach to effecting ‘technology transfers’ from academic labs to industrial firms, there also are doubts as to whether the likelihood of innovative success ensuing from such transactions is raised by having universities rather than firms own the patents on academic inventions. There are theoretical arguments about this, pro and con, because the issue turns essentially on the comparative strength of opposing effects: are firms likely to make a better job of the innovation process because they have greater control over the development of their own inventions? Or is it less likely that viable academic inventions will be shelved if the inventor’s institution retains control of the patent and has incentives to find a way of licensing it to a company that will generate royalty earnings by direct exploitation? Since the issue is one that might be settled on empirical grounds, it is fortunate that Crespi et al. (2006) have recently carried out a statistical analysis of the effects of university ownership on the rate of commercial application (diffusion) of patents, and on patents’ commercial values, based upon the experience of European academic inventions for which patents were issued by the EPO. Their analysis controls for the different (ex ante observed) characteristics of university-owned and non-universityowned patents, and therefore accords with theoretical considerations that suggest one should view university ownership of a patent as the endogenously determined outcome of a bargaining game.5 Both before and after controlling for such differences between patents, they find no statistically significant effects of university ownership of patents. The only significant (positive) effect reported is that university-owned patents are more often licensed out, but this does not lead to an overall increase in the rate of commercial use. Hence the authors conclude that they can find no evidence of ‘market failure’ that would call for additional legislation in order to make university patenting more attractive in Europe. Their inference is that whether or not universities own commercially interesting patents resulting from their research makes little difference, because whatever private economic benefits might be conveyed by ownership per se are being adjusted for by the terms set in the inter-organizational bargaining process. This interpretation of the findings surely should gratify admirers of the Coase Theorem’s assertion that the locus of ownership of valuable property does not carry efficiency implications when transactions costs are not very high. Nonetheless, even though impelled by misconceptions of the realities both in the USA and in Europe, there is now a general sense that, by


77

following the American example, governments in the EU have forced a reconsideration of the administrative system into which Europe’s universities had settled in the era following the rapid post-World War II proliferation of new institutional foundations; and that this shock has been on balance salutary in its effects for the longer term. Perhaps that is so. It certainly has encouraged fresh thinking about the potential payoffs of publicly funded research in terms of commercial innovation in small and medium-sized industries, and of the support of applied research in areas where new science might spawn new technologies of interest to major new industries. It has precipitated and legitimized the assertion of university rights to ownership of intellectual property vis-à-vis the claims of their employees – an alteration in institutional norms that had occurred almost universally in the USA before the 1970s. More significantly, perhaps, it had the effect of encouraging a general re-examination of university regulations affecting the activities of academic researchers in Europe. The liberalization – for the benefit of universities – of many rules that had been imposed uniformly on state institutions and their employees, in turn, has opened the way to a broader consideration of the need for greater institutional independence and autonomy. That appears to have brought more realistic attention from university leaders to the possibilities of adopting or creating new incentive mechanisms that would redirect individual activities and raise productivity among those who worked within those organizations.6

PRODUCTIVE SHOCKS AND LASTING PROBLEMATIC TENSIONS These have been important steps toward the flexibility needed for R&D collaborations throughout the European Research Area (ERA), even though a considerable distance remains to be traveled by the respective national government authorities along the path towards granting greater autonomy to their institutions; and also by consortia and regional coalitions of the institutions themselves to remove the impediments to collaboration and inter-university mobility of personnel that continues to fragment the European market for academic science and engineering researchers. Furthermore, although European governments have not hesitated to urge business corporations to accept the necessity of investments in ‘organizational re-engineering’ to take full advantage of new technologies and consequent new ways of working, they have not been so quick to put this good advice into practice ‘closer to home’ – when urging

78


‘modernization’ upon their respective educational and research institutions. Yet it is now more widely recognized that the ‘modernizing’ of university governance and management is not a costless process, and, like ‘business re-engineering’, requires up-front incremental expenditures to effect the transformations that are expected to yield sustainable future gains in the efficiency of resource use. There is thus an obvious tension between two key assertions about university–business interactions in many current policy recommendations, and in the programs that seek to respond to their advice. Insistence giving priority to ‘market-driven’ technology transfers – based upon the licensing or direct exploitation of intellectual property arising from university research – creates impediments to inter-organizational collaboration, and, at the very least, tends to inhibit the recommendation that universities strive to develop more frequent interpersonal collaborative contacts to encourage exchange of scientific and technological information with industry. That this tension remained unresolved is not surprising, but that it continued to pass without much comment in policy circles for so long was nonetheless unfortunate. Most welcome, therefore, are the growing signs of a shift of thinking in Europe that is evidenced in such statements as the one below, in which the view expressed by the Report of the Forum on University-based Research (European Commission, 2005, p. 28) is in harmony with that in the 2005 report by the Forum on University-based Research: ‘From a societal perspective, more will be gained by letting our universities excel in knowledge creation while encouraging closer links with the rest of society, than by insisting that they should fund themselves mainly through commercializing their knowledge.’ This may intimate that the orientation of policy development for the ERA, particularly that aiming to ‘strengthen the link between the public research base and industry’,7 is now moving into closer alignment with what appears to be the emergent trend in industry–university collaboration in the USA. The latter, however, is not another new institutional model. Quite the opposite, in fact, as the signs are indicating a growing movement to recover a mode of interaction that seemed to have been all but lost in the post-Bayh–Dole era. One harbinger of this trend reversal might be seen in the recently announced Open Collaborative Research Program, under which IBM, Hewlett-Packard, Intel, Cisco Systems and seven US universities have agreed to embark on a series of collaborative software research undertakings in areas such as privacy, security and medical decision-making.8 The intriguing feature of the agreement is the parties’ commitment to make their research results freely and publicly available. Their avowed purpose in this is to be able to begin cooperative


79

work, by freeing themselves from the lengthy delays and costly, frustrating negotiations over IPR that proposals for such collaborative projects typically encounter. This development reflects a growing sense in some corporate and university circles during the past five years that the Bayh–Dole legislation had allowed (and possibly encouraged) too great a swing of the pendulum towards IP protection as the key to appropriating economic returns from public and private R&D investments alike; that the vigorous assertion of IPR was being carried too far, so that it was impeding the arrangement of inter-organization collaborations involving researchers in the private and publicly funded spheres. As Stuart Feldman, IBM’s vice-president for computer science, explained to the New York Times: ‘Universities have made life increasingly difficult to do research with them because of all the contractual issues around intellectual property . . . We would like the universities to open up again.’ A computer scientist at Purdue University is quoted in the same report as having echoed that perception: ‘Universities want to protect their intellectual property but more and more see the importance of collaboration [with industry].’ The empirical evidence about the effects of Bayh–Dole-inspired legislation in the EU that has begun to appear points, similarly, to some negative consequences for research collaboration. Thus a recent study has investigated the effect of the January 2000 Danish Law on University Patenting and found that it led to a reduction in academic–industry collaboration within Denmark (Valentin and Jensen, 2007). But the new law, which gave the employing university patent rights to inventions produced by faculty scientists and engineers who had worked alone or in collaboration with industry, appears also to have been responsible for an increase in Danish biotech firms’ readiness to enter into research collaborations with scientists working outside Denmark – an outcome that must have been as surprising as it was unwelcome to the legislation’s proponents. Clearly, the transfer of institutional rules from the USA to Europe is not a matter to be treated lightly; their effects in different regimes may not correlate at all well. It remains to be seen just how widely shared are these skeptical ‘second thoughts’ about the wisdom of embracing the spirit of the Bayh–Dole experiment, and how potent they eventually may become in altering the modus of industry–university interactions that enhance ‘technology knowledge transfers’, as distinguished from ‘technology ownership transfers’. At present it is still too early to speculate as to whether many other academic institutions will spontaneously follow the example of the Open Collaborative Research Program. Moreover, it seems unlikely that those with substantial research programs in the life sciences and portfolios of biotechnology and medical device patents will find themselves impelled to

80


do so by the emergence of enthusiasm for such open collaboration agreements on the part of drug development firms and major pharmaceutical manufacturers. Nevertheless, from the societal viewpoint, the issue of whether IPR protection is getting in the way of the formation of fruitful collaborations between industry and university researchers is fundamentally a question about the conditions that would maximize the marginal social rate of return on public investment in exploratory research. This could be achieved by making it more attractive for R&D-intensive firms with interests and capabilities in the potential commercial applications to collaborate with publicly funded academic research groups because they hoped subsequently to exploit the knowledge base thereby created. This issue is not unrelated to an important aspect of the concerns that have been raised in regard to potential ‘anti-commons effects’ of the academic patenting of research tools, and the resulting impediments to downstream R&D investment that are created not only by ‘blocking patents’, but by ‘patent thickets’ formed by a multiplicity of IP ownership rights that are quite likely to be distributed among different public research organizations (PROs). The latter would contribute to prospects of ‘royalty stacking’ that would reduce the prospective revenues from a technically successful innovation, and to higher investment costs due to the transactions costs of conducting extensive patent searches and multiple negotiations for the rights to use the necessary set of upstream patents.9 It would seem possible to address the source of this particular problem by allowing, or indeed encouraging, the cooperative formation of efficient ‘common-use pools’ of PRO patents on complementary collections of research tools. While this would strengthen the bargaining position of the collectivity of patent-owning institutions, and it would be necessary to have supervision of the competition authorities to present abuses, it might well increase the licensing of those technologies to downstream innovators. Of course, it is a second-best solution from the societal viewpoint, as the award of ownership rights on inventions that have resulted from publicly funded academic research will result in a ‘deadweight loss’ – due to the effect of the licensing charges that curtail the downstream exploitation of those inventions.10 The specific functionality of the information-disclosure norms and social organization of open science, which until very recently (by historical standards) was strongly associated with the ethos and conduct of academic, university-based research, rests upon the greater efficacy of data and information-sharing as a basis for the cooperative, cumulative generation of eventually reliable additions to the stock of knowledge. Treating new findings as tantamount to being in the public domain fully


81

exploits the ‘public-goods’ properties that make it possible for data and information to be concurrently shared in use and reused indefinitely, and thereby promote the faster growth of the stock of reliable knowledge. This contrasts with the information control and access restrictions that generally are required in order to appropriate private material benefits from the possession of (scientific and technological) knowledge. In the proprietary research regime, discoveries and inventions must either be held secret or be ‘protected’ by gaining monopoly rights to their commercial exploitation. Otherwise, the unlimited entry of competing users could destroy the private profitability of investing in R&D.11 One may then say, somewhat baldly, that the regime of proprietary technology (qua social organization) is conducive to the maximization of private wealth stocks that reflect current and expected future flows of economic rents (extra-normal profits). While the prospective award of exclusive ‘exploitation rights’ have this effect by strengthening incentives for private investments in R&D and innovative commercialization based on the new information, the restrictions that IP monopolies impose on the use of that knowledge perversely curtail the social benefits that it will yield. By contrast, because open science (qua social organization) calls for liberal dissemination of new information, it is more conducive to both the maximization of the rate of growth of society’s stocks of reliable knowledge and to raising the marginal social rate of return from research expenditures. But it, too, is a flawed institutional mechanism: rivalries for priority in the revelation of discoveries and inventions induce the withholding of information (‘temporary suspension of cooperation’) among close competitors in specific areas of ongoing research. Moreover, adherents to open science’s disclosure norms cannot become economically self-sustaining: being obliged to quickly disclose what they learn and thereby to relinquish control over its economic exploitation, their research requires the support of charitable patrons or public funding agencies. The two distinctive organizational regimes thus serve quite different purposes within a complex division of creative labor, purposes that are complementary and highly fruitful when they coexist at the macro-institutional level. This functional juxtaposition suggests a logical explanation for their coexistence, and the perpetuation of institutional and cultural separations between the communities of researchers forming ‘the Republic of Science’ and those who are engaged in commercially oriented R&D conducted under proprietary rules. Yet these alternative resource allocation mechanisms are not entirely compatible within a common institutional setting; a fortiori, within the same project organization there will be an unstable competitive tension between the two and the tendency is for the more fragile, cooperative micro-level arrangements and incentives to be undermined.

82


SOCIAL INTERACTIONS ACROSS ORGANIZATIONAL BOUNDARIES AND THE FACILITATION OF COLLABORATIVE RESEARCH Beyond the overtly commercial and explicitly contractual interactions involving IP, whose role at the macro-system level in supporting R&D investment and innovation tends to be accorded prime place in general policy prescriptions, the importance of other channels of ‘interaction’ with business is often stressed in discussions of what the leadership of Europe’s universities should be doing in that regard. Prominent on this list are the variety of interpersonal and inter-organizational connections that bring participants in academic research into regular contact with members of the local, regional and national business communities. Under the heading ‘The role of the universities in promoting business–university collaboration’, the Lambert Review (2003, p. 41), for example, remarked on the growing role that universities (in the UK) have taken in their cities and regions during recent decades: Vice-chancellors often have links with the CEOs of major local companies, with chambers of commerce, with their development agency and with NHS Trusts and other community service providers in their region. Academics work with individual businesses through consultancy, contract or collaborative research services. University careers services cooperate with the businesses which wish to recruit their graduates or provide work placements for their students.

The trend toward organized institutional involvement – as distinct from personal connections between university professors and industrial and financial firms in their locale – is indeed an ongoing process for many of Europe’s HEIs (higher education institutions). But the reader of the Lambert Review, who was familiar with the US university scene, especially that among the public (land grant) institutions, would be struck by its suggestion of the novelty of top-level administrators having links with CEOs and local business leaders, inasmuch as this would be presumed to be the case for their American counterparts. Also noteworthy, as a reflection of the ‘top-down’ impetus for the establishment of such relationships, is the quoted passage’s emphasis on the cooperation of university careers services with recruiters from business firms. At most major US research universities – where the organized placement services of the professional schools, as well as those of the undergraduate colleges, have long been established – the important recruiting contacts with graduate scientists and engineers are typically arranged at the level of the individual departments, and often are linked with a variety of ‘industrial affiliates’ programs. This is significant in view of the expert


83

screening functions that are implicitly performed for potential employers by universities’ graduate educational programs and faculty supervisors. Screening of scientific and engineering talent, as well as assessment of graduate and postdoctoral research contributions, is a publicly subsidized service (provided as it is without fee) that is especially valuable for companies seeking promising researchers who have been working in frontier areas. All but the largest firms are likely to lack the internal expertise, and are unable to arrange the extended opportunities that internships could provide for observation and evaluation of the capabilities of current graduates. Even the large R&D-intensive corporations find it worthwhile to contribute to the ‘industry associates programs’ of leading departments in the sciences and engineering, if only for the opportunities these afford to form contacts with advanced students and younger faculty who may be approached with offers of employment in the near future.12 The formation of enduring ties for the transfer of knowledge through the movement of personnel gives business organizations access to the craft aspects of applying new techniques, contacts with new recruits’ personal network of other young researchers, and an advantage in spotting exceptional capabilities to conduct high-caliber research. Such ties are sustained by personal relationships with the student’s professors, and strengthened by ‘repeat play’, which tend to inhibit the latter’s inclination to ‘oversell’ members of their current crop of PhDs and postdocs; similarly, the prospects of having to try to recruit next year from the same source works to induce the firms to be more candid in describing the nature of the employment opportunities that professors may recommend to their good students. The point here is that the direct participation of the parties, rather than institutionally provided third-party intermediation services, will generally be a requirement for successful ‘relationship management’ in the market for young research talent. Perhaps the greater prevalence of such arrangements that can be observed in science and engineering departments and research groups at US universities can be attributed to the greater degree of autonomy that university administrations there have allowed to these units, permitting them (and indeed providing them with initial help) to create special programs of lectures, seminars and gatherings of ‘industry associates’ by soliciting and using funds contributed by the business invitees who participate as sponsors of those events. Initiatives of this kind, it must be said, are also an aspect of the traditions of local community and regional involvement that were developed in the agricultural and engineering schools of state (public land grant) universities in America. This form of direct engagement with the society beyond the precincts of the academy has been further reinforced and extended to the private HEIs in the USA

84


by the generally more intense competition among them in the placement of graduating students in national and regional job markets – which is especially pronounced in the cases of the professional schools and graduate science and engineering faculties. Whatever the precise sources of these contrasts may be, the obvious suggestion to be registered here is that interesting interactions and productive engagements of this kind arise under conditions that have allowed greater scope for initiative, and attached rewards to actions taken not by vice-chancellors, but at the levels within these institutions where one is most likely to find the specific information and technical judgments about the subjects of mutual interest to academic researchers and knowledgeseeking corporate personnel. It implies also that, when they work successfully, they do so within an ecological niche that provides a web of supporting connections and mutually reinforcing incentives that need to be studied and understood before attempting to transplant and adapt this important mechanism for ‘connectivity’ in new institutional and cultural settings.13 It is only reasonable that considerable effort will be needed in order to properly align mutual expectations among the parties to a collaboration when they approach the negotiating table with quite different, and conflicting, goals that have been organizationally mandated. Nevertheless, the extent to which that investment is undertaken by both sides does appear to shape strongly whether, and how well, business–university research collaborations turn out to benefit both parties, and whether they are able to evolve into more enduring ‘connected’ relationships. When one starts the alignment process at the upper echelons of the administrative hierarchies of organizations that are differentiated in their purposes and concerns as business companies and universities, the conflicts are likely to appear most salient and the prospective negotiation process more difficult and protracted, and uncertain in outcome, whereas the existence or absence of common interests and appreciation of the magnitude and division of prospective gains from cooperation usually will be quite readily established. The question then is whether the benefits in terms of the enhanced capacity to carry out the projected line of research are deemed sufficiently important to their respective (academic and business) organizations that mutual accommodations will be reached to ‘make it happen’. The organizational structure of most research universities, in which the upper levels of administration typically have at best only a derived interest in pursuing the particular substantive research programs that animate members of their research faculty, are likely to eschew any attempt to evaluate and prioritize among them on the basis of their comparative scientific interest or societal worth. Accordingly, university administrators rarely


85

if ever approach firms with proposals to engage in particular research projects that would involve collaborations between specified groups or individual faculty scientists and engineers and counterparts employed in the business R&D labs. Instead, the research director of a company that has decided that sponsoring a collaborative project with certain universitybased research scientists would be beneficial to the organization’s ‘bottom line’ will usually have authority to take the initiative of approaching the prospective academic partners to discuss such an arrangement. But, as the latter, in their capacities of research faculty members rather than officers of the university, do not usually have corresponding authority to negotiate formal inter-organizational agreements, the business firm’s representatives find themselves told that they must deal with the university administration, and more precisely with one or a number of ‘service units’ within the institution (variously described as the office of external relations, ‘sponsored research office’, ‘university research services’, ‘technology transfer office’, all of which will in one way or another be equipped with legal counsel and contract negotiators). Reasonable as this may appear as a procedure reflecting the different specializations of the people whose expertise the university calls upon, problems with its operation in practice often arise precisely because the primary concerns of these specialized services typically have little to do with the specifics of the professors’ interests in the research collaboration.14 Rather, their professional purpose is to secure such financial benefits as can be extracted by ‘the university’ (directly or indirectly) in exchange for agreeing that its facilities and faculty resources will be permitted to perform their part of the contemplated collaborative work, and that the university will bear responsibility should they refuse to perform in accordance with the terms of the contract. Their competence and role also require their performance of ‘due diligence’ – by trying to identify all the conceivable risks and costs that could stem from their institution’s exposure to legal liabilities and adverse publicity occasioned by participating in the proposed collaboration. The uncertainties about the nature of the products and processes of research, conjoined with the professional incentives of those charged with performing ‘due diligence’ – and their inability to calculate the countervailing value of the losses entailed in not doing the research – tend to promote behaviors that reflect extreme risk aversion. In other words, these agents of the university are predisposed to advocate and adopt a tough bargaining stance, trying to get the other collaborating party (or parties) to bear the liabilities, or the costs of insuring against them; and when that appears to be infeasible, they are not hesitant to counsel that the project not be undertaken by their institution. That this can be an unwelcome

86


surprise to corporate representatives who were under the impression that ‘the university’ would be symmetrically responding to the interest of the faculty counterparts of their own research group is perhaps responsible for the shocked and disparaging terms in which research directors of large, R&D-intensive US companies relate their experiences in negotiations with universities over the IPR to joint R&D ventures.15 What happens in such cases appears to depend upon whether or not the faculty researchers who are keen to do the science are able to persuade people at some higher levels in the university administration that it would not be in the institution’s long-term interest to refuse to allow their research groups to seize the opportunity of a collaboration with the firm in question. When the individuals concerned are valued by their university administration, whether for their academic prestige or for their ability to recruit talented young faculty, or for their track record of success in securing large public research grants and the overhead support that these bring, their persuasive efforts to find a compromise arrangement in which the university does not try to extract the maximum concessions from the firm, or bears more of the risk than its lawyers think is prudent, are likely to be successful. This is especially likely if there is a credible threat that professors will go to another research institution – where, as the formulaic expression puts it in such conversations, they ‘will feel really wanted’. The point of entering into these seemingly sordid details is to highlight the way that complex innovation systems emerge. In the case at hand it will be seen that more active competition among research institutions for productive scientists – especially where it receives additional impetus from the usefulness of their talents in their university’s competition for public research funding – will have the indirect effect of working as a countervailing force against the internal organizational impediments to the formation of spontaneous ‘connectivity’ between academic and business researchers. Regulatory structures that permit universities to compete to attract and retain research faculties that have attained great peer esteem, and public research funding programs whose allocation criteria give weight to excellence and thereby provide high-level administrators with justifications for being seen to depart from risk-averse institutional guidelines in order to accommodate those individuals’ pursuit of interesting research opportunities, therefore affect positively the formation of university– industry connections that are likely to give rise to future innovations. The perspective thus gained might be contrasted favorably with the thrust of the enthusiastic notice given by the Lambert Review (2003, p. 42) to the recent trend toward the opening of ‘corporate liaison offices’ at UK universities:


87

Partly in recognition of the number and complexity of these [business– university] relationships, many universities have developed corporate or business liaison offices, with a specific remit to act as the interface with business. These offices have taken on an increasing number of tasks as universities’ engagement with their wider community has developed. These include developing networks of businesses; marketing the research strengths of the university; advising on consultancy agreements and contract research; arranging complex collaborative research agreements or major joint ventures.

For the university to present business corporations’ representatives with a well-organized corporate academic face, and a central office whose concerns are regulation of external relationships and internal management control of the exploitation of the university’s marketable ‘knowledge assets’, may succeed in making European upper-level executives at both institutions feel increasing ‘at home’ in their new contacts. Yet this organizational measure strikes one as perhaps neither so important nor so well designed to respond to the challenge of drawing R&D managers and research personnel into dense and fruitful networks of knowledge exchange with university-based experts.16 Viewed against these findings, the emphasis that was placed by the text of the Lambert Review upon the mission (‘remit’) of the newly established corporate liaison office to be the university ‘interface with business’ is quite striking. To have a liaison officer advising firms of the formal requirements the university will impose upon consultancy agreements and contract research, particular those involving complex collaborative research agreements, certainly is appropriately instructive when there is no room for flexibility. Yet putting this function in the hands of a central liaison office encourages pre-commitment of the university to the inflexibility of ‘standard-form contracts’, and thus tends to reduce the scope for exploring a variety of possible legal arrangements for the assignment of intellectual property rights, obligations and liabilities that would be responsive to the particular needs of the research collaborators, as well as the concerns of the participating corporate entities. Liaison officers, as the agents of university administrations, are likely to have much stronger career incentives to attend to the priorities of those responsible for monitoring and regulating the formalities of the university’s external transactions than to seek ways of fulfilling the actual research raison d’être that provides the impetus for the formation of successful and more sustained inter-organizational connections.17 To appreciate the tangled lines of influence and indirect effects is to recognize why systems analysis is so necessary in the diagnosis of institutional problems and the design of corrective measures for the innovation process.

88


‘ONLY CONNECT’ – TOWARDS FOSTERING A VIBRANT ORGANIZATIONAL ECOLOGY OF INNOVATION To form ‘a system of innovation’, the organizations and the individuals in them have to interact in a way that contributes solutions to innovation problems. Systems depend on connections (interactions) and cannot be described or understood simply in terms of their components. What is at stake here is an idea that goes back to Alfred Marshall’s concept of the internal and external organization of a firm (in Industry and Trade, 1919, and Principles of Economics (8th edition) 1920). Flows of knowledge from outside the firm’s boundaries are important determinants of its capabilities and actions, but this information is not simply ‘in the ether’. A firm has to invest in the organization to gather this information and feed it into and adapt it to its internally generated information. Innovative activity is perhaps the most important case of the firm’s reliance on external sources of information, and leads to the idea that the firm is embedded in a wider matrix of relations that shape its ability to innovate. Hence the concern for the various ways in which universities may contribute to the innovation process that we have outlined above. But it is important to recognize that a firm’s internal and external organization constitutes an operator that is simultaneously facilitating and constraining. The codes and information structuring routines that firms invest in to interact with other external sources of information may also serve to filter and blinker the firm’s appreciation of the information that is important and that which isn’t (Arrow, 1974). Thus the innovation systems that a firm is part of are not always plastic in the face of changes in the knowledge environment; and, as a consequence, they fail because their reading of new information is deficient. We should not lose sight of the probability that an innovation system generated to solve one set of problems may prove counterproductive in the context of a new and different set of problems, which is why the processes for flexibly assembling and disassembling specific innovation systems add greatly to the adaptability of any economy (see Metcalfe, 2007). The policy problem may then be put starkly: ‘Is it possible to improve on the spontaneous self-organization process of the already existing and refined interaction between firms and research universities in Europe?’ That the answer may be in the affirmative suggests that the innovation policy response falls into two related branches: 1.

Policy to improve the chances of innovation systems being formed from the knowledge ecologies of the member states, a problem that is


2.

89

largely about barriers and incentives to collaborate in the solution of innovation problems. Policy to improve the quality of the knowledge ecologies in the member states assessed in terms of the overall supply of researchers in different disciplines and the way in which they are organized to produce knowledge.

The preceding sections have focused on the role of research universities, and less so on other public research organizations in relation to these two policy problems. But two general points should be recognized to underlie the whole discussion: it is business firms that occupy the central role in the realization of innovations, and it is the mix of market and non-market interactions that shapes the incentives, the available resources and the opportunities to innovate. Innovation, obviously, is more than a matter of invention, and so it is particularly important not to equate innovation policy with policy for science and technology. University–business linkages form only part of this system and their influence on innovation cannot be independent of the many other factors at play. Thus, for example, the competitive implications of the single market will influence the incentives to innovate, whether interpreted as opportunities or threats to a firm’s position. Consequently, competition policy is de facto an important component of a broad innovation policy just as innovation policy is de facto an important component of competition policy. The fact that the knowledge ecology of the EU has been changing rapidly in the past two decades, and that there are important differences in the richness of the ecologies in different member states, adds further problems in understanding the implications for the innovation process. The prevailing division of labor in the European knowledge ecology has not arisen by chance, but rather as a reflection of many years of evolution in the comparative advantages of different organizations in producing and using knowledge. Firms, for example, have evolved in ways quite different from universities because they perform different sets of tasks and fulfill quite different societal functions. This division of labor needs to be respected and understood, for it would be as foolish to make universities behave like firms as it would be economically disastrous to make private firms operate like universities. The differences in their respective modes of operation are not accidental, but have a functional purpose. The origins of the current ecology, of which the governance of universities is a part, can be traced back to a historical epoch when the knowledge foundations of industrial processes owed little to systematic scientific understanding and the formal organization and conduct of R&D activities. The modern age is different, however: the great expansion of

90


organized public and private science and engineering research activities that took place during the second half of the twentieth century, and accelerated the shift in the structures of the ‘industrialized’ economies toward ‘services’ and away from commodity production, are two important transformations that have in a sense made the university as an institution appear to be, at least outwardly, less distinct from other corporate entities than was formerly the case.18 The relevant issue, then, must remain how best to achieve coordination of this division of labor and thereby enhance innovation processes. As we have explored above, the different ‘cultures’ of business and the public research sector need special attention. The distinguishing feature of fundamental research in science and technology is its open nature, its nature as a science commons (see Cook-Deegan, 2007; Nelson, 2004). Open science (including engineering technology) is a collective endeavor that bases the reliability of the knowledge production processes on widespread agreement as to methods of evaluation and replication, but bases radical progress of knowledge on disagreement, the scientific equivalent of creative enterprise. This tension between order and agreement and change and disagreement is at the core of the institutions that shape science. Similarly, in regard to commercial innovation, disagreement is the defining characteristic of any significant innovative enterprise that is necessarily based on a conjecture that imagines that the economic world can be ordered differently. It is the open market system that facilitates adaptations to such disagreement and generates powerful incentives to disagree: the instituted procedures of science and business are open ‘experiment generating systems’; both work within different principles of order and both depend for their progress on the productive channeling of disagreement. The consequences are that the knowledge-generating and using processes of businesses and of PROs operate with different cultures, different value systems, different time frames, and different notions of what their principal activities are. Thus the principal outputs of universities are educated minds and new understandings of the natural and artificial worlds, economy, society and so on. The outputs of business are different, and involve new understandings of productive and commercial processes for the purpose of producing outputs of goods and services to be sold at a profit. Universities operate with one kind of governance system to achieve their aims, private firms with quite different governance systems, and these differences materially influence their interactions in the pursuit of innovation. As has been pointed out, this results in very different norms for the production and sharing of knowledge within and between the two systems. In both business and the academy, positive feedback processes are in


91

operation so that success breeds success. The profits from existing activities that provide the basis for subsequent innovation in a firm have their equivalent in the university in terms of research reputations that serve to attract high-quality staff and funding. Indeed, the institutions of science are partly designed to create and reinforce this process. The currently articulated attempts by some member states to accelerate this reputation effect through the competitive allocation of teaching and research funds are bound to further concentrate reputations on a relatively small number of universities. Because there are strong potential complementarities between the conduct of exploratory, fundamental research in institutions organized on the ‘open science’ principle, and closed proprietary R&D activities in the private business sector, it is doubly important to establish market and non-market arrangements that facilitate information flows between the two kinds of organization. The returns on public investment in research carried on by PROs can be captured through complementary, ‘valorizing’ private R&D investments that are commercially oriented, rather than by encouraging PROs to engage in commercial exploitation of their knowledge resources. This is why the strategy that has been expressed in the EU’s Barcelona targets is important: by raising the rate of business investment in R&D, Europe can more fully utilize the knowledge gained through its public research and training investments, and correspondingly capture the (spillover) benefits that private producers and consumers derive from the application of advances in scientific and technological knowledge. Knowledge transfer processes can be made more effective by attention to the arrangements that are in place at the two main points of the public research institutions’ connections with their external environments. That a research institute or a university may acquire the attributes of an isolated, inward-looking ‘ivory tower’ is well understood, and their internal processes in many cases tend to encourage this. Universities in the EU are frequently criticized for operating with internal incentive structures that reward academic excellence in teaching and research independently of any potential application to practice in the business or policy realms. This concern is reflected in the newly attributed ‘third stream’ or ‘triangulation’ of the university system, defined as ‘the explicit integration of an economic development mission with the traditional university activities of scholarship, research and teaching’.19 Third-stream activities are of many different kinds, and here it is important to distinguish those activities that seek the commercialization of university research (technology licenses, joint ventures, spin-offs and so on) from activities of a more sociopolitical nature that include professional advice to policy-makers, and contributions to cultural and social life (see OEU, 2007). What is significant about the

92


current debate is the emphasis on the commercialization activities. What is less well understood, and possibly will remain elusive, is how to design institutional arrangements that successfully support commercialization while not inhibiting the performance of research and teaching functions that are the primarily social raison d’être for the continued maintenance of the universities as a distinctive organizational and cultural form.

A SUMMARY Researchers in Western universities do make important, fruitful connections with business firms, and indeed have done so for many, many years. But the pressures and changes that Europe’s universities now face in markedly different innovation ecologies have raised questions that focus attention on the purposes and efficacy of the current extent and modes of university–business interactions. Innovation ecologies only form into innovation systems when the different organizations from the ecology are connected for the purpose of solving innovation problems. Since universities and firms are part of a complex division of labor in which each has evolved unique characteristics relative to their primary functions, it is not to be wondered at that the practices that support these functions do not automatically facilitate interactions among these differentiated organizations. Therefore public and private policy consequently has an important role to play in respect of the richness and diversity of Europe’s innovation ecology, and with respect to the ways in which connections can be formed and re-formed to promote a higher rate of innovation. But a fresh look, and a ‘rethink’ within an explicit ‘systems’ or organizational ecology framework is in order, because some of the main institutional innovations that have been promoted with a view to enhancing the exploitation of university research do not seem to be the most beneficial ways of ensuring that university knowledge is translated into greater economic wealth. Indeed, continuing to seek to overcome the barriers to connecting publicly funded research conducted in academic institutions with commercial application, by having those organizations become dependent upon commercialization of research findings and behaving as a proprietary performer of R&D, simply is not sensible. It would jeopardize the open science arrangements that are more effective for the conduct of fundamental, exploratory research – a function that must be fulfilled by some institution if a basis for long-run productivity growth is to be sustained. Policies that would add to existing pressures on academic communities and their leaders to take on new and different missions, for which their


93

historical evolution and specialized characteristics have not equipped them, run the risk of damaging their ability to fulfill critical functions that no other organizations in the society are prepared to perform with comparable effectiveness. The recognition of a need for new missions in the generation and transmission of knowledge suited to solve problems of innovation in the economy, therefore, should redirect attention to more creative solutions. These are likely to involve the development of alternative equally specialized bridging organizations that would gain expertise in the forging of diverse inter-organizational links between the worlds of the academy and the worlds of business. To explore that option as a promising way forward, however, lies well beyond the scope of the present chapter and so must be left to stand as a grand challenge for ‘mechanism design’.

ACKNOWLEDGMENT This chapter was previously published in Research Policy, Vol. 35, No. 8, 2006, pages 1110–21.

NOTES 1. 2. 3.

4.

5.

It is unnecessary to review the details of these misunderstandings, which are discussed in David (2007). For a comparison of the questionable effects of Bayh–Dole on the licensing activities of three major US universities see Mowery et al. (2001). Balconi et al. (2004); Geuna and Nesta (2006); Crespi et al. (2006). The statistics presented by Balconi et al. (2004) in table 3 refer to 20 specific science and engineering research fields in which at least 20 academic inventors (of all nationalities) could be observed in the EPO patent data for the years 1978–99. The proportions referred to in the following text pertain to Italian academic inventors as a fraction of total faculty enrolments in the corresponding fields at Italian universities and polytechnics on 31 October 2000. Paradoxically, this was the practice despite the fact that at that time Italian universities had titular rights to own the patents filed by their employees, which was anomalous in the context of the German, Dutch and other national universities at the time; the practice in Italy removed the anomaly by permitting professors to assign the rights directly to industrial companies – a practice that was subsequently ratified by a change in the Italian law. That change seemed, quixotically, to run against the stream of Bayh–Doleinspired ‘reforms’ that were under way in other nations’ university systems at the time, giving patent rights formerly held as the professors’ prerogatives, to their employers. Operationally, however, the Italian reform was more in accord with the intention of facilitating the transfer of new technologies to industry, legalizing the way it had previously been done. The identity of the parties in such bargains, of course, is defined by the regulations governing the initial assignment of the inventor’s patent rights, which, in the general situation, reside in the first instance with the institution at which the R&D work was conducted. Italy is the signal exception in the EU, as changes in Italian law placed the

94

6.

7.

8. 9.

10.

11. 12.

13.

The capitalization of knowledge patent in the hands of the inventing professor(s). In principle the latter could assign the rights to a university, which could in turn bargain with a firm over the terms of a license to exploit it. In this regard it is significant that the latter considerations led the Italian government to award ownership rights in patents to their faculty employees, whereas the industrial treatment of ‘work for hire’ by employed inventors was applied to university faculty by all the other European states. Thus, in Denmark, PROs including universities were given the rights to all inventions funded by the Ministry of Research and Technology (in 1999); French legislation authorized the creation of TTOs (technology transfer offices) at universities (in 1999), and university and PRO assertion of rights to employee inventions was ‘recommended’ by the Ministry of Research (in 2001); the ‘professor’s privilege’ was removed in Germany by the Ministry of Science and Education (in 2002); in Austria, Ireland, Spain and other European countries the employment laws have been altered to removed ‘professor’s exemption’ from the assignment to employers of the IP rights to the inventions of their employees. See OECD (2003); Mowery and Sampat (2005). The quoted phrase is the single most frequently cited national policy development among those listed in a country-by-country summary of the 25 EU member states’ ‘National policies toward the Barcelona Objective’, in European Commission (2003), Table 2.1, pp. 29ff. See Lohr (2006). The universities involved are UC Berkeley, Carnegie Mellon, Columbia University, UC Davis, Georgia Institute of Technology, Purdue University and Rutgers University. For further discussion of the literature on the economics of the so-called ‘anticommons’, and the critical importance of ‘multiple-marginalization’ as a source of inefficiency that is potentially more serious than that which would result from the formation of a cartel, or profit-maximizing pool among the holders of complementary patents, see David (2008). There was something not so foolish, after all, in the old-fashioned idea of upstream public science ‘feeding’ downstream research opportunities to innovative firms. The worry that this will not happen in the area of nanotechnology (see Lemley, 2005) brings home the point about the unintended consequences of the success of national policies that aimed at building a university-based research capacity in that emerging field. The idea was not to allow domestic enterprise to be blocked by fundament patents owned by other countries. That they might now be blocked by the readiness of PROs on their home terrain seeking to exploit their control of those tools is a disconcerting thought. For points of entry into the growing economics literature on the impact of academic patenting upon exploratory research investments, and the ‘anti-commons’ question (specifically, the ambiguities of recent empirical evidence regarding its seriousness), see David (2003); Lemley and Shapiro (2007). This and the following discussion draw upon Dasgupta and David (1994) and David (2003). The value of the screening function for employers is the other side of the coin of the ‘signaling’ benefits that are obtained by young researchers who trained and chose to continue in postdoctoral research positions in academic departments and labs where publication policies conform to open science norms of rapid and complete disclosure. On job market signaling and screening externalities in this context see, e.g. Dasgupta and David (1994), section 7.1, pp. 511–513. This caution might be subsumed as part of the general warning against the ‘mix-andmatch’ approach to institutional reform and problem selection in science and policymaking, a tendency that is encouraged by international comparative studies that seek to identify ‘best practices’, as has been pointed out by more than one observer of this fashionable practice (but see, e.g., David and Foray (1995). Examining particular institutions, organizational forms, regulatory structures, or cultural practices in isolation from the ecologies in which they are likely to evolve, and searching for correlations


95

between desired system-level outcomes and their presence in the country or regional cross-section data, has been fashionable but as a rule offers little if any guidance about how to move from one functional configuration to another that will be not only viable but more effective. 14. The difficulties occasioned by this internal organizational structure of universities, which contributes to separating the interest of the institution as a ‘research host’ from that of its faculty researchers, thereby placing these research ‘service units’ in a regulatory role vis-à-vis the latter, are considerable. But they are far from arbitrary or capricious, in view of the potential legal complexities that contractual agreements for collaborative research performance may entail. For further discussion see David and Spence (2003). 15. See the 2003 survey results reported by Hertzfeld et al. (2006). See also David (2007), esp. table 1 and text discussion. 16. It is consequently a bit surprising to find the following statement, attributed to the Lambert Review of Business–University Collaboration (HM Treasury, 2003), p. 52, n. 110: ‘Indeed, the best forms of knowledge transfer involve human interaction, and European society would greatly benefit from the cross-fertilization between university and industry that flows from the promotion of inter-sectoral mobility.’ 17. These issues are examined in some detail in David and Spence (2003). 18. While this does not imply that other institutions and organizations are more interchangeable with the universities in the performance of a number of the latter’s key functions in modern society, it has contributed to the recent tendency of some observers to suggest that universities as deliverers of research and training services might be more effective if they emulated business corporations that perform those tasks. 19. See Minshull and Wicksteed (2005). Activities of this nature are not linked solely to academy–industry interactions. The tripartite missions in health care to link biomedical research with clinical service delivery and clinical education across hospitals and university medical schools have been widely adopted in the USA and UK. In the latter they are known as academic clinical partnerships, and they provide the framework within which much NHS-funded research is carried out. See Wicksteed (2006).

REFERENCES Arrow, K.J. (1974), The Limits of Organization, London and New York: W.W. Norton. Balconi, M., S. Breschi and F. Lissoni (2004), ‘Networks of inventors and the role of academia: an exploration of Italian patent data’, Research Policy, 33 (1), 127–45. Cook-Deegan, R. (2007), ‘The science commons in health research: structure, function and value’, Journal of Technology Transfer, 32, 133–56. Crespi, G.A., A. Geuna and B. Verspagen (2006), ‘University IPRs and knowledge transfer. Is the IPR ownership model more efficient?’, presented to the 6th Annual Roundtable of Engineering Research, Georgia Tech College of Management, 1–3 December, available at http://mgt.gatech.edu/news_room/ news/2006/reer/files/reer_university_iprs.pdf. Dasgupta, P. and P.A. David (1994), ‘Toward a new economics of science’, Research Policy, 23, 487–521. David, P.A. (2003), ‘The economic logic of “open science” and the balance between private property rights and the public domain in scientific data and information: a primer’, in J. Esanu and P.F. Uhlir (eds), The Role of the Public

96


Domain in Scientific and Technical Data and Information: A National Research Council Symposium, Washington, DC: Academy Press. David, P.A. (2007), ‘Innovation and Europe’s academic institutions – second thoughts about embracing the Bayh–Dole regime’, in F. Malerba and S. Brusoni (eds), Perspectives on Innovation, Cambridge: Cambridge University Press. (Available as SIEPR Policy Paper 04-027, May 2005, from http://siepr.stanford. edu/papers/pdf/04-27.html, accessed January 2008.) David, P.A. (2008), ‘New moves in “legal jujitsu” to combat the anti-commons’, Keynote presentation to the COMMUNIA Conference on the Public Domain in the Digital Age, Louvain-le-Neuve, Belgium, 30 June–1 July, available at http://communia-project.eu/node/115. David, P.A. and D. Foray (1995), ‘Accessing and expanding the science and technology knowledge base’, STI Review: O.E.C.D. Science, Technology, Industry, 16 (Fall), David, P.A. and M. Spence (2003), Towards Institutional Infrastructures for e-Science: The Scope of the Challenges, A Report to the Joint Information Systems Committee of the Research Councils of Great Britain, Oxford Internet Institute Report No. 2, September, available at http://www.oii.ox.ac.uk/resources/publications/OIIRR_E-Science_0903.pdf, last accessed 28 January 2008.. European Commission (2003), ‘National policies toward the Barcelona Objective’, in Investing in research: an action plan for Europe, Brussels EUR 20804 (COM, 2003, 226 final). European Commission (2005), ‘European universities: enhancing Europe’s research base’, Report by the Forum on University-based Research, EC-DG Science and Society, May, available at http://eur-lex.europa.eu.Result. do?idReg=14&page=7}. European Commission (2007), ‘EC Staff Working Document’ (COM, 2007, 161/2). Geuna, A. and L. Nesta (2006), ‘University patenting and its effects on academic research: the emerging European evidence’, Research Policy, 35 (June–July), [P.A. David and B.H. Hall (eds), Special Issue on Property and the Pursuit of Knowledge: IPR Issues Affecting Scientific Research]. Hertzfeld, H.R., A.N. Link and N.S. Vonortas (2006), ‘Intellectual property protection mechanisms in research partnerships’, Research Policy, 35 (June–July) [P.A. David and B.H. Hall (eds), Special Issue on Property and the Pursuit of Knowledge: IPR Issues Affecting Scientific Research]. HM Treasury (2003), The Lambert Review of Business–University Collaboration, London: HMSO, available at http://www.lambertreview.org.uk. Lambert, Richard (2007), Lambert Review in EC Staff Working Document (COM, 2007, 161/2). Lemley, M.A. (2005), ‘Patenting nanotechnology’, October, available at: http:// siepr.stanford.edu/programs/SST_Seminars/index.html. Lemley, M.A. and C. Shapiro (2007), ‘Royalty Stacking and patent hold-up’, January, available at http://siepr.stanford.edu/programs/SST_Seminars/index. html. Lohr, Steven (2006), ‘I.B.M. and university plan collaboration’, New York Times, 14 December, available at http://www.nytimes.com/2006/12/14/technology/14blue. html. Marshall, A. (1919), Industry and Trade, London: Macmillan. Marshall, A. (1920), Principles of Economics, London: Macmillan.


97

Metcalfe, J.S. (2007), ‘Innovation systems, innovation policy and restless capitalism’, in F. Malerba and S. Brusoni (eds), Perspectives on Innovation, Cambridge: Cambridge University Press. Minshull, T. and B. Wicksteed (2005), University Spin-Out Companies: Starting to Fill the Evidence Gap, Cambridge: SQW Ltd. Mowery, D.C. and B.N. Sampat (2005), ‘Bayh–Dole Act of 1980 and university– industry technology transfer: a model for other OECD governments?’, Journal of Technology Transfer, 20 (1–2), 1115–27. Mowery, D.C., R.R. Nelson, B. Sampat and A.A. Ziedonis (2001), ‘The growth of patenting and licensing by US universities: an assessment of the effects of the Bayh–Dole act of 1980’, Research Policy, 30, 99–119. Nelson, R.R. (2004), ‘The market economy and the scientific commons’, Research Policy, 33, 455–71. Observatory of the European University (OEU) (2007), Position Paper, PRIME Network: http://www.prime-noe.org. OECD (2003), Turning Science into Business: Patenting and Licensing at Public Research Organizations, Paris: OECD. Valentin, F. and R.L. Jensen (2007), ‘Effects on academia–industry collaboration of extending university property rights’, Journal of Technology Transfer, 32, 251–76. Wicksteed, S.Q. (2006), The Economic and Social Impact of UK Academic Clinical Partnerships, Cambridge: SQW.co.uk.

3.

Venture capitalism as a mechanism for knowledge governance1 Cristiano Antonelli and Morris Teubal

1.

INTRODUCTION

New dedicated capital markets specialized in the public transactions of the stocks of ‘science-based companies’ emerged in the USA during the 1970s. These new financial markets enable the anticipation of returns stemming from the economic applications of technological knowledge, bundled with managerial competence, but non-embodied in either capital or intermediary goods. As such the financial markets have, for the first time in history, promoted the creation and growth of a specialized segment of ‘inventor’ companies and favored public transactions in technological knowledge as an activity per se. These new financial markets are becoming a key component of an innovation-driven novel institutional system termed ‘venture capitalism’. This is key for a new model of ‘knowledge-based’ growth relevant not only for information and communication technologies but also for biotechnologies and new radical technologies at large (Perez, 2003). As such, venture capitalism can be considered a major institutional innovation that enables higher levels of knowledge governance. The basic ‘innovation’ here is not technological but rather institutional, as it consists in a new hybrid organization based upon the bundling of knowledge, finance and competence into new science-based startup firms and in the trade of their knowledge-intensive property rights in dedicated institutional financial markets (Hodgson, 1998; Menard, 2000, 2004; Menard and Shirley, 2005). In order to grasp the process that has led to its introduction we shall rely upon the complexity approach to the economics of innovation. The application of the tools of complex system dynamics to the economics of innovation enables us to analyze the role of new multi-agent structures such as the new financial markets characterized by higher-level organizations. These ‘higher levels of organization’ in fact are forms of organized complexity that favor the generation and dissemination of technological 98

Venture capitalism and knowledge governance

99

knowledge into economic systems. Specifically venture capitalism can be considered a major institutional innovation that provides a platform for the more effective exploitation of technological knowledge bringing together into a coalition for innovation a variety of complementary players such as ‘inventors’, venture capital companies, managerial skills and investment funds, large incumbents searching for new sources of technological knowledge and families looking for new financial assets, and stirring their participation and active contribution to a collective undertaking (Lane, 1993; Lane and Maxfield, 2005; Antonelli, 2008; Lane et al., 2009). This chapter elaborates the view that venture capitalism has improved the governance of technological knowledge within economic systems, and hence has reshaped the prime mechanism by which the generation of new knowledge can lead to economic growth (Nelson, 1994, 1995; Quéré, 2004). The rest of the chapter is organized as follows. Sections 2 and 3 provide the analytical background. Specifically, Section 2 provides the basic economics of the relationship between finance and innovation, and highlights the advantages of the new financial markets in providing funds to sciencebased startup companies with respect to previous institutional arrangements such as banks and incumbent corporations. Section 3 explores the basic elements of the economics of markets as economic institutions. Section 4 shows the complexity of interactions that led to the emergence of the new financial markets. The conclusions highlight the main results.

2.

FINANCE AND INNOVATION: THE FRAMEWORK

Knowledge as an economic good exhibits major limitations in terms of radical uncertainty, non-divisibility, non-excludability, non-exhaustibility, non-appropriability and non-rivalry in use. Much economic analysis has explored the implications with respect to the tradability of knowledge (Arrow, 1962). Yet the limitations of knowledge as an economic good have major implications also in terms of the provision of finance to fund its generation and use. Major asymmetries shape the interaction between prospective funders and prospective innovators. The access to financial markets for innovative projects is seriously limited by the radical uncertainty that characterizes both the generation and the exploitation of new knowledge. Prospective lenders and investors are worried by the combined high levels of risk: (1) that the activities that have been funded with their own money will not succeed, and (2) that the new knowledge, occasionally generated, will

100


not be appropriated by the inventor, at least to an extent that makes it possible to repay the credits and remunerate the capital invested. Even in the case of a successful generation, funders have good reasons to worry about dissipation stemming from uncontrolled leakages of proprietary knowledge. As a consequence, worthy inventive activities and innovative projects risk being jeopardized because of the lack of financial resources (Hall, 2002). Stiglitz has provided two fundamental tools to analyze the relationship between finance and innovation. With the first stream of contributions, Stiglitz (Stiglitz and Weiss, 1981; Stiglitz, 1985) has shown that equity finance has an important advantage over debt in the provision of funds to innovative undertakings because investors have the right to claim a share of the profits of successful companies. While lenders can claim only their credits, investors can participate to the bottom tail of the highly skewed distribution of positive returns stemming from the generation of new knowledge and the introduction of new technologies. This has important consequences in terms of reduction of both the risks of credit rationing and the costs of financial resources for research activities. Lenders need to charge high interest rates in order to compensate for the risks of failure and to discriminate among new research activities to avoid as many ‘lemons’ as possible. Equity investors instead find an equilibrium rate of return at much lower levels because they can participate in the huge profits of a small fraction of the new ventures. The fraction of lemons that equity can support is much larger than that of debt, hence financial equity can provide a much larger amount of funding for research activities. With a second line of analysis, Stiglitz (Sah and Stiglitz, 1986, 1988) has provided the distinction between hierarchies and polyarchies as alternative mechanisms to manage different types of risk. Hierarchical decision-making is better able to avoid the funding of bad projects. Yet the ability of hierarchies is limited by the scope of their competence: their decision-making tends to favor minor, incremental changes. Polyarchic decision-making, on the other hand, experiences higher risks of including bad projects, for example Type 1 errors, but yields higher chances of inclusion of outstanding projects. According to Stiglitz, hierarchical decisionmaking fits better in economic environments characterized by low levels of entropy and radical uncertainty. Conversely, polyarchic decision-making applies better in times when the levels of radical uncertainty are higher. The distinction between Type 1 and Type 2 errors proves very useful in assessing the working of alternative mechanisms and forms of decisionmaking in the selection and implementation of new technological knowledge. The argument elaborated by Stiglitz can be used upside-down so as to investigate what type of decision-making yields higher results in


101

terms of the generation of new technological knowledge and the eventual introduction of innovations. Hierarchies are more likely to incur Type 2 errors that arise when good innovative projects are excluded. Hence hierarchical decision-making has higher chances of favoring incremental innovations and to excluding innovative undertakings that are disruptive and may engender problems in terms of discontinuities both with respect to the existing knowledge base and sunk costs. Polyarchic decision-making, based on a variety of competences, selected on a professional basis according to their expertise, and less exposed to vested interests, on the contrary, favors the inclusion of a wider range of projects. As a consequence, polyarchies tend to include also bad projects. But the likelihood that outstanding projects are retained is much higher. The occurrence of radical innovations seems higher with polyarchic architectures. The combination and implementation of the two tools provided by Stiglitz enable the comparative assessment of the alternative institutional mechanisms designed to handle the relationship between finance and innovation, and identified by Schumpeter: banks and corporations. The analysis of their limitations, with the tools provided by Stiglitz, enables us to identify the emerging venture capitalism as a third distinctive mechanism. In his Theory of Economic Development, Schumpeter stresses the central role of the provision of appropriate financial resources to entrepreneurs. The natural interface of the entrepreneur, as a matter of fact, is the innovative banker. The banker is innovative when he or she is able to spot new opportunities and select among the myriad of business proposals that are daily submitted, those that have higher chances of getting through the system. With a given quantity of financial resources, the innovative banker should be able to reduce the flow of funds towards traditional activities and switch them towards the new firms. The innovative banker should be able to identify the obsolete incumbents that are going to be forced to exit by the creative destruction that follows the entry of successful innovators. Banks can be considered much closer to polyarchic decision-making. They can rely upon a variety of expertise and competence, hired on a professional basis. Their competence is much less constrained by a given scope of expertise, and the effects of irreversibilities and vested interests are much lower. As such, banks seem better able to avoid Type 2 errors. Banks have a clear advantage in the screening process, but their action is limited by clear disadvantages in the participation in the profits stemming from new innovative undertakings. Banks are exposed to the intrinsic asymmetry between debt and equity in the provision of funds to innovative undertakings. This is true especially when radical innovations occur. The higher the discontinuity brought about by radical innovations, the

102


larger the risks of failure of new companies. Banks bear the risks of the failure of firms that had access to their financial support but cannot share the benefits of radical breakthroughs. As Schumpeter himself realized, this model, although practiced with much success in Germany in the last decades of the nineteenth century, suffered from the severe limitations brought about by this basic asymmetry. Schumpeter not only realized the limits of the first model but identified the new model emerging in the US economy at the beginning of the twentieth century. The analysis of the corporation as the institutional alternative to the ‘innovative banker’ has been laid down in Capitalism, Socialism and Democracy. Here Schumpeter identifies the large corporation as the driving institution for the introduction of innovations. His analysis of the corporation as an innovative institutional approach to improving the relationship between finance and innovation has received less attention than other facets (King and Levine, 1993). The internal markets of the Schumpeterian corporation substitute external financial markets in the key role of the effective provision and correct allocation of funds combining financial resources and entrepreneurial vision within competent hierarchies. Corporations, however, are much less able to manage the screening process. Internal vested interests and localized technological knowledge help reduce the risks of funding bad projects but risk reducing the chances that radical innovations are funded. The Schumpeterian corporation confirms that equity finance is more effective than debt finance for channeling resources towards innovative undertakings, but with a substantial bias characterized by continuity with the existing knowledge base. The model of finance for innovation based upon the corporation ranks higher than the model based upon banks in that equity finance is more efficient than debt-based finance with respect to risk-sharing, but has its own limitations arising from the reduction of the centers able to handle the decision-making and the ensuing reduction of the scope of competence that filters new undertakings. In the second part of the twentieth century a few corporations concentrated worldwide a large part of the provision of finance for innovation. The limited span of competence of a small and decreasing number of incumbents became less and less able to identify and implement new radical technologies: a case of lock-in competence could be observed. The corporation has been able for a large proportion of the twentieth century to fulfill the pivotal role of intermediary between finance and innovations, but with a strong bias in favor of incremental technological change. The screening capabilities of corporations fail to appreciate radical novelties. The integration of these two strands of analysis highlights the radical mismatch between the distinctive competence and the competitive advantage


103

of the two traditional modes of provision of financial resources to innovation. Both in equity and debt finance, exploitation conditions on the one hand and competence on the other, are not aligned and are actually divergent. Banks, as polyarchies, are better able to identify and fund radical innovations but cannot participate into their extraprofits, as they provide debt and not equity. On the contrary they are exposed to the high rates of failures stemming from Type 1 errors, for example, the higher incidence of ‘lemons’ into their portfolios of funded projects. Corporate provision of funds to internal R&D projects selected by internal and hierarchical decision-making is less inclined to identifying and funding radical innovations that would benefit larger firms as equity providers. Corporations are better able to fund minor, incremental innovations where their competitive advantage in exploitation is lesser because the latter are less likely to earn extra profits. This misalignment between the distinctive exploitation conditions and the intrinsic competence of the two traditional institutions has the clear effect of reducing the incentives to the provision of funds for innovation, and of increasing the interest rates for debt finance. Together with the limits of knowledge as an economic good, this institutional misalignment is one of the main causes of underinvestment in the generation of technological knowledge and hence undersupply of innovations. A mechanism based upon a screening procedure performed by competent polyarchies and the equity-based provision of finance to new undertakings would clearly combine the best aspects of each model. Venture capitalism seems more and more likely to emerge as the third major institutional set-up able to manage the complex interplay between finance and innovation when radical changes take place. As a matter of fact, venture capitalism combines the advantages of distributed processing typical of polyarchies with the advantages of equity-based finance over debt-based finance. Venture capitalism makes it possible to combine the more effective identification of radical innovations with the more effective sharing of risks associated with the provision of funds. Table 3.1 provides a synthetic account of the analysis conducted so far. The bank-based provision of funds to innovation suffers the limits of debt-based finance but ranks higher in terms of distributed processing. The advantages of distributed processing are larger, the larger the number of banks, and the larger the number of independent agents that participate in the screening process. The corporation model is less able to avoid Type 2 errors but enjoys the advantages of the equity-based provision of finance to innovation. The corporation model suffers especially from the grip of the past that sunk costs and the irreversibilities of tangible and intangible capital exert upon the appreciation of new disruptive technologies. It is also clear that the smaller the number of corporations that control the

104


Table 3.1

Limits and advantages of alternative financial systems for innovations Polyarchies

Debt finance

Equity finance

Banks experience more Type 1 errors funding bad projects because of low competence levels but favor the introduction of radical innovations; as lenders however they cannot participate into their extra profits Venture capitalism favor the introduction of radical innovations and participate into the fat tails of profits of new ventures

Hierarchies

Corporations can participate into the fat tail of profits of new ventures, and are better able to sort out bad projects, but are limited by higher probability to commit Type 2 errors reducing the rate of introduction of radical innovations

funding of innovative undertaking, the higher the risks of Type 2 errors at the system level. Venture capitalism seems able to combine the advantages of the corporation model in terms of equity-based provision of funds for innovation, with the distributed processing typical of the banking system. The emergence of the new, dedicated financial markets specialized in the public transactions of the knowledge-intensive property rights of new science-based startup companies is a key aspect of venture capitalism. As such it requires a dedicated analysis.2 In order to grasp the emergence of the new financial markets specialized in the transactions of knowledge-intensive property rights, it is necessary to revisit the basic elements of the economics of markets.

3.

MARKETS AS ECONOMIC INSTITUTIONS

Markets as an Economic Problem Markets are economic institutions that emerge when an appropriate combination of complementary conditions occurs. Markets are the product of social and institutional change. As such, they evolve over time: they


105

can decline and emerge. At each point in time, markets differ. Markets can be classified according to their characteristics and their functionality. The emergence and upgrading of a market is the result of an articulated institutional process that deserves to be analyzed carefully. There are three basic notions of market in the literature: (1) in the textbook theory of exchange, markets exist and are self-evident; and any transaction presupposes the existence of an underlying market; (2) markets as devices for reducing transaction costs (Coase); (3) markets as social institutions promoting division of labor, innovation and economic growth. A major contribution to the discussion of markets comes from Coase whose work clarifies both (1) and (2) above. ‘In mainstream economic theory the firm and the market are for the most part assumed to exist and are not themselves the subject of investigation’ (Coase, 1988, p. 5; italics added). By mainstream economic theory Coase means an economic theory without transaction costs. Transaction costs are the costs of market transactions that include ‘search and information costs, bargaining and decision costs, and policing and enforcement costs’ (Dahlman, 1979, quoted by Coase), which, of course, includes the costs of contracting. In Coase’s theory, transaction costs exist and can be important; and they explain the existence of the firm.3 In the old neoclassical theory of exchange that Coase refers to, the existence of markets (and also the creation of new markets) is assumed but not analyzed. It is an axiom, a self-evident truth, similar to Coase’s criticism of the notion of consumer utility, which is central to the above theory: ‘a non existing entity which plays a part similar, I suspect, to that of ether in the old physics’ (Coase, 1988, p. 2; italics added). This view of markets implies that any transaction assumes an underlying market, or that there is no such thing as a transaction without a market. This is not only not correct but, following Coase or the implications of his analysis, we assert that the distinction between individual transactions and a market is important.4 For our purposes, markets are social institutions where at least a critical mass of producers and a critical mass of consumers interact and transact. There is an important element of collective interaction and of collective transacting; that is, any one transaction takes into account the conditions of all other transactions. From this viewpoint a market contrasts with an institutional context characterized by three relevant conditions. First, it is a lower set of transactions than that of the subsequent market. Second, transactions are isolated and sporadic, both synchronically and diachronically. Third, agents do not rely upon exchanges but on self-sufficiency; that is, users produce the products they consume/use.

106


Originally markets were defined only in geographical terms as locations where a large number of sellers and buyers would meet to trade. Since then, markets have grown into sophisticated institutions characterized by an array of functions and characteristics.5 The extent to which the process has grown differs. Different stratifications of institutional evolution can be found according to the characteristics of products and agents involved (Menard, 2004). Markets differ across countries, industries and contexts. Markets differ according to the functions they can perform and their structural characteristics. The emergence and evolution of markets is the result of a process that takes place over time and is shaped by institutional innovations of different kinds. Towards a Classification of Markets Markets have properties and characteristics. According to such characteristics, markets are more or less able to perform their functions. The properties of markets do not coincide with the properties of the products being exchanged and the characteristics of agents engaged in trade. Yet there is a high degree of overlap between the characteristics of the products and agents and the properties of the markets. The reputation of agents is an essential condition for the emergence and the working of markets. The certification of agents and the ex ante assessment of their reliability and sustainability provide both tentative customers and suppliers with information necessary to perform transactions. Without the provision of information about the reliability of partners in trade, both customers and suppliers must bear the costly burden of relevant search and assessment activities. From the viewpoint of the effective working of the marketplace, moreover, the symmetric distribution of reputation, as a carrier of information, plays a key role. It is clear that in a system where reputation is distributed unevenly, transactions are likely to privilege the few agents that enjoy the advantages of good reputation. A star system is likely to emerge, with clear monopolistic effects. Systems where the reputation of agents is certified are likely to work better than systems where reputation is asymmetrically distributed. The latter systems, in turn, perform better than systems where average levels of reputation are low. Reputation is a key element in the definition of social capital precisely for its positive effects in terms of reduction of transaction costs. Products differ widely with respect to their characteristics, and exhibit different levels of general tradability and hence influence the performances of the corresponding markets with respect to the number and quality of the functions provided to the rest of the system. In this context it is consequently clear that a central property is the


107

category of products that are being exchanged. We can identify markets built around a specific need category or user segment (encompassing many different products and technologies); and markets built around a particular industry or segment of producers (encompassing many user segments and need categories). In the first profile of a market, users of substitute products relating to the satisfaction of a basic category of need converge; in the second, producers of products related to a basic set of technologies converge. In the former market, the products traded are substitutes on the demand side. In the latter market, defined by a particular producer technology category, for example the chemical industry, the products traded are substitutes on the supply side. Beyond the characteristics of the products being exchanged in the marketplace, and of agents engaged in trade, we can identify at least six main characteristics of markets: the time horizon of markets plays a central role. Spot markets are far less effective than regular markets. In effective markets, future prices can be identified and a full intertemporal string of prices and quantities can be set. Market density is defined by the number of agents both on the demand and on the supply side. It is clear that markets with one player either on the demand or the supply side are highly imperfect. Market thickness is relevant both on the demand and the supply side with respect to the volume of transactions. With respect to thickness, there is an important issue about the levels of the critical mass necessary for a good performance of the market. When transactions take place with high levels of frequency, the users of markets, both on the demand and the supply side, and prices and quantities can adjust swiftly to changing economic conditions. Sporadic transactions limit the performances of markets. Recurrence of transactions is most important to reduce opportunistic behavior and to make comparisons possible. Recurrence of transactions is a major source of transparency and hence information. The concentration of transactions increases the density, thickness, frequency and recurrence of transactions: as such it can be enforced by means of compulsory interventions, or emerge as the consequence of a spontaneous process. The role of concentration is vital for the emergence of new effective markets, and hence it is at the same time a prerequisite and a threshold factor. The Functions of Markets Markets differ greatly with respect to their characteristics, and as a consequence with respect to the functions they can perform. A well-functioning market is able to perform a variety of functions that a set of isolated transactions cannot. At least four basic functions can be identified:

108


1

Markets as signaling mechanisms to actual or potential users or suppliers/producers Markets with appropriate levels of thickness and robustness signal to the rest of the economy the need for the specific products being traded; and that the need-satisfying category of good not only exists but is traded and therefore accessible. The signaling involves a qualitative dimension (the ‘need’ and the ‘product class’ satisfying it) and a quantitative dimension reflected in quantities and values purchased and sold. Existence of a market also minimizes volatility and swings concerning persistence of the ‘need’ or possibility of obtaining the good. This is because a market or an industry operating in it is presumably more stable than a single user or a single firm; and a market – compared to a single transaction – provides relative assurance about the possibility of repetitive transactions, purchases or sales, in the future. Signaling existence and persistence of need to be satisfied and product class to be supplied helps any firm/supplier and any user/consumer respectively, actual or potential, to focus his or her search process on the relevant space where the market exists or operates. It also facilitates users’ (producers’) long-run decisions concerning purchase (sale) of a new particular product class or service or system traded in a particular market (‘the product’). The decisions involve investment decisions concerning or involving the product or its supply. Nobody wants to create dependence on a product purchased (sold) whose sources of supply (demand) and mechanisms of purchase (sale) are not highly reliable and stable.6 2 Markets as selection and incentive mechanisms Markets are able to perform relevant screening functions when many different products, manufactured with different technologies, are being confronted. Best products emerge and lower-quality products are screened. The extent to which selection is dynamically efficient depends on characteristics of users, for example on whether or not users are willing or not to take risks in trying novel products. It also depends on characteristics of producers, for example whether they are innovative or not and whether or not competition (as a process) among producers both generates variety and leads individual firms to rapidly adapt and improve their products in response to other firms’ products. Good selection mechanisms enable the allocation of effective incentives to agents, via entry, expansion and invention/innovation, and symmetrically exit when losses emerge both on the demand and the supply side. 3 Markets as coordination mechanisms By means of their signaling functions, markets make possible coordination in the production of complementary products. Specialization


109

of agents in the narrow spectrum of activities where each firm has a competitive advantage can be done by means of efficient markets. This because in the market all the relevant users are present, so that a firm can easily know the potential market for that specific component (or components) in the production of which it enjoys a competitive advantage (it will also save on selling costs). The mechanisms in operation seem to be: signaling and selection with interactive learning. More generally, markets facilitate both specialization and integration by producers. Moreover, markets also provide integration opportunities on the demand side: they facilitate integration and specialization of users that can combine specialized products into more elaborated consumption and usage. 4 Markets as risk management mechanisms By reducing transaction costs and through the enhancement of variety of firms and products, some markets (as opposed to transactions without markets) make possible the distribution of risks across a variety of firms and products. Hence they reduce the risks of opportunistic behavior and information and knowledge asymmetries. Only a few markets can reach all the necessary levels of time horizon, density, thickness, frequency, recurrence and concentration. The analysis of the broad array of characteristics and functions of markets as economic institutions enables the analysis of the emergence of the market as the result of a process of convergent and complementary innovations. Markets emerge and consolidate as specialized institutions.7 From this viewpoint, the emergence of a viable market can be considered the result of an articulated, institutional process that deserves to be analyzed carefully. Markets are social institutions that perform a variety of functions and exhibit different forms, organizations and characteristics. Moreover, markets are a dynamical construct. Hence markets are being created, they emerge, occasionally their performances and functions improve, yet they can decline. In other words, markets evolve (Richter, 2007). In turn, the emergence of new specialized markets has an impact on the economic system. This leads us to appreciate the notion of ‘market’ originally proposed by Adam Smith, namely ‘a device that promotes division of labor, learning/innovation, and economic growth’. An effort to understand the institutional characteristics of markets in a general context seems necessary in order to grasp all the implications of the creation of the new financial markets associated with venture capitalism. The analysis of their emergence should be the center-piece in any theory of economic development nowadays: markets perform a central role not

110


only in the allocation of resources but also in promoting ‘knowledge-based growth’ (De Liso, 2006).

4.

THE EMERGENCE OF NASDAQ FOR VENTURE CAPITALISM

The creation of a surrogate market for knowledge where knowledgeintensive property rights can be traded as financial products can be considered one of the key features and contributions of venture capitalism. The new financial markets specialized in knowledge-intensive property rights are based on a new intermediation form that emerges from the mutual adaptation of different groups of actors both on the supply and the demand side, and with the underlying institutional structure. This has led to a multilayer super-market such as NASDAQ, which enables participants to relate to a large number of markets for individual stocks simultaneously, thereby better coordinating their needs with the capabilities offered. A new market may emerge when a set of previously isolated precursor transactions sparks an emergence process. For this to happen, a number of conditions are required. Frequently these will include pre-emergence processes of interaction and information flow among agents, together with experimentation and learning concerning product characteristics and user/ producer organization and strategy. Emergence may also require a critical mass of precursor transactions both to underpin the above-mentioned interactions, learning and experimental processes, and to enhance the expected ‘benefits’ derived from creating a new market.8 Moreover, the successful emergence of a new market may depend critically on the converging action of agents towards emerging platforms able to provide the required dynamic coordination (Richardson, 1972, 1998). The evolutionary process leading to the emergence of a new market is seen as an autocatalytic, cumulative process with positive feedback, or, alternatively, a process characterized by dynamic economies of scale. This process involves the creation and utilization of externalities that explain the acceleration of growth. The cumulative process does not end with creation of the new market; rather it continues afterwards at least for a time (provided that external conditions do not deteriorate).9 The new (more complex) structure created by the interaction among elementary components (firms and users) will, once it is emerged, positively further stimulate such components. This phenomenon provides us with an additional, and much less recognized, characteristic of ‘a market’: once created it will stimulate the creation of new firms.10


111

The Phases of the Process The emergence of the new financial markets is the result of a continued process of convergent and complementary steps that can be visualized as comprising four phases. Phase I. Bundling finance and competence with innovation Since the early days, venture capital firms specialized in the provision of ‘equity finance’ to new science-based startup companies as distinct from ‘loans’, which were the prevailing product offered by existing financial institutions (banks). Equity finance was offered to science-based startup companies bundled together with business services and management advice, management services, certification and networking functions. This was exchanged for limited partnership. Limited partnership is a key ‘precursor’ dimension to the emergence of the new market. In the USA during the 1960s and 1970s, limited partnerships were the dominant form of organization for new science-based startup companies. Limited partnership allowed for dilution of founder equity positions and a capital (jointly with the prevailing product) market orientation. Gompers and Lerner (1999) stress the role at this stage of the changing features of intellectual property right regimes. The increasing depth, width and duration of patents has led to higher levels of appropriability for knowledge that is embodied in new science-based companies, and traded in the form of knowledge-intensive property rights rather than bundled within large diversified incumbents. Large incumbents were able to rely much less on the protection provided by intellectual property rights because of the advantages of existing barriers to entry that would delay the dissipation of innovation rents. Large incumbents, moreover, can take advantage of lead times and secrecy as effective mechanisms of knowledge appropriation. New science-based startup companies, on the other hand, need to disclose information about the advantages of their knowledge base: patents perform a key signaling function. The protection of hard intellectual property right regimes is much more important for science-based startup companies that are newcomers themselves. The radical changes in intellectual property right regimes introduced in the 1980s and 1990s clearly favored venture capitalism because they reduced for investors the levels of risks associated with the non-appropriability of the strong knowledge component of the intangible assets of the new science based firms (Hussinger, 2006). Phase II. Knowledge-intensive property rights Phase II is marked by the evolution of limited partnership as the leading form of organization of startup into private stock companies based upon

112


knowledge-intensive property rights shares of the new science-based startup companies and other rights concerning the management of the company. Limited partnership converges progressively into stock-holding. The personal participation of partners in the startup declines and is substituted by the professional services of managers organized by venture capital companies. The new bundling of equity with managerial competence into knowledge-intensive property rights of science-based startup companies that can be traded can be considered the dominant (product) design that lies at the origin of what will become a new market. In this early phase, venture capital companies co-evolve with the organization of the new science-based startup companies. The development of venture capital companies and the growth of the syndications as a way to collect funds for new science-based startup companies have played a key role in this phase. Private investors and financial companies that had contributed to the fundraising activities for new companies were eager to elaborate exit strategies for collecting the value of the new firms after their creation and growth, and participate fully in the profits of the ‘blockbusters’. The search for ‘exit’ strategies acts as a powerful dynamic factor at this stage. Phase III. Trading knowledge-intensive property rights in private markets Exits took place principally through the sale of knowledge-intensive property rights in the so-called trade sales to individuals or organizations. These are ‘private transactions’. During the first half of the 1970s we can observe the growing number of over-the-counter (OTC) initial offerings of knowledge-intensive property rights. Here a critical mass of transactions slowly builds up and triggers, through variation, a more systematic and focused search and experimentation process leading to the emergence of a public market. Large companies become progressively aware of the important opportunities provided by the new small public companies whose shares are traded over the counter as a source of technological knowledge. Mergers and acquisitions increase as corporations rely more and more systematically upon the takeover of the new science-based companies, after initial public offering (IPO), as a source of technological knowledge that has already been tested and proved to be effective. The acquisition of external knowledge, embodied in the new firms, complements and partly substitutes internal activities conducted intra muros within the traditional research laboratories. Specifically, incumbents rely on the new source of external technological knowledge as an intermediary input that can be combined with other internal knowledge sources. Hence it is clear that the new, dedicated financial markets implement


113

a new central functionality in the economic system in terms of increased division of labor in the generation of new technological knowledge, and higher levels of specialization in the production of the bits of knowledge that each company is better able to command. From this viewpoint it is also clear that the new markets favor the coordination among different firms specialized in the generation of complementary modules of knowledge that can be exchanged and traded. The new financial markets favor the reorganization of the generation of knowledge, away from high levels of internal vertical integration, towards open innovation architectures (Chesbrough, 2003). The changing organization of the generation of technological knowledge on the new financial markets attracts increasing flows of firms on the demand side. Consequently, the growing demand of the new knowledge-intensive property rights by large incumbents increases the frequency of transactions and hence the thickness of the new markets (Avnimelech and Teubal, 2004, 2006, 2008). Phase IV. Emergence of a public capital market focused on IPOs The increasing size of OTC exchanges led the National Association of Dealers to introduce an automatic quoting mechanism to report the prices and quantities of the private transactions. Eventually the mechanisms, better known by an acronym, evolved into a marketplace. NASDAQ became a new market for selling knowledge-intensive property rights to the public at large rather than only to private individuals or organizations. NASDAQ became the specialized market for IPOs of the shares of the new science-based startup companies nurtured by venture capital companies and funded with their assistance by groups of financial investors. Significant adaptations of the institutional environment, for example modifications of the ERISA (Employment Retirement Income Security Act), including the 1979 amendment to the ‘prudent man’ rule governing pension fund investments in the USA (Gompers and Lerner, 2004, pp. 8, 9) involved liberalization of the constraints on pension fund investment in the stock of new science-based startups. In parallel, the increasing liberalization of international financial and currency markets had the twin effect of increasing both the demand and the supply in the NASDAQ. On the demand side, a growing number of investment funds entered the NASDAQ to place their capital. On the supply side, the high levels of liquidity, the thickness of transactions and the low levels of volatility, together with the high quality of the professional services available in NASDAQ, attracted the entry of venture capital companies of other countries (in the Israeli case the dynamics is impressive) that eventually represented a large and growing share of the total figure of IPOs of science-based startup companies. An increasing

114


concentration of exchanges, a key feature of a marketplace, has been taking place at the global level (Bozkaya et al. 2008). By means of global concentration, sparse, rare and occasional transactions by a myriad of isolated and dispersed agents, scattered around many local markets, were progressively brought into the same physical and institutional context with clear advantages in terms of the number of transactions that occur and hence can be compared and observed. Here the analysis of Schmookler (1966) on the role of demand in pulling technological innovation applies to explain the final stages of this process of institutional change. Schmookler found strong empirical evidence of a link between capital-good market size (as indicated by gross investment) on the one hand and capital-good improvement inventions (as indicated by patents on capital goods, with a lag) on the other (Schmookler, 1966). Moreover, when it comes to explain the distribution of patents on capitalgoods improvement inventions across industries, ‘demand’ overrides any differences in the ‘supply’ side of inventions. His analysis suggests that the emergence of new product markets in general and not only capital-goods markets will, through a ‘demand’ effect, induce improvement inventions in the underlying product and process technology. Here it is clear that demand for the new knowledge-intensive property rights by investment funds, pension funds and eventually family pulled the final diffusion of NASDAQ with a snowball effect in terms of the overall level of transactions. The new levels of mass transactions favored the frequency of IPOs and attracted qualified professional and financial companies specialized in market management. This in turn led to substantial increase in the thickness of the markets, reduction in volatility and eventually global concentration of exchanges. The concentration of transactions, the thickness of the new markets, and, most important, the ensuing recurrence of transactions on individual stocks have important effects in terms of reduction of volatility. The entry on the demand side of large investment funds, pension funds and ultimately even private investors has the important effect of providing large flows of transactions on the shares of individual companies. The size of the new financial markets makes it possible to better manage uncertainty by means of the distribution of small bets across a variety of actors and of firm-specific equity markets. In the previous phases, characterized by the preponderance, on the demand side, of large incumbents searching for new science-based companies able to complement their internal knowledge base in order to organize takeovers and subsequent delisting, transactions on individual stocks were sporadic, with high levels of volatility. This enables NASDAQ to become an efficient mechanism for the


115

identification of the correct value of knowledge-intensive property rights. This in turn leads it to perform the key function of appreciation of the large share of intangible assets in the value of the new science-based companies (Campart and Pfister, 2007; Bloch, 2008). The expansion/transformation of NASDAQ is clearly the result of a cumulative process with positive feedback involving a number of processes that make the market more and more attractive to increasingly larger sets of agents (both demand side and supply side). The reasons are similar to some extent to those invoked to explain the dynamics of venture capital or cluster emergence. The new sets of agents that participate in the new markets include specialized agents providing services to investors or companies, for example investment banks, brokers, consultants, and so on; specialized new intermediaries such as venture capital/ private equity funds, financial investors and so on. The enhanced volume that their entry induces further reduces transaction costs, which further increases the thickness and frequency of transactions. This also reduces uncertainty to individual investors as well as market volatility, and so on. Thus, once a new market emerges (e.g. as a result of venture capitalism) and begins to grow, a point may be reached when the private ‘benefit’ from developing a disruptive technology may become such as to induce ‘technology suppliers’ like science-based startup companies to undertake disruptive technology development. This in turn enabled exploitation of significant economies of scale and scope, and a momentum for further expansion (dynamic economies or cumulative processes with positive feedback). NASDAQ thereby eventually became the market for transactions on knowledge-intensive property rights in general. NASDAQ in effect became a ‘super-market’ for products generating income streams for the general public. The emergence of venture capitalism, defined as the combination of venture capital companies able to screen, fund and assist the growth of new science-based startup companies complemented by a dedicated financial market specialized in the transactions of their property rights, marks important progress in knowledge governance. Venture capitalism has significant advantages with respect to the system architecture prevailing in the second part of the twentieth century, when innovations were mainly selected, developed and commercialized by existing incumbent companies. The new, dedicated financial markets seem better able than the previous knowledge governance mechanisms to appreciate the economic value of technological knowledge, to signal the new directions of technological change, to select the new blueprints and, most important, to provide better incentives respectively to ‘inventors’, to venture capital firms and to

116


investors in directing their resources and capabilities towards the generation and use of new technological knowledge. The new, dedicated financial markets seem able to reduce the limitations of both the hierarchical corporate and the credit-based polyarchic model based upon the banking system. They also seem able to combine the advantages of screening radical innovations of polyarchic decisionmaking with the advantages stemming from direct participation in the profits of new outperforming science-based startup that are characteristic of the equity provision of finance to innovation, typical of the corporate model.

5.

CONCLUSIONS

Venture capitalism can be understood as a new mechanism for the governance of technological knowledge that is the result of a system dynamics where a variety of complementary and localized innovations introduced by heterogeneous agents aligned and converged towards a collective platform. The new mechanism has improved the governance of technological knowledge within economic systems, through the combination of new science-based startups and new, dedicated financial markets specialized in the transactions of knowledge-intensive property rights. Hence it has reshaped the prime mechanism by which the generation of new knowledge can lead to economic growth. The relationship between technological and institutional change is strong and allows for bidirectional causality. Technological change can be considered the cause of institutional change, as much as institutional change can be considered as the origin of technological change. A large literature has explored the view that the discontinuities brought about by the radical technological breakthrough that took place in the late 1970s with the emergence of the new technological systems based upon information and communication technologies (ICT) can be thought to be at the origin of the progressive demise of the Chandlerian model of innovation centered on large corporations. Venture capitalism has been often portrayed as the consequence of the ICT revolution. In this chapter we have articulated the alternative hypothesis. The emergence of venture capitalism based upon new dedicated financial markets specialized in the trading of knowledge-intensive property rights and hence in the systematic appreciation of new science-based startups can be considered a major institutional innovation in the governance of technological knowledge and as such a key factor in hastening the pace of introduction of more radical technological innovations.


117

The analysis has highlighted the advantages of the new mechanism of knowledge governance based upon venture capital companies able to screen, fund and implement new science-based startup companies and new dedicated financial markets specialized in knowledge-intensive property rights. It has also shown how the emergence of such new markets has been the result of a complex process of system dynamics where a plurality of actors and interests aligned and converged towards a common platform able to integrate and valorize the complementarities between their different profit functions. The emergence of the new financial markets can be considered as a major institutional innovation that is likely to have important effects on the pace of technological change. Following our line of investigation, we can summarize the main reasons why the process of transformation of radical inventions into new product markets is likely to become more certain, frequent and routinized under venture capitalism: (1) increased numbers of new science-based startup companies with radical inventions; (2) new systemic and generic mechanisms of direct or indirect transformation of such inventions into new product markets; (3) the effect of new markets and more rapid market growth on invention, including radical (both disruptive and nondisruptive) inventions; (4) the possible emergence of unbundled markets for technological improvements. Venture capitalism creates a cumulative process of innovation-based economic growth. The combination of continued generation of new opportunities and the mechanism for ‘unlocking’ the system from potential, strong past dependence, is evidence that venture capitalism could become a feature of sustainable innovation-based growth.

NOTES 1.

2. 3.

Morris Teubal acknowledges the funding and support of ICER (International Center for Economic Research) where he was a Fellow in 2005 and 2008 and the Prime (NoE) Venture Fun Project. Preliminary versions have been presented at the Fifth Triple Helix Conference ‘The capitalization of knowledge: cognitive, economic, social and cultural aspects’ organized in Turin by the Fondazione Rosselli, May 2005 and the following workshops: ‘The emergence of markets and their architecture’, jointly organized by CRIC (University of Manchester) and CEPN-IIDE (University Paris 13) in Paris, May 2006; ‘Instituting the market process: innovation, market architectures and market dynamics’ held at the CRIC of the University of Manchester, December 2006; ‘Search regimes and knowledge based markets’ organized by the CEPN Centre d’Economie de Paris Nord at the MSH Paris Nord, February 2008. So far, this contribution complements and integrates Antonelli and Teubal (2008), which focuses on the emergence of knowledge-intensive property rights. Concerning the nature and function of markets, again following Coase: ‘Markets are institutions that exist to facilitate exchange, that is they exist in order to reduce the

118

4.

5.

6. 7.

8. 9.

10.

The capitalization of knowledge cost of carrying out exchange transactions. In Economic Theory which assumes that transaction costs are non-existent markets have no function to perform’ (Coase, 1988, p. 7); and ‘when economists do speak about market structure, it has nothing to do with markets as an institution, but refers to such things as the number of firms, product differentiation and the like, the influence of the social institutions that facilitate exchange being completely ignored’. Coase (1988) discusses the elements comprising a market, e.g. the medieval fairs and markets that comprise both physical facilities and legal rules governing the rights and duties of those carrying out transactions. Modern markets will also involve collective organizations, that is technological institutes and mechanisms for the provision of market-specific public goods. They also require a critical mass of buyers and sellers, and institutions assuring standards and quality on the one hand and transparency of transactions and inter-agent information flow on the other. Marshall makes it clear that markets are themselves the product of a dynamic process: ‘Originally a market was a public place in a town where provisions and other objects were exposed for sale; but the word has been generalized, so as to mean any body of persons who are in intimate business relations and carry on extensive transactions in any commodity. A great city may contain as many markets as there are important branches of trade, and these markets may or may not be localized. The central point of a market is the public exchange, mart or auction rooms, where the traders agree to meet and transact business. In London the Stock Market, the Corn Market, the Coal Market, the Sugar Market, and many others are distinctly localized; in Manchester the Cotton Market, the Cotton Waste Market, and others. But this distinction of locality is not necessary. The traders may be spread over a whole town, or region of country, and yet make a market, if they are, by means of fairs, meetings, published price lists, the post-office or otherwise, in close communication with each other’ (Marshall, 1920, pp. 324–5). Markets can also signal new product or product feature requirements (‘unmet needs’) within the ‘product category’ being traded. Our agenda is therefore not only to define and explain the role of markets but also to identify the processes of emergence of new markets. This will include analyzing the conditions under which a set of ‘precursor’ transactions will not lead to the emergence of a new market. In terms of system dynamics, this could be termed ‘left-hand truncation’. Moreover, explaining emergence will require making reference to other variables, that is scale economies in building the marketplace (Antonelli and Teubal, 2008). The benefits include savings in transaction costs that should cover the fixed costs of creating and the variable costs of operating a new market (see above). The above framework suggests that failed market emergence could be the result of two general causes. One is failed selection processes resulting from too little search/experimentation and/or inappropriate selection mechanisms due to institutional rigidity. The other is failure to spark or sustain an evolutionary cumulative emergence process (e.g. due to system failures that policy has not addressed). Not all radical inventions, even those leading to innovations and having potential, will automatically lead to new product markets. Students of regional high-tech clusters such as Saxenian (1994) and Fornahl and Menzel (2004) have intuitively recognized the relevance of such dynamics, but not quite elaborated it.

REFERENCES Antonelli, C. (2008), Localized Technological Change: Towards the Economics of Complexity, London: Routledge.


119

Antonelli, C. and M. Teubal (2008), ‘Knowledge intensive property rights and the evolution of venture capitalism’, Journal of Institutional Economics, 4, 163–82. Arrow, K. (1962), ‘Economic welfare and the allocation of resources to invention’, in R.R. Nelson (ed.), The Rate and Direction of Inventive Activity: Economic and Special Factors, Princeton, NJ: Princeton University Press for the National Bureau of Economic Research, pp. 609–25. Avnimelech, G. and M. Teubal (2004), ‘Venture capital start-up co-evolution and the emergence and development of Israel’s new high tech cluster’, Economics of Innovation and New Technology, 13, 33–60. Avnimelech, G. and M. Teubal (2006), ‘Creating venture capital industries that coevolve with high tech: Insights from an extended industry life cycle perspective of the Israeli experience’, Research Policy, 35, 1477–98. Avnimelech, G. and M. Teubal (2008), ‘From direct support of business sector R&D/innovation to targeting of venture capital/private equity: a catching up innovation and technology policy cycle perspective’, Economics of Innovation and New Technology, 17, 153–72. Bloch, C. (2008), ‘The market valuation of knowledge assets’, Economics of Innovation and New Technology, 17, 269–84. Bozkaya, A. and Bruno Van Pottelsberghe De La Potterie (2008), ‘Who funds technology-based small firms? Evidence from Belgium’, Economics of Innovation and New Technology, 17, 97–122. Campart, S. and E. Pfister (2007), ‘Technology cooperation and stock market value: an event study of new partnership announcements in the biotechnology and pharmaceutical industries’, Economics of Innovation and New Technology, 17, 31–49. Chesbrough, H. (2003), Open Innovation. The New Imperative for Creating and Profiting from Technology, Boston, MA: Harvard Business School Press. Coase, R. (1988), The Firm, the Market and the Law, Chicago, IL: The University of Chicago Press. Dahlman, C.J. (1979), ‘The problem of externality’, Journal of Law and Economics, 22, 141–62. De Liso, N. (2006), ‘Charles Babbage, technological change and the “National System of Innovation”’, Journal of Institutional and Theoretical Economics, 162, 470–85. Fornahl, D. and M.P. Menzel (2004), ‘Co-development of firms founding and regional cluster’, Discussion Paper No. 284, University of Hanover, Faculty of Economics. Gompers, P. and J. Lerner (1999), The Venture Capital Cycle, Cambridge, MA: The MIT Press. Gompers, P. and J. Lerner (2004), The Venture Capital Cycle, 2nd edn, Cambridge, MA: The MIT Press. Hall, B.H. (2002), ‘The financing of research and development’, Oxford Review of Economic Policy, 18, 35–51. Hodgson, G.M. (1998), ‘The approach to institutional economics’, Journal of Economic Literature, 36, 166–92. Hussinger, K. (2006), ‘Is silence golden? Patents versus secrecy at the firm level’, Economics of Innovation and New Technology, 15, 735–52. King, R.G. and R. Levine (1993), ‘Finance and growth: Schumpeter might be right’, Quarterly Journal of Economics, 107, 717–37.

120


Lane, D.A. (1993), ‘Artificial worlds and economics, part II’, Journal of Evolutionary Economics, 3, 177–97. Lane, D.A. and R.R. Maxfield (2005), ‘Ontological uncertainty and innovation’, Journal of Evolutionary Economics, 15, 3–50. Lane, D.A., S.E. van Der Leeuw, A. Pumain and G. West (eds.) (2009), Complexity Perspectives in Innovation and Social Change, Berlin: Springer, pp. 1–493. Marshall, A. (1890), Principles of Economics, London: Macmillan (8th edn, 1920). Menard, C. (ed.) (2000), Institutions Contracts and Organizations. Perspectives from New Institutional Economics, Cheltenham, UK and Northampton, MA, USA: Edward Elgar. Menard, C. (2004), ‘The economics of hybrid organizations’, Journal of Institutional and Theoretical Economics, 160, 345–76. Menard, C. and M.M. Shirley (eds) (2005), Handbook of New Institutional Economics, Dordrecht: Springer. Nelson, R.R. (1994), ‘The co-evolution of technology, industrial structure and supporting institutions’, Industrial and Corporate Change, 3, 47–63. Nelson, R.R. (1995), ‘Recent evolutionary theorizing about economic change’, Journal of Economic Literature, 23, 48–90. Perez, C. (2003), Technological Revolutions and Financial Capital: The Dynamics of Bubbles and Golden Ages, Cheltenham, UK and Northampton, MA, USA: Edward Elgar. Quéré, M. (2004), ‘National systems of innovation and national systems of governance: a missing link?’, Economics of Innovation and New Technology, 13, 77–90. Richardson, G.B. (1972), ‘The organization of industry’, Economic Journal, 82, 883–96. Richardson, G.B. (1998), The Economics of Imperfect Knowledge, Cheltenham, UK and Northampton, MA, USA: Edward Elgar. Richter R. (2007), ‘The market as organization’, Journal of Institutional and Theoretical Economics, 163, 483–92. Sah, R.K. and J.E. Stiglitz (1986), ‘The architecture of economic systems’, American Economic Review, 76, 716–27. Sah, R.K. and J.E. Stiglitz (1988), ‘Committees, hierarchies and polyarchies’, Economic Journal, 98, 451–70. Saxenian, A. (1994), Regional Development: Silicon Valley and Route 128, Cambridge, MA: Harvard University Press. Schmookler, J. (1966), Invention and Economic Growth, Cambridge, MA: Harvard University Press. Schumpeter, J.A. (1934), The Theory of Economic Development, Cambridge, MA: Harvard University Press. Schumpeter, J.A. (1942), Capitalism, Socialism and Democracy, New York: Harper and Brothers. Stiglitz, J.E. (1985), ‘Credit markets and capital control’, Journal of Money, Credit and Banking,), 133–52. Stiglitz, J.E. and A. Weiss (1981), ‘Credit rationing in markets with imperfect information’, American Economic Review, 71, 912–27.

4.

How much should society fuel the greed of innovators? On the relations between appropriability, opportunities and rates of innovation Giovanni Dosi, Luigi Marengo and Corrado Pasquali

1.

INTRODUCTION

This chapter attempts a critical assessment of both theory and empirical evidence on the role and consequences of the various modes of appropriation, with particular emphasis on intellectual property rights (IPR), as incentives for technological innovation. That profit-motivated innovators are fundamental drivers of the ‘unbound Prometheus’ of modern capitalism (Landes, 1969) has been well appreciated since Smith, Marx and, later, Schumpeter. For a long time such an acknowledgment has come as an almost self-evident ‘stylized fact’. Finer concerns about the determinants of the propensity to innovate by entrepreneurs and business firms came much later with the identification of a potentially quite general trade-off underlying the economic exploitation of technological knowledge: in so far as the latter is a nonrival and hardly excludable quasi-public good, pure competitive markets are unable to generate a stream of quasi-rents sufficient to motivate profitseeking firms to invest resources in its production (Arrow, 1962). In order to provide such incentives, a general condition is to depart from pure competition (as was indeed quite naturally acknowledged by Smith, Marx and Schumpeter). Granted that, however, what is empirically the extent of such a departure? And, from a normative point of view, what is the desirable degree of appropriability able to fuel a sustained flow of innovations undertaken by business firms? And through which mechanisms? Moreover, what is the impact of different institutional and technological

121

122


conditions upon the profitability and competitive success of innovators themselves? The last angle is the one tackled in the seminal paper of David Teece (1986), who argues that profits from innovation depend upon the interaction of three families of factors, namely, appropriability regimes, complementary assets and the presence or absence of a dominant paradigm. Note that appropriability conditions, in addition to patent and copyright protection, include secrecy, lead times, costs and time required for duplication, learning, sales and service assets. Moreover, as Teece emphasizes, such appropriability regimes are largely dictated by the nature of technological knowledge (Teece, 1986, p. 287). These fundamental observations on the mechanisms through which firms ‘benefit from innovation’, however, have been lost in a good deal of contemporary literature on the incentive to innovate, wherein, first, appropriability conditions are reduced almost exclusively to IPR regimes, and, second, the award of IPR themselves is theoretically rooted in a framework – in our view deeply misleading – namely that of ‘market failures’. In what follows, we start from a critical assessment of such a perspective and of the related notion of a monotonic relation between IP protection and rates of innovation (Section 2). Next, after an overview of the recent changes in IPR regimes (Section 3), in Section 4 we review the empirical evidence on the relationship between appropriability in general and IP protection in particular, on the one hand, and rates of innovation on the other. Such evidence, we shall argue, suggests that, first, appropriability conditions are just one of several factors (possibly second-order ones) shaping the propensity to innovate. Together, the relative importance of the various factors and their interaction is highly sector- and technologyspecific. Second, appropriability is likely to display a threshold effect, meaning that a minimum degree of appropriability is necessary to motivate innovative effort, but above such a threshold further strengthening of appropriability conditions will not determine further increases of R&D investments and rates of innovation. Rather, social inefficiencies such as ‘anti-commons’ effects (Heller and Eisenberg, 1998), rent-seeking behaviors, dissipation of quasi-rents into litigation and so on are much more likely to emerge. Third, and relatedly, there seems to be no clear evidence of a positive relation between the tightening of IPR regimes and the rates of innovation. Conversely, there is good evidence on the (perverse) links between IPR protection and income distribution. The rates of innovation, we suggest, fundamentally depend on paradigm-

Appropriability, opportunities and innovation

123

specific opportunities rather than on mere appropriability conditions (at least above some threshold) and even less so on the specific subset of appropriability devices represented by legal IPR protection. Note that observed rates of innovation at the level of an industry or an economy are only remotely related to any ‘equilibrium’ rate of R&D investment by the ‘representative’ firm, whatever that means. Given whatever incentive profile, one typically observes quite varied search responses (as very roughly measured by R&D investments) and also quite different technological and economic outcomes, well beyond what a statistician would interpret as independent realizations of the same underlying random process. We thus conclude (Section 5) that while the first-order determinants of the rates of innovation rest within the technology-specific and sector-specific opportunity conditions, the differential ability of individual firms to benefit from them economically stem from idiosyncratic organizational capabilities. But if this is the case, the answer to the question we ask in the title of this chapter is also straightforward: fueling the greed of innovators might be at best irrelevant for the ensuing rates of innovation, while of course bad from a social point of view.

2.

SOME FAILURES OF THE ‘MARKET FAILURE’ ARGUMENTS

The economic foundations of both theory and practice of IPR rest upon a standard market failure argument, without any explicit consideration of the characteristics of the knowledge whose appropriation should be granted by patent or other forms of legal monopoly. The proposition that a positive and uniform relation exists between innovation and intensity of IP protection in the form of legally enforced rights such as patents holds only relative to a specific (and highly disputable) representation of markets, their functioning and their ‘failures’, on the one hand, and of knowledge and its nature on the other. The argument falls within the realm of the standard ‘Coasian’ positive externality problem, which can be briefly stated in the following way. There exists a normative set of efficiency conditions under which markets perfectly fulfill their role of purely allocative mechanisms. The lack of externalities is one of such conditions because their appearance amounts (as with positive externalities) to underinvestment and underproduction of those goods involved in the externality itself. Facing any departure from efficiency conditions, a set of policies and institutional devices must be put in place with the aim of re-establishing them in order to

124


achieve social efficiency. Knowledge generation is one of the loci entailing such an externality: since knowledge is (to a great extent) a public good, it will be underproduced and will receive insufficient investments. Hence an artificial scarcity is created to amend non-rivalry and non-excludability in its use, yielding an appropriate degree of appropriability of returns from investments in its production. The core of the matter then becomes one of balancing out the detrimental effect of the deadweight loss implied by a legally enforced monopoly, on the one hand, and the beneficial effect of investments in R&D and more generally in knowledge generation, on the other. A number of general considerations can be made about this argument. First, the argument rests fundamentally on the existence of a theoretical (but hardly relevant in terms of empirical and descriptive adequacy) benchmark of efficiency against which policy and institutional interventions should be compared as to their necessity and efficacy. Second, the efficiency notion employed is a strict notion of static efficiency that brings with it the idea that markets do nothing except (more or less efficiently) allocate resources. Third, a most clear-cut distinction between market and non-market realms is assumed, together with the idea that non-market (policy, institutional) interventions can re-establish perfect competition using purely market-based ‘tools’. Fourth, it is assumed that the nature of ‘knowledge’ is totally captured by the notion of ‘information’, thus setting the possibility of treating it institutionally in uniform ways, neglecting any dimension of knowledge that relates to its ‘non public good’ features. According to this perspective, the transformation of the public-good ‘knowledge’ in the private-good ‘patent’ will perfectly set incentives for its production by way of legally enforced conditions and possibilities of appropriability. However, if one starts questioning that markets solely allocate resources, one may begin to consider them as performing a wider set of activities such as being the places in which ‘novelty’ is (imperfectly) produced, (imperfectly) tested and (imperfectly) selected. In this alternative perspective, it becomes hard to reduce any efficiency consideration to static efficiency, so that, for instance, it is not necessarily true that allocative patterns that are efficient from a static perspective have the same property from a dynamical point of view. It thus follows that the institutional attribution of property rights (whether efficient or not in a static allocative perspective) may strongly influence the patterns of technological evolution in directions that are not necessarily optimal or even desirable. In this sense, any question about the appropriate level of IP protection and degree of appropriability would be better grounded in a theory of


125

innovative opportunities and productive knowledge (issues on which the theory of allocative efficiency is rather silent: see Winter, 1982 and Stiglitz, 1994 from different angles). In addition, viewing markets as embedded and depending upon a whole ensemble of non-market institutions allows us to appreciate the fact that technological innovation is highly dependent on a variety of complementary institutions (e.g. public agencies, public policies, universities, communities and of course corporate organizations with their rich inner structure) that can hardly be called ‘markets’ and hardly be regulated by pure market incentives. Precisely this institutional embeddedness of innovative activities makes it very unlikely that a ‘market failure’ approach such as the one we sketched above could provide any satisfactory account of the relationship between appropriability and propensity to innovate. Finally, the (misleading) identification of knowledge with information (i.e. the deletion of any reference to cognitive and procedural devices whose role is to transform sheer information into ‘useful knowledge’ and which are to a large extent tacit and embedded in organizations) makes one forget that processes through which new knowledge is generated are strongly dependent on the specificities of each technological paradigm (which can hardly be reduced to ‘information’ categories). One question that seems to be rarely asked (and answered) in precise terms is: what is (if any) the increase in the value of an innovation realized by way of patenting it? A straightforward answer to this question would be: in a perfectly competitive market, any innovation has no value (i.e. its price equals zero) as its marginal cost of reproduction equals zero. As a consequence, the whole and sole value of an innovation comes from its being patented. On this perspective, one is forced to conclude that a straightforward positive relation exists between innovative activities and patents: a relation in which patents are the one and only source of value of technological innovations (given perfect competition). That is, in Teece’s words, patents would be the only way of ‘profiting from technological innovation’. Under more careful scrutiny, however, this argument is subject to a series of limitations and counter-examples. A first class of counterarguments arises from the many instances of innovations that, in spite of not being patented (or patented under very weak patent regimes), have most definitely produced considerable streams of economic value. Relevant examples can be drawn from those technologies forming the core of ICT (information and communication technology). For instance, the transistor, while being patented from Bell Labs, was liberally licensed also as a consequence of antitrust litigation and pressure from the US Justice Department: its early producers nonetheless obtained

126


enough revenue to be the seeds of the emergence of a whole industry (Grandstrand, 2005). The early growth of the semiconductor industry had been driven to a great extent by public procurement in a weak IP regime. The software industry, certainly a quite profitable one, similarly emerged under a weak IP regime. The telecom industry was, until the 1990s largely operated by national monopolies that were also undertaking a good deal of research, and IPR played little role in the rapid advance of technology in this industry. Mobile telephony also emerged under a weak IP regime (until the late 1980s). We suggest indeed that strong IPR did not play a pivotal role either in the emergence of ICT or as a means of value generation. On the contrary, in the early stage of those sectors it might have been the very weakness of the patent regime that spurred their rapid growth. Conversely, the strengthening of the IP regime in recent years (soon after the ICT boom in the late 1980s) might well have been (in terms of political influence) a consequence rather than a cause of the fast pace at which the ICT sector expanded. Back to our opening question, it is worth noting how (some) economists have been at least cautious with respect to the adoption of the patent system as the only means to foster innovative activity and to its uniform effectiveness. As Machlup (1958, p. 80) put it: ‘If we did not have a patent system, it would be irresponsible, on the basis of our present knowledge of its economic consequences, to recommend instituting one. But since we have had a patent system for a long time, it would be irresponsible, on the basis of our present knowledge, to recommend abolishing it.’ Similar doubts are expressed in David (1993, 2002), who argues that IPR are not necessary for new technologies and suggests that different institutional mechanisms more similar to open science might work more efficiently. Of course, the cautious economist is well aware that even from a purely theoretical point of view, the innovation/patent relation is by no means a simple one. And similarly tricky from a policy point of view is the identification of balance between gains and losses of any system of IP protection. As a matter of fact, on the one hand it may be argued that IP monopolies afforded by patents or copyright raise prices above unit production costs, thus diminishing the benefits that consumers derive from using protected innovations. On the other hand, the standard argument claims that the same rights provide a significant incentive to produce new knowledge through costly investments in innovative research. However, such a purported trade-off might well apply also at the micro level. Whether or not a firm has the profitability of its own innovations


127

secured by IPR, its R&D behavior and its IPR enforcement strategies cannot be unaffected by the actions of other firms acquiring and exploiting their own IPR. The effect of firms exploiting IPR invariably raises the costs that other firms incur when trying to access and utilize existing knowledge. Similar dilemmas apply to the effects of a strong IP system on competition process. Static measures of competition may decrease when a monopoly right is granted but dynamic measures could possibly increase if this right facilitates entry into an industry by new and innovative firms. Are these trade-offs general features of the relationship between static allocative efficiency and dynamic/innovative efficiency? There are good reasons to think that such trade-offs might not theoretically even appear in an evolutionary world, as Winter (1993) shows. On the grounds of a simple evolutionary model of innovation and imitation, Winter (1993) compares the properties of the dynamics of a simulated industry with and without patent protection to the innovators. The results show that, first, under the patent regime the total surplus (i.e. the total discovered present value of consumers’ and producers’ surplus) is lower than under the non-patent one. Second, and even more interesting, the non-patent regime yields significantly higher total investment in R&D and displays higher best-practice productivity. More generally, an evolutionary interpretation of the relation between appropriability and innovation is based on the premise that no model of invention and innovation and no answer to the patent policy question is possible without a reasonable account of inventive and innovative opportunities and their nature. The notion of technological paradigm (Dosi, 1982), in this respect, is precisely an attempt to account for the nature of innovative activities. There are a few ideas associated with the notion of paradigm worth recalling here. First, note that any satisfactory description of ‘what technology is’ and how it changes must also embody the representation of the specific forms of knowledge on which a particular activity is based and cannot be reduced to a set of well-defined blueprints. It primarily concerns problem-solving activities involving – to varying degrees – also tacit forms of knowledge embodied in individuals and in organizational procedures. Second, paradigms entail specific heuristics and visions on ‘how to do things’ and how to improve them, often shared by the community of practitioners in each particular activity (engineers, firms, technical societies and so on), that is, they entail collectively shared cognitive frames. Third, paradigms often also define basic templates of artifacts and systems, which over time are progressively modified and improved. These basic artifacts can also be described in terms of some fundamental technological and

128


economic characteristics. For example, in the case of an airplane, their basic attributes are described not only and obviously in terms of inputs and production costs, but also on the basis of some salient technological features such as wing-load, take-off weight, speed, distance it can cover etc. What is interesting here is that technical progress seems to display patterns and invariances in terms of these product characteristics. Hence the notion of technological trajectories associated with the progressive realization of the innovative opportunities underlying each paradigm. In turn, one of the fundamental implications of the existence of such trajectories is that each particular body of knowledge (each paradigm) shapes and constrains the rates and direction of technical change, in a first rough approximation, irrespective of market inducements, and thus also irrespective of appropriability conditions.

3.

THE GROWTH IN PATENTING RATES AND THE (MIS)USES OF PATENT PROTECTION

Needless to say, such a lack of any robust theory-backed relation between appropriability (and even less IPR forms of appropriability) and rates of innovation puts the burden of proof upon the actual empirical record. Indeed, the past two decades have witnessed the broadening of the patenting domain, including the application of ‘property’, to scientific research and its results. This has been associated with an unprecedented increase in patenting rates. Between 1988 and 2000, patent applications from US corporations have more than doubled. The relation between the two phenomena, however, and – even more important – their economic implications are subject to significant controversy (for discussion, see Kortum and Lerner, 1998; Hall, 2005; Lerner, 2002; Jaffe and Lerner, 2004; and Jaffe, 2000). A first hypothesis is that the observed ‘patent explosion’ has been linked to an analogously unprecedented explosion in the amount and quality of scientific and technological progress. A ‘hard’ version of that hypothesis would claim that the increase of patents has actually spurred the acceleration of innovation, which otherwise would have not taken place. A ‘softer’ version would instead maintain that the increase of patents has been an effect rather than a cause of increased innovation, as the latter would also have taken place with weaker protection. The symmetrically opposite hypothesis is that the patent explosion is due to changes both in the legal and institutional framework and in firms’ strategy, with little relation to the underlying innovative activities. While it is difficult to come to sharp conclusions in the absence


129

of counterfactual experiments, circumstantial evidence does lend some support to the latter hypothesis. Certainly part of the growth in the number of patents is simply due to the expansion of the patentability domain to new types of objects such as software, research tools, business methods, genes and artificially engineered organisms (see also Tirole, 2003, on the European case). Moreover, new actors have entered the patenting game, most notably universities and public agencies (more on this in Mowery et al., 2001). Finally, corporate strategies vis-à-vis the legal claim of IPR appear to have significantly changed. First, patents have acquired importance among the non-physical assets of firms as a means to signal the enterprise’s value to potential investors, even well before the patented knowledge has been embodied in any marketable good. In this respect, the most relevant institutional change is to be found in the so-called ‘Alternative 2’ under the NASDAQ regulation (1984). This allowed ‘market entry and listing of firms operating at a deficit on the condition that they had considerable intangible capital composed of IPRs’. At the same time, patents seems to have acquired a strategic value, quite independently from any embodiment in profitable goods and even in those industries in which they were considered nothing more than a minor byproduct of R&D: extensive portfolios of legal rights are considered means for entry deterrence (Hall and Ziedonis, 2001) and for infringement and counter-infringement suits against rivals. Texas Instruments, for instance, is estimated to have gained almost US$1 billion from patent licenses and settlements resulting from its aggressive enforcement policy. It is interesting to note that this practice has generated a new commercial strategy called ‘defensive publishing’. According to this practice, firms that find it too expensive to build an extensive portfolio of patents tend to openly describe an invention in order to place it in the ‘prior art’ domain, thus preserving the option to employ that invention free from the interference of anyone who might eventually patent the same idea. Kortum and Lerner (1998) present a careful account of different explanations of recent massive increases in patenting rates, comparing different interpretative hypotheses. First, according to the ‘friendly court hypothesis’, the balance between costs related to the patenting process (in terms, e.g., of loss of secrecy) and the value of the protection that a patent affords to the innovator had been altered by an increase in the probability of successful application granted by the establishment in the USA of the Court of Appeals for the Federal Circuit (CAFC) specialized in patent cases – regarded by most observers as a strongly pro-patent institution (see Merges, 1996).

130


Second, the ‘regulatory capture’ tries to explain the surge of US patent applications, tracking it back to the fact that business firms in general and larger corporations (whose propensity to patent has traditionally been higher than average) in particular succeeded in inducing the US government to change patent policy in their favor by adopting a stronger patent regime. The third hypothesis grounds the interpretation in a general increase in ‘technological opportunities’ related, in particular, to the emergence of new technological paradigms such as those concerning information technologies and biotechnologies. Remarkably, Kortum and Lerner (1998) do not find any overwhelming support either for the political/institutional explanations or for the latter one driving the surge in patenting to changes in the underlying technological opportunities. At the same time there is good evidence that the cost related to IP enforcement has gone up, together with the firms’ propensity to litigate: the number of patents suits instituted in the US Federal Courts has increased from 795 in 1981 to 2573 in 2001. Quite naturally, this has led to significative increases in litigation expenditures. It has been estimated by the US Department of Commerce that patent litigation begun in 1991 led to total legal expenditures by US firms that were at least 25 percent of the amount of basic research by these firms in that year.

4.

THE BLURRED RELATIONS BETWEEN APPROPRIABILITY AND INNOVATION RATES: SOME EVIDENCE

What is the effect of the increase in patent protection on R&D and technical advance? Interestingly, in this domain also, the evidence is far from conclusive. This is for at least two reasons. First, innovative environments are concurrently influenced by a variety of different factors, which makes it difficult (for both the scholar and the policy-maker) to distinguish patent policy effects from effects due to other factors. Indeed, as we shall argue below, a first-order influence is likely to be exerted by the richness of opportunities, irrespective of appropriability regimes. Second, as patents are just one of the means to appropriate returns from innovative activity, changes in patent policy might often be of limited effect. At the same time, the influence of IPR regimes upon knowledge dissemination appears to be ambiguous. Hortsmann et al. (1985) highlight the cases in which, on the one hand, the legally enforced monopoly rents should induce firms to patent a large part of their innovations, while, on


131

the other hand, the costs related to disclosure might well be greater than the gain eventually attainable from patenting. In this respect, to our knowledge, not enough attention has been devoted to question whether the diffusion of technical information embodied in inventions is enhanced or not by the patent system. The somewhat symmetric opposite issue concerns the costs involved in the imitation of patent-protected innovations. In this respect, Mansfield et al. (1981) find, first, that patents do indeed entail some significant imitation costs. Second, there are remarkable intersectoral differences. For example, their data show 30 percent in drugs, 20 percent in chemicals and only 7 percent in electronics. In addition, they show that patent protection is not essential for the development of at least three out of four patented innovations. Innovators introduce new products notwithstanding the fact that other firms will be able to imitate those products at a fraction of the costs faced by the innovator. This happens both because there are other barriers to entry and because innovations are felt to be profitable in any case. Both Mansfield et al. (1981) and Mansfield (1986) suggest that the absence of patent protection would have little impact on the innovative efforts of firms in most sectors. The effects of IPR regimes on the propensity to innovate are also likely to depend upon the nature of innovations themselves and in particular whether they are, so to speak, discrete ‘standalone’ events or ‘cumulative’. So it is widely recognized that the effect of patenting might turn out to be a deleterious one on innovation in the case of strongly cumulative technologies in which each innovation builds on previous ones. As Merges and Nelson (1994) and Scotchmer (1991) suggest, in this realm stronger patents may represent an obstacle to valuable but potentially infringing research rather than an incentive. Historical examples, such as those quoted by Merges and Nelson on the Selden patent of a light gasoline in an internal combustion engine to power an automobile, and the Wright brothers’ patent on an efficient stabilizing and steering system for flying machines, are good cases to the point, showing how the IPR regime probably slowed down considerably the subsequent development of automobiles and aircraft. The current debate on property rights in biotechnology suggests similar problems, whereby granting very broad claims on patents might have a detrimental effect on the rate of innovation, in so far as they preclude the exploration of alternative applications of the patented invention. This is particularly the case with inventions concerning fundamental pieces of knowledge: good examples are genes or the Leder and Stewart patent on a genetically engineered mouse that develops cancer. To the extent that such techniques and knowledge are critical for further research that proceeds cumulatively

132


on the basis of the original invention, the attribution of broad property rights might severely hamper further developments. Even more so if the patent protects not only the product the inventors have achieved (the ‘onco-mouse’) but all the class of products that could be produced through that principle (‘all transgenic non-human mammals’) or all the possible uses of a patented invention (say, a gene sequence), even though they are not named in the application. More generally, the evidence suggests that the patents/innovation relation depends on the very nature of industry-specific knowledge bases, on industry stages in their life cycles and on the forms of corporate organizations. Different surveys highlight, first, such intersectoral differences and second, on average, the limited effectiveness of patents as an appropriability device for the purpose of ‘profiting from innovation’. Levin et al. (1987), for instance, report that patents are by and large viewed as less important than learning-curve advantages and lead time in order to protect product innovation and the least effective among appropriability means as far as process innovations are concerned (see Table 4.1). Cohen et al. (2000) present a follow-up to Levin et al. (1987) just cited addressing also the impact of patenting on the incentive to undertake R&D. Again, they report on the relative importance of the variety of mechanisms used by firms to protect their innovations – including secrecy, lead time, complementary capabilities and patents – see again Table 4.1. The percentage of innovations for which a factor is effective in protecting competitive advantage deriving from them is thus measured. The main finding is that, as far as product innovations are concerned, the most effective mechanisms are secrecy and lead time, while patents are the least effective, with the partial exception of drugs and medical equipment. Moreover, the reasons for the ‘not patenting’ choice are reported to be (1) demonstration of novelty (32 percent); (2) information disclosure (24 percent); and (3) ease of inventing around (25 percent). The uses of patents differ also relative to ‘complex’ and ‘discrete’ product industries. Complex product industries are those in which a product is protected by a large number of patents while discrete product industries are those in which a product is relatively simple and therefore associated with a small number of patents. In complex product industries, patents are used to block rival use of components and acquire bargaining strength in cross-licensing negotiations. In discrete product industries, patents are used to block substitutes by creating patent ‘fences’ (see Gallini, 2002; Ziedonis, 2004). It is also interesting to compare Cohen et al. (2000) with Levin et al. (1987), which came before the changes in the IPR regime and before the


Table 4.1

133

Effectiveness of appropriability mechanism in product and process innovations, 1983 and 1994 surveys, USA, 33 manufecturing industries

(a) Product innovations Mechanism

Patents Secrecy Lead time Sales & service Manufacturing

1st

2nd

3rd

4th

1983

1994

1983

1994

1983

1994

1983

1994

4 0 14 16 n.a.

7 13 10 4 3

3 0 14 16 n.a.

5 11 8 4 3

17 11 5 1 n.a.

7 2 7 7 14

9 22 0 0 n.a.

4 5 7 10 7

(b) Process innovations Mechanism

Patents Secrecy Lead time Sales & service Manufacturing Note: Source:

1st

2nd

3rd

4th

1983

1994

1983

1994

1983

1994

1983

1994

2 2 26 4 n.a.

1 21 3 0 10

4 10 5 16 n.a.

5 10 7 0 12

3 19 2 7 n.a.

3 1 16 3 10

24 2 0 6 n.a.

16 0 3 11 0

n.a. = not available. Levin et al. (1987) and Cohen et al. (2000), as presented in Winter (2002).

massive increase in patenting rates. Still, in Cohen et al. (2000) patents are not reported to be the key means to appropriate returns from innovations in most industries. Secrecy, lead time and complementary capabilities are often perceived as more important appropriability mechanisms. It could well be that a good deal of the increasing patenting activities over the last two decades might have gone into ‘building fences’ around some key invention, thus possibly raising the private rate of return to patenting itself (Jaffe, 2000), without however bearing any significant relation to the underlying rates of innovation. This is also consistent with the evidence discussed in Lerner (2002), who shows that the growth in (real) R&D spending pre-dates the strengthening of the IPR regime. The apparent lack of effects of different IPR regimes upon the rates of innovation appears also from broad historical comparisons. So, for example, based on the analysis of data from the catalogues of two nineteenth-century world fairs – the Crystal Palace Exhibition in London

134


in 1851, and the Centennial Exhibition in Philadelphia in 1876 – Moser (2003) finds no evidence that countries with stronger IP protection produced more innovations than those with weaker IP protection and strong evidence of the influence of IP law on sectoral distribution of innovations. In weak IP countries, firms did innovate in sectors in which other forms of appropriation (e.g. secrecy and lead time) were more effective, whereas in countries with strong IP protection significantly more innovative effort went to the sectors in which these other forms were less effective. Hence the interesting conclusion that can be drawn from Moser’s study is that patents’ main effect could well be on the direction rather than on the rate of innovative activity. The relationship between investment in research and innovative outcomes is explored at length in Hall and Ziedonis (2001) in the case of the semiconductor industry. In this sector, the limited role and effectiveness of patents – related to short product life cycles and fast-paced innovation which make secrecy and lead time much more effective appropriability mechanisms – also makes the surge in patenting (dating back to the 1980s) particularly striking. As Hall and Ziedonis report, in the semiconductor industry patenting per R&D dollar doubled over the period 1982–92. (Incidentally, note that, over the same period, patenting rates in the USA were stable in manufacturing as a whole and declined in pharmaceuticals.) Semiconductors are indeed a high-opportunity sector whose relatively low propensity to patent is fundamentally due to the characteristic of the knowledge base of the industry. Thus it could well be that the growth in patents might have been associated with the use of patents as ‘bargaining chips’ in the exchanges of technology among different firms. Such a use of (low-quality) patents – as Winter (2002) suggests – might be a rather diffused phenomenon: when patents are used as ‘bargaining chips’, that is, as ‘the currency of technology deals’, all the ‘standard requirements’ about such issues as non-obviousness, usefulness, novelty, articulability (you can’t patent an intuition), reducibility to practice (you can’t patent an idea per se) and observability in use turn out to be much less relevant. In Winter’s terms, ‘if the relevant test of a patent’s value is what it is worth in exchange, then it is worth about what people think it is worth – like any paper currency. “Wildcat patents”1 work reasonably well to facilitate exchanges of technology. So, why should we worry?’ One of the worries concerns the ‘tragedy of anti-commons’. While the quality of patents lowers and their use bears very little relation to the requirements of stimulating the production and diffusion of knowledge, the costs devoted to untie conflicting and overlapping claims on IP are likely to


135

increase, together with the uncertainty about the extent of legal liability in using knowledge inputs. Hence, as convincingly argued by Heller and Eisenberg (1998) and Heller (1998), a ‘tragedy of anti-commons’ is likely to emerge wherein the IP regime gives too many subjects the right to exclude others from using fragmented and overlapping pieces of knowledge, with no one having ultimately the effective privilege of use. In these circumstances, the proliferation of patents might turn out to have the effect of discouraging innovation. One of the by products of the recent surge in patenting is that, in several domains, knowledge has been so finely subdivided into separate property claims (on essentially complementary pieces of information) that the cost of reassembling constituent parts/ properties in order to engage in further research charges a heavy burden on technological advance. This means that a large number of costly negotiations might be needed in order to secure critical licenses, with the effect of discouraging the pursuit of certain classes of research projects (e.g. high-risk exploratory projects). Ironically, Barton (2000) notes that ‘the number of intellectual property lawyers is growing faster than the amount of research’. While it is not yet clear how widespread are the foregoing phenomena of a negative influence of strengthening IPR protection upon the rates of innovation, a good deal of evidence does suggest that, at the very least, there is no monotonic relation between IPR protection and propensity to innovate. So, for example, Bessen and Maskin (2000) observe that computers and semiconductors, while having been among the most innovative industries in the last 40 years, have historically had weak patent protection and rapid imitation of their products. It is well known that the software industry in the USA experienced a rapid strengthening of patent protection in the 1980s. Bessen and Maskin suggest that ‘far from unleashing a flurry of new innovative activity, these stronger rights ushered in a period in which R&D spending leveled off, if not declined, in the most patent-intensive industries and firms’. The idea is that in industries such as software, imitation might be promoting innovation and that, on the other hand, strong patents might inhibit it. Bessen and Maskin argue that this phenomenon is likely to occur in those industries characterized by a relevant degree of sequentiality (each innovation builds on a previous one) and complementarity (the simultaneous existence of different research lines enhances the probability that a goal might be eventually reached). A patent, in this perspective, actually prevents non-holders from the use of the idea (or of similar ideas) protected by the patent itself, and in a sequential world full of complementarities this turns out to slow down innovation rates. Conversely, it might well happen that firms would be better off in an environment characterized by easy imitation, whereby it would be true that imitation reduced current profits but it would be also true that easy

136


imitation raised the probability of further innovation to take place and of further profitable innovations to be realized. A related but distinct question concerns the relationship between IPRs, the existence of markets for technologies and the rates of innovation and diffusion (see Arora et al., 2001, for a detailed analysis of the developments). While it is certainly true that some IPR protection is often a necessary condition for the development of markets for technologies, there is no clear evidence suggesting that more protection means more markets. And neither is there general evidence that more markets drive higher rates of innovation. Rather, the degree to which technological diffusion occurs via market exchange depends to a great extent on the nature of technological knowledge itself, e.g. its degree of codifiability (Arora et al., 2001). So far we have primarily discussed the relations between the regimes of IPR protection and rates of innovations, basically concluding that either the relation is not there, or, if it is there it might be a perverse one, with strong IPR enforcement actually deterring innovative efforts. However, we know also that IPR protection is only one of the mechanisms for appropriating returns from innovation, and certainly not the most important one. What about the impact of appropriability in general? Considering together the evidence on appropriability from survey data (see Cohen et al., 2000; Levin et al., 1987), the cross-sectoral evidence on technological opportunities (see Klevorick et al., 1995) and the evidence from multiple sources on the modes, rates and directions of innovation (for two surveys, see Dosi, 1988; and Dosi et al., 2005), the broadbrush conclusion is that appropriability conditions in general have only a limited effect on the pattern of innovation, if any. This clearly applies above a minimum threshold: with perfectly zero appropriability, the incentive to innovate for private actors would vanish, but with few exceptions such a strict zero condition is hardly ever encountered. And the threshold, as the open source software shows, might be indeed very low.

5.

OPPORTUNITIES, CAPABILITIES AND GREED: SOME CONCLUSIONS ON THE DRIVERS OF INNOVATION AND ITS PRIVATE APPROPRIATION

There are some basic messages from the foregoing discussion of the theory and empirical evidence on the relationship between degrees of IPR protection and rates of innovation. The obvious premise is that some private expectation of ‘profiting from innovation’ is and has been throughout the history of modern capitalism,


137

a necessary condition for entrepreneurs and business firms themselves to undertake expensive and time-consuming search for innovations. That was already clear to classical economists and has been quite uncontroversial since. However, having acknowledged that, there are neither strong theoretical reasons nor any strong empirical evidence suggesting that tuning up or down appropriability mechanisms of innovations, in general, and appropriability by means of IPR in particular, has any robust effect upon the resources that private self-seeking agents devote to innovative search and upon the rates at which they discover new products and new production processes. As pointed out by the already mentioned survey by Jaffe (2000) on the effects of the changes in IPR regimes in recent years, ‘there is little empirical evidence that what is widely perceived to be a significant strengthening of intellectual property protection had significant impact on the innovation process’ (Jaffe, 2000, p. 540). Note that any tightening of IPR is bound to come together with a fall in ‘consumer surplus’: making use somewhat uneasily of such a static tool for welfare analysis, it is straightforward that as producers’ rents and prices on innovation rise, the consumer surplus must fall. Conversely, on the producers’ side, to the extent that firms’ attention and resources are, at the margin, diverted from innovation itself toward the acquisition, defense and assertion against others of property rights, the social return to the endeavor as a whole is likely to fall. While the evidence on all sides is scant, it is fair to say that there is at least as much evidence of these effects of patent policy changes as there is evidence of stimulation of research. (Jaffe, 2000, p. 555)

But if IPR regimes have at best second-order effects upon the rates of innovation, what are the main determinants of the rates and directions of innovation? Our basic answer, as argued above and elsewhere (see Dosi, 1988, 1997, Dosi et al., 2005) is the following. The fundamental determinants of observed rates of innovation in individual industries/technologies appear to be nested in levels of opportunities that each industry faces. ‘Opportunities’ capture, so to speak, the width, depth and richness of the sea in which incumbents and entrants go fishing for innovation. In turn, such opportunities are partly generated by research institutions outside the business sector, partly stem from the very search efforts undertaken by incumbent firms in the past and partly flow through the economic system via supplier/user relationships (see the detailed intersectoral comparisons in Pavitt, 1984, and in Klevorick et al., 1995). Given whatever level of innovative opportunities is typically associated with particular technological paradigms, there seems to be no general lack of appropriability

138


conditions deterring firms from going out and fishing in the sea. Simply, appropriability conditions vary a great deal across sectors and across technologies, precisely as highlighted by Teece (1986). Indeed, one of the major contributions of that work is to build a taxonomy of strategies and organizational forms and map them into the characteristics of knowledge bases, production technologies and markets of the particular activity in which the innovative/imitative firm operates. As these ‘dominant’ modes of appropriation of the returns from innovation vary across activities, so also should vary the ‘packets’ of winning strategies and organizational forms: in fact, Teece’s challenging conjecture still awaits a thorough statistical validation on a relatively large sample of statistical successes and failures. Note also that Teece’s taxonomy runs counter to any standard ‘IPRleads-to-profitability’ model according to which turning the tap of IPR ought to move returns up or down rather uniformly for all firms (except for noise), at least within single sectors. Thus the theory is totally mute with respect to the enormous variability across firms even within the same sector and under identical IPR regimes, in terms of rates of innovation, production efficiencies and profitabilities (for a discussion of such evidence see Dosi et al., 2005). The descriptive side – as distinguished from the normative, ‘strategic’ one – of the interpretation by Teece (1986) puts forward a promising candidate in order to begin to account for the patterns of successes and failures in terms of suitability of different strategies/organizational arrangements to knowledge and market conditions. However, Teece himself would certainly agree that such interpretation could go only part of the way in accounting for the enormous interfirm variability in innovative and economic performances and their persistence over time. A priori, good candidates for an explanation of the striking differences across firms even within the same line of business in their ability to both innovate and profit from innovation ought to include firm-specific features that are sufficiently inertial over time and only limitedly ‘plastic’ to strategic manipulation so that they can be considered, at least in the short term, ‘state variables’ rather than ‘control variables’ for the firm (Winter, 1987). In fact, an emerging capability-based theory of the firm to which Teece himself powerfully contributed (see Teece et al., 1990; and Teece et al., 1997) identifies a fundamental source of differentiation across firms in their distinct problem-solving knowledge, yielding different abilities of ‘doing things’ – searching, developing new products, manufacturing and so on (see Dosi et al., 2000, among many distinguished others). Successful corporations, as is argued in more detail in the introduction to Dosi et al. (2000), derive competitive strength from their above-average performance


139

in a small number of capability clusters where they can sustain a leadership. Symmetrically, laggard firms often find hard the imitation of perceived best-practice production technologies difficult because of the problem of identifying the combination of routines and organizational traits that makes company x good at doing z. Such barriers to learning and imitation, it must be emphasized, have very little to do with any legal regime governing the access to the use of supposedly publicly disclosed but legally restricted knowledge such as that associated with patent-related information. Much more fundamentally, it relates to collective practices that in every organization guide innovative search, production and so on. In fact, in our view, given the opportunities for innovation associated with a particular paradigm – which approximately also determine the ensuing industry-specific rates of innovation – who wins and who loses among the firms operating within that industry depends on both the adequacy of their strategic choices – along the lines of the taxonomy of Teece (1986) – and on the type of idiosyncratic capabilities that they embody. In our earlier metaphor, while the ‘rates of fishing’ depend essentially on the size and richness of the sea, idiosyncratic differences in the rates of success in the fishing activity itself depend to a large extent on firm-specific capabilities. Moreover, the latter, jointly with complementary assets, fundamentally also affects the ability to ‘profit from innovation’. Conversely, if we are right, this whole story has very little to do with any change in the degree to which society feeds the greed of the fishermen, in terms of prices they are allowed to charge for their catch. That is, the tuning of IPR-related incentives is likely to have only second-order effects, if any, while opportunities, together with the capabilities of seeing them, are likely to be the major drivers of the collective ‘unbound Prometheus’ of modern capitalism and also to shape the ability of individual innovators to benefit from it.

ACKNOWLEDGMENT This chapter was previously published in Research Policy, Vol. 35, No 8, 2006, pp. 1110–21.

NOTE 1. Winter is here pursuing an analogy between patents and ‘wildcat banknotes’ in the US free banking period (1837–65).

140


REFERENCES Arora, A., A. Fosfuri and A. Gambardella (2001), Markets for Technology: Economics of Innovation and Corporate Strategy, Cambridge, MA: MIT Press. 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. Barton, J. (2000), ‘Reforming the patent system’, Science, 287, 1933–34. Bessen, J. and E. Maskin (2000), ‘Sequential innovation, patents and imitation’, Working Paper 00-01, Cambridge, MA: MIT Department of Economics. Cohen, W., R.R. Nelson and J. Walsh (2000), ‘Protecting their intellectual assets: appropriability conditions and why US manufacturing firms patent or not’, Discussion Paper 7552, NBER. David, P. (1993), ‘Intellectual property institutions and the panda’s thumb: patents, copyrights, and trade secrets in economic theory and history’, in M. Wallerstein, M. Mogee and R. Schoen (eds), Global Dimensions of Intellectual Property Protection in Science and Technology, Washington, DC: National Academy Press. David, P. (2002), ‘Does the new economy need all the old IPR institutions? Digital information goods and access to knowledge for economic development’, presented at Wider Conference on the New Economy in Development, Helsinki. Dosi, G. (1982), ‘Technological paradigms and technological trajectories. A suggested interpretation of the determinants and directions of technological change’, Research Policy, 11, 147–62. Dosi, G. (1988), ‘Source, procedures and microeconomic effects of innovation’, Journal of Economic Literature, 26, 1120–71. Dosi, G. (1997), ‘Opportunities, incentives and the collective patterns of technical change’, Economic Journal, 107, 1530–47. Dosi, G., R.R. Nelson and S. Winter (eds.) (2000), The Nature and Dynamics of Organizational Capabilities, Oxford and New York: Oxford University Press. Dosi, G., L. Orsenigo and M. Sylos Labini (2005), ‘Technology and the economy’, in N. Smelser and R. Swedberg (eds), The Handbook of Economic Sociology, Princeton NJ: Princeton University Press/Russell Sage Foundation. Gallini, N. (2002), ‘The economics of patents: lessons from recent U.S. patent reform’, Journal of Economic Perspectives, 16, 131–54. Grandstrand, O. (2005), ‘Innovation and intellectual property Rights’, in I. Fagerberg, D. Mowery, and R. Nelson (eds), The Oxford Handbook of Innovation, Oxford: Oxford University Press, pp. 266–90. Hall, B. (2005), ‘Exploring the patent explosion’, Journal of Technology Transfer, 30, 35–48. Hall, B. and R. Ziedonis (2001), ‘The patent paradox revisited: firm strategy and patenting in the US semiconductor industry’, Rand Journal of Economics, 32, 101–28. Heller, M. (1998), ‘The tragedy of the anticommons: property in transition from Marx to markets’, Harvard Law Review, 111, 698–701. Heller, M. and R. Eisenberg (1998), ‘Can patents deter innovation? The anticommons in biomedical research’, Science, 280, 698–701. Hortsmann, I., J. Mac Donald and A. Slivinski (1985), ‘Patents as information


141

transfer mechanisms: to patent or (maybe) not to patent’, Journal of Political Economy, 93, 837–58. Jaffe, A. (2000), ‘The US patent system in transition: policy innovation and the innovation process’, Research Policy, 29, 531–57. Jaffe, A. and J. Lerner (2004), Innovation and its Discontents, Princeton, NJ: Princeton University Press. Klevorick, A., R. Levin, R.R. Nelson and S. Winter (1995), ‘On the sources and interindustry differences in technological opportunities’, Research Policy, 24, 185–205. Kortum, S. and J. Lerner (1998), ‘Stronger protection or technological revolution: what is behind the recent surge in patenting?’, Rochester Conference Series on Public Policy, 48, 247–307. Landes, D. (1969), The Unbound Prometheus, Cambridge, MA: Cambridge University Press. Lerner, J. (2002), ’150 years of patent protection’, American Economic Review: Papers and Proceedings, 92, 221–25. Levin, R., A. Klevorick, R.R. Nelson and S. Winter (1987), ‘Appropriating the returns from industrial R&D’, Brookings Papers on Economic Activity, pp. 783–820. Machlup, F. (1958), ‘An economic review of the patent system’, Discussion Paper, US Congress, Washington DC: Government Printing Office. Mansfield, E. (1986), ‘Patents and innovation: an empirical study’, Management Science, 32, 173–81. Mansfield, E., M. Schwartz and S. Wagner (1981), ‘Imitation costs and patents: an empirical study’, Economic Journal, 91, 907–18. Merges, R. (1996), ‘Contracting into liability rules: intellectual property rights and collective rights organizations’, California Law Reviews, 84, 1293–386. Merges, R. and R.R. Nelson (1994), ‘On limiting or encouraging rivalry in technical progress: the effects of patent scope decisions’, Journal of Economic Behavior and Organization, 25, 1–24. Moser, P. (2003), ‘How do patent laws influence innovation? Evidence from nineteenth-century world fairs’, Discussion Paper, NBER. Mowery, D., R.R. Nelson, B. Sampat and A. Ziedonis (2001), ‘The growth of patenting and licensing by US universities: an assessment of the effects of the Bayh–Dole Act of 1980’, Research Policy, 30, 99–119. Pavitt, K. (1984), ‘Sectoral patterns of innovation: toward a taxonomy and a theory’, Research Policy, 13, 343–73. Scotchmer, S. (1991), ‘Standing on the shoulders of giants: cumulative research and the patent law’, Journal of Economic Perspectives, 5, 29–41. Stiglitz, J. (1994), Whither Socialism?, Cambridge, MA: MIT Press. Teece, D. (1986), ‘Profiting from technological innovation: implications for integration, collaboration, licensing and public policy’, Research Policy, 15, 285–305. Teece, D., G. Pisano and A. Shuen (1997), ‘Dynamic capabilities and strategic management’, Strategic Management Journal, 18, 509–33. Teece, D., R. Rumelt, G. Dosi and S. Winter (1990), ‘Understanding corporate coherence: theory and evidence’, Journal of Economic Behavior and Organization, 23, 1–30. Tirole, J. (2003), Protection de la propriété intellectuelle: une introduction et quelques pistes de réflexion, Rapport pour le Conseil d’Analyse Economique.

142


Winter, S. (1982), ‘An essay on the theory of production’, in H. Hymans (ed.), Economics and the World around It, Ann Arbor, MI: University of Michigan Press, pp. 55–93. Winter, S. (1987), ‘Knowledge and competences as strategic assets’, in D. Teece (ed.), The Competitive Challenge: Strategies for Industrial Innovation and Renewal, New York: Harper & Row. Winter, S. (1993), ‘Patents and welfare in an evolutionary model’, Industrial and Corporate Change, 2, 211–31. Winter, S. (2002), ‘A view of the patent paradox’, Presentation at London Business School, 20 May. Ziedonis, R. (2004), ‘Don’t fence me in: fragmented markets for technology and the patent acquisition strategies of firms’, Management Science, 50, 804–20.

5.

Global bioregions: knowledge domains, capabilities and innovation system networks Philip Cooke

INTRODUCTION This chapter explores a field, ‘globalization of bioregions’, that is of growing interest to social science and policy alike. A number of special journal issues have been published (Cooke, 2003, 2004; Lawton Smith and Bagchi-Sen, 2004), and others may be anticipated from different stables. Outside the spatial field but well informed by it, an earlier one was published in Small Business Economics (2001), subsequently enlarged into a book (Fuchs, 2003). Numerous other books are published in this increasingly dynamic field (Orsenigo, 1989; McKelvey, 1996; De la Mothe and Niosi, 2000; Carlsson, 2001). There has also been a surge in paper publication, too large to cover in this brief review, but that of Powell et al. (2002) is noteworthy in showing close spatial interactions among biotechnology SMEs (small and medium-sized enterprises) and venture capitalists from non-spatially designed survey data. Most economists writing about the evolution of the biotechnology sector (many authors refuse to think of a technology as an ‘industry’) accept that it is a classic case of a sciencedriven sector, highly active in R&D expenditure, patenting of discoveries, with research increasingly dominated by public (mainly university) laboratories, but exploitation mainly dominated not by large but by small firms.1 This is so for the biggest part of biotechnology, accounting for about 70 per cent of sales, biopharmaceuticals, which is the subject of much of what follows. Biopharmaceuticals, as a large subsector of biotechnology, is fascinating. It belies the prediction that multinational corporations are bound to dominate all aspects of any industry due to economies of scale. While the likes of Pfizer and Glaxo are certainly large global actors by any standards, they are rapidly losing capabilities to lead research. This function has moved – rapidly in the 1990s – to university ‘Centres of Excellence’ 143

144


and particularly to networks of dedicated biotechnology firms (DBFs) that often lead important research projects (Orsenigo et al., 2001) relying on ‘big pharma’ for grants, marketing and distribution. All are agreed that DBFs cluster around leading-edge university campuses and that even if they are not the kind of DBFs that depend on immediate interaction with research laboratories, they still agglomerate. A good example of such an agglomerated non-cluster occurs to the south-west of London, centred on Guildford. Why there? Static externalities like good access to Heathrow airport and a suitable talent pool from which to poach. But firms scarcely interact except with global clients. This contrasts with the dynamic externalities from proximate co-presence with many and various significant others in the sector. In the following brief sections, these will be highlighted by reference to key issues of relevance to cluster theory more generally, the specifics of science-driven industry, the post-Schumpeterian symbiotics that characterize the double dependence of small businesses on corporate financial clout and ‘big pharma’ on DBF and laboratory research excellence. Also posed is the question – raised in the ‘spillovers’ debate – as to whether most cluster interactions in biotechnology (and yet to be asked for elsewhere) are really forms of market rather than milieu-type transaction, pecuniary rather than non-pecuniary spillovers, and traded rather than untraded interdependencies (Zucker et al., 1998; Owen-Smith and Powell, 2004).

BIOTECHNOLOGY AND CLUSTER THEORY Biotechnology is easily seen to be an economic activity, technology or sector both of the present but more for the future. Here, as elsewhere, we shall call it an enabling technology although, for example, biopharmaceuticals clearly also constitutes a subsector of the broader healthcare industry along with research, pharmaceuticals, hospitals, medical schools and a further panoply of support services that together comprise in some countries over 15 per cent, and sometimes more, of GDP, and in the UK up to 25 per cent of national R&D expenditure (OECD, 2000; DTI, 1999). Agricultural bioscience is similarly part of the wider agro-food industry. Biotechnology can be shown to have its own value chain, especially in biopharmaceuticals, although some parts of even that are indistinguishable from agro-food biotechnology or environmental biotechnology, especially regarding enzymes, recombinant DNA and the like. Indeed, agricultural ‘farming’ of DNA antibody sequences for healthcare application has been patented by San Diego firm Epicyte through its patented Plantibodies product. Although objections have been raised by researchers about the

Global bioregions

145

scope of such a ‘broad patent’ for restricting academic research freedom, it does not take extra-sensory insight to see that such technologies can be the saving of some agricultural areas assailed by global competition provided they become more ‘knowledge-intensive’ in this and other ways. Precisely this sort of Galilean leap seems to be at the heart of the clustering process in this industry. Thus Epicyte’s actual patent-holder is San Diego’s renowned Scripps Research Institute, where its founders Hiatt and Hine worked. Nearby Dow AgroSciences is the licensing and funding partner. This is the ‘holy trinity’ of interaction around biotechnology discovery – basic research, ‘pharma’ or agro-chemicals funding and DBF exploitation for commercial purposes, with ‘pharma’ earning returns through marketing and distribution. In San Diego, to continue there for a moment, are some 400 SMEs engaged in some aspect of biotechnological core or peripheral activity and, as Lee and Walshok (2001) show, 45 were founded by the Scripps, Salk, Burnham or La Jolla research institutes, while 61 were founded directly from the University of California. Many more were founded in the area as a consequence of knowledge flows and investment opportunities for incoming and resident entrepreneurs advertised through activities such as those promoted by the much-emulated UC San Diego Connect network. Equally characteristic is the well-connected non-local network of the typical DBF. This is obviously not geographically embedded in the topography of the cluster but seems to be consistently a feature of DBFs in such clusters, namely their engagement in strong functional or organizational proximity with other DBFs or research laboratories elsewhere, even globally. Thus Epicyte has, for Plantibodies, strong research or production links with DBFs in Princeton, NJ, Malvern, PA, Baltimore, MD and Biovation in Aberdeen, Scotland. Its research affiliations are with Johns Hopkins University, Baltimore, MD, Oklahoma Medical Research Foundation, Cornell University, Ithaca, NY and its closely affiliated agro-biotech centre, The Boyce Thompson Institute. Hence it has more non-local than local ‘partners’, or from a markets viewpoint, customers and suppliers. However, Scripps and Epicyte are specific research leaders and knowledge originators for Plantibodies and the others are suppliers in the knowledge value chain of research, funding and specific products and/or technologies, testing and trialling capabilities, and marketing and distributional expertise (Cooke, 2002). What differentiates biotechnology clusters like this from others in different sectors is the science-driven nature of the business, and it is worth remembering that there are roughly 400 other DBFs in this specific cluster with many board and advisory cross-links, some of them local but more non-local, each to some extent aping the profile of Epicyte, with strong

146


localized and stronger globalized, or at least North American, business linkages. It has recently been argued by Wolter (2002) that this characteristic differentiates biotechnology clusters from Porterian ‘market’-driven clusters: the Porterian ‘diamond’ of factor conditions for competitiveness postulates transparent competitors, clear market opportunities, known production and supply schedules and appropriate market support to enable competing firms to perform effectively. Bioscience lacks this kind of market stability. Most of its incumbents do not make profits, indeed new stock market rules and indeed stock markets like NASDAQ and (in the UK) AIM (Alternative Investment Market) had to be set up to allow firms not making profits to be publicly invested in. Knowledge is unstable and it is prone to being swiftly superseded. Skills are in short supply and requirements change rapidly. Deep tacit knowledge of indeterminate but potentially enormous value is in the minds and hands of academics; hence their capacity to generate upfront input returns from licensing income or patent royalties. But many DBFs presently earn little from product or treatment outputs. The key to this unusual kind of economy is proximity to knowledge ‘fountainheads’ such as those described for Maryland by Feldman and Francis (2003), Munich by Kaiser (2003), Montreal by Niosi and Bas (2003), Cambridge by Casper and Karamanos (2003), and key knowledge centres in Israel by Kaufmann et al. (2003). The largest of these may be termed ‘megacentres’ after a French term for science-led innovation complexes like Grenoble that combine ‘ahead-of-the-curve’ science with disruptive research, heavy investment in scientific and technological ‘talent’, and incentives for entrepreneurs to exploit such discoveries commercially (Johnson, 2002). This is a feature of biotechnology that steps beyond the notion of ‘clusters’ and may presage new developmental opportunities for other science-driven, research-intensive and heavily publicly funded complexes in future, nanotechnology, for example, being the actual trigger for Grenoble’s ‘megacentre’ appellation.

BEYOND SCHUMPETER? Reference to disruptive research and technologies recalls Schumpeter’s view of innovation and entrepreneurship as characterized by ‘creative destruction’ (Schumpeter, 1975). That view has been enormously influential among the Neo-Schumpeterian School of innovation theorists and, particularly, innovation systems analysts such as Freeman (1987), Lundvall (1992) and Nelson (1993). Inspired by Schumpeter’s insights into the heroic efforts of the innovator, whether as the lone entrepreneur

Global bioregions

147

or, later, the equally heroic R&D team of the large corporation as the key driver of competition, they developed a heterodox perspective that attracted a large academic and policy following. This was not least because of the School’s rejection of the desiccated theorems of neoclassical economics that treated Schumpeter’s wellspring of capitalism as a commodity to be acquired ‘exogenously’ off the shelf. This is no longer believed, even by neoclassicals. Although the national innovation systems authors mentioned in this section had no spatial brief – other than for a poorly theorized ‘nation state’ which in the case of Taiwan and South Korea could look odd, their adherence to Schumpeterian insights implied, by his concept of ‘swarming’, clusters. These are logical consequences where imitators pile in to mimic the innovator and seek early superprofits before over-competition kills the goose that lays the golden eggs. Hence ‘creative destruction’, as some older technology – sailing ships – gives way to newer – steamships. Many shipyards arise only to be competed into a few or none, as competitive advantage moves their epicentre to cheaper regions around the world. It can be argued that the swift destruction of value observable in such processes, and more recently in the bursting of the dot.com and financial bubble (see below), with their negative upstream implications for world trade, constitutes a problem of ‘over-innovation’ rather as, in a completely opposite way, German car and machinery makers used to be accused of ‘over-engineering’. Mobile telephony and computing may equally be accused of having too many ‘bells and whistles’ that most people never use. Biotechnology is not like this. It is in many ways either non-Schumpeterian or post-Schumpeterian, although, as we have seen, it certainly clusters, even concentrates in ‘megacentres’ of which there are very few in the world. But it does this not to outcompete or otherwise neutralize competitor firms, but rather to maximize access to knowledge at an early stage in the knowledge exploration phase that is difficult to exploit in the absence of multidimensional partnerships of many kinds. Feldman and Francis’s (2003) paper presents a remarkable, even original, picture of the manner in which the driver of cluster-building in Washington and Maryland’s Capitol region was the public sector. Importantly, the National Institutes of Health conduct basic research as well as funding others to do it in university and other research laboratories. Politics means that sometimes bureaucracies must downsize, something that was frequently accompanied by scientists transferring knowledge of many kinds into the market, not least in DBFs that they subsequently founded. Interactions based on social capital ties with former colleagues often ensued in the marketplace. Before founding Human Genome decoding firm Celera, Craig Venter was one of the finest exploiters of such networks, not least with the founding

148


CEO of Human Genome Sciences Walter Haseltine, who had been a colleague of Walter Gilbert at Harvard and James Watson, co-discoverer of DNA, and later of Nobel Prize-winner David Baltimore (Wickelgren, 2002). Thus, as Feldman and Francis (2003) show, inter-organizational capabilities networks from government to firms as distinct from more normally understood laboratory–firm or firm–firm links may also be pronounced in biotechnology. This is seldom observed in the classical Schumpeterian tradition, which rests to a large degree on engineering metaphors, where entrepreneurship is privileged. Moreover, few Neo-Schumpeterians devote attention to the propulsive effect of public decisions (for an exception, see Gregersen, 1992), even though regional scientists had long pointed out the centrality of military expenditure to the origins of most ICT clusters (Hall and Markusen, 1985). In Niosi and Bas (2003) it is shown how relations are symbiotic rather than creatively destructive among firms in the Montreal and Toronto clusters. Arguing for the ‘technology’ definition of biotechnology, they say it is not a generic but a diverse set of activities, each with specificity. Of interest is that in biopharmaceuticals they interact in networks to produce products for different markets. Hence there is crosssectoral complexity in markets as well as research, although healthcare is the main market. Accordingly, they too point to the ‘regional innovation system’ nature of biotechnology, spreading more widely and integrating more deeply into the science base than is normally captured by the localized, near-market idea of ‘cluster’. Thus CROs and biologics suppliers may exist in the orbit, not the epicentre of the cluster. This collaborative aspect of biotechnology is further underlined by Casper and Karamanos (2003). For example, they show that, regarding joint publication, while there is substantial variation in the propensity to publish with laboratory, other academic or firm partners by Cambridge (UK) biotechnology firms, only 36 per cent of firms are ‘sole authors’ and the majority of partners are those from the founders or their laboratory. Academic collaborators are shared equally between Cambridge and the rest of the UK, with international partners a sizeable minority. Hence the sector is again rather post-Schumpeterian in its interactive knowledge realization, at least in respect of the all-important publication of results that firms will probably seek to patent. Moreover, they will, in many cases, either have or anticipate milestone payments from ‘pharma’ companies with whom they expect to have licensing agreements. Finally, biotechnolgy is interesting for the manner in which epistemological, methodological and professional shifts in the development of the field created openings for DBFs when it might have been thought that already large pharmaceuticals firms would naturally dominate the new

Global bioregions

149

field. This is an important sub-text in Kaiser’s (2003) paper on the slow development of biotechnology in Germany, and specifically its main centre at Munich. Pharmaceuticals arose from chemistry, not biology, and although, as others show, modern biotechnology relies on molecular expertise that chemists pioneered, most of its discoveries derive ultimately from biology. Hence the ‘molecular biology revolution’, as Henderson et al. (1999) call it, somehow superseded the epistemological primacy of chemistry. Methodologically, chemists retain a hold over such activities as screening for particular molecules that act as inhibitors for those that are disease-causing. But methodologically that process has become infinitely swifter and more automated with the advent of supercomputer-strength high-throughput screening. Such equipment was also developed by specialist ICT companies applying expertise to bioinformatics and the like. A good example of the advantage of the application of such examination knowledge is the race to decode the human genome. Here, Celera pitted itself against, and for a long time led, an enormous transatlantic university research centre-based consortium. The reason it could do this was Craig Venter’s innovative approach, sequencing segments ‘shotgun’ fashion rather than whole gene sequences, and Californian firm Perkin-Elmer’s decision to produce equipment to enable this to be done rapidly. It was only when the consortium’s budget was massively increased so that they, too, could afford such equipment that they caught up and the race ended neck and neck. This, of course, meant that, professionally speaking, ICT equipment producers, specialist software engineers and information scientists have become integral to biotechnology.2

CLUSTERING: MARKET OR MILIEU? Strong cases can straightforwardly be made for the characteristic spatial proximity emphasis there is in biotechnology in North America, Europe and, in the guise of Israel and Singapore, Asia. But global linkages are also pronounced (see below). In Kaufmann et al. (2003) it is shown that the academic nature of the background of most firm founders means that they have a weak business expertise and cluster accordingly, but near to their laboratory home base in the early stages. The importance of public subsidy in the form of regional support policies is also stressed. However, it is also shown that with maturity, complex business and technological needs cause expansion of networks overseas. In this, the need to access business partners for examination knowledge expertise, particularly in testing and trialling for clinical research, is paramount as it is unavailable

150


in such a small-country setting. As with early financing requirements that also contribute to a clustering disposition, there is a strong business market impulse for clustering in Israel, one that also drives expansion of networks. In other words, partnering is strategic but primarily for market reasons. Thus the milieu is in quite pronounced ways one in which business is done with trusted others, rather than, as may sometimes be presumed, through exchanging of much untraded resource. This is echoed in Casper and Karamanos’, (2003) study of Cambridge, UK, where they conclude that a major factor in clustering is the presence of an ‘ideas market’ and access to specialized labour markets. This attracts spin-ins as well, a feature shown to be pronounced in Cambridge.3 Nevertheless, in Germany, after a number of false starts, the BioRegio programme earned praise for rapid stimulation of biotechnology startup businesses in 1997–2002. Of importance in this success was the preexistence of institutional partnership and a special institution, such as Munich’s BioM, capable of promoting startup activity. But, as Kaiser (2003) notes, this dynamism was equally predicated on the emergence of a venture capital market, itself assisted by the co-funding support available from BioRegio and other regional and local funds. Here, then, public investment was required to stimulate market activity to promote clusters in areas that possessed both organizational and institutional capabilities for clustering to occur.4 We may say that the institutional dimension constitutes aspects of a milieu effect, but a rather different one from the type hypothesized by the GREMI group, for whom the public sector seldom plays a leading role (Maillat, 1998). Of course a different but crucial role was, as discussed, shown to have been played out in the Capitol region in the USA in the account of biotechnology growth provided by Feldman and Francis (2003). In Canada, one could say that the clustering developed in classic style for the sector, namely large-scale publicly funded research of high quality created opportunities for entrepreneurship by academics and venture capitalists. Strong links with pharmaceuticals firms strengthened the funding base of DBFs, some of which have been acquired by larger companies, while newer spin-outs continue to emerge.

BIOTECHNOLOGY’S GLOBALIZATION: FROM BELOW? Today, biotechnology is seen as a sine qua non of regional economic development. Paradoxically, as ministers and many scientists bemoan the aspirations of regional agencies everywhere to establish a biotechnology presence, their words are drowned in the rush to implement policies that

Global bioregions

151

seek to ensure future regional competitiveness from biotechnology applications in all their myriad forms. This is, of course, reminiscent of the rush to clone Silicon Valley by policy-makers worldwide during the 1980s and 1990s. Most of these efforts were unsuccessful on a narrow evaluation, in the sense that Silicon Valley remains unique in both the scale of its local and global impact, both during the upswing and more recently the downswing in the ICT industry. It is worth recalling that job losses in Silicon Valley from 2001 to 2003 were 400 000 and San Jose lost 20 per cent of its jobs. The latter statistic, according to a UCLA Anderson Business School report, constitutes ‘the single largest loss of jobs by any major metropolitan area since at least the second world war’ (Bazdarich and Hurd, 2003; Parkes, 2004). Despite the roller-coaster ride that high-technology industry clearly provides, these are, on balance, growth sectors over the business cycle, and on a broad evaluation regions that sought to anchor high-technology growth during the past two decades have often proved remarkably successful. The Kista science park, north of Stockholm, founded in 1972, still has some 600 companies with 27 000 employees (down from 29 000 in 2000), more than 3000 students in IT, and numerous university and research institutions active in wireless communication and the mobile Internet. Other such complexes in Helsinki, Ottawa and Cambridge could be mentioned, let alone Austin and Dallas, Texas or Dulles, Virginia and Bangalore, Hyderabad, Shenzen and Shanghai. None of these is an exact clone of the original, but they are important contributors to the regional economies they inhabit. Biotechnology has begun to evolve its ecology likewise, albeit at a lesser scale, and with different specificities. Thus, first, agro-food biotechnology locales are different from those specializing in biopharmaceuticals – although occasionally they overlap (e.g. San Diego). An interesting case of the latter phenomenon is presented in the Medicon Valley (Sweden–Denmark) study of Coenen et al. (2004). The proximate cause of locational variation beween biopharmaceutical and agro-food clusters in general is the fundamentally public funding of research in the former and private funding in the latter. Moreover, firm structure, organization and capabilities bring a heavy influence to bear. Thus agro-food is dominated by a few multinational corporations like Monsanto, Bayer, Dow, Unilever, Nestlé, Syngenta and Aventis. Joining them are former chemicals companies transforming themselves into ‘biologics’ (fermentation and contract DNA manufacturing specialists) firms, like Avecia, DSM, Lonza and rising state-backed firms like BioCon in India and China Biotech. Significant agglomerations of such companies occur in out of the way places like Guelph and Saskatoon in Canada, Wageningen, the

152


Netherlands and Basel, Switzerland. More substantial but somehow still unusual places from a high-tech viewpoint such as St Louis, Missouri also appear on the radar in the form of ‘BioBelt’, the world’s biggest agro-food agglomeration, as Ryan and Phillips (2004) have shown. Second, unlike the biopharmaceuticals industry, these corporates conduct much leading-edge research in house, although they are also found in juxtaposition to national research institutes, themselves often located historically in agricultural zones. This explains the concentration in Saskatoon, where oilseed rape innovation capabilities have been concentrated since the 1940s, and St Louis with comparable competences in the corn and soybean sectors. Syngenta – the Basel-based agro-food firm formed from the merger of ICI and Novartis’s agro-food divisions – established a major cotton research centre in Raleigh–Durham, a traditional centre of expertise in cotton research, now leading in cotton genomics. The same firm was first to decode the rice genome, albeit at its Torrey Mesa Research Institute in San Diego. As many published papers have shown, biopharmaceuticals is unlike this. University research departments and centres of research expertise lead the field in bioscientific research and the exploitation for commercial purposes of biotechnology. Thus corporate ‘pharma’ is a supplicant to these leading-edge research centres. Recent papers by Lawton Smith (2004) and Bagchi-Sen and Scully (2004) underline the centrality of university and medical school research as a magnet for specialist biotechnology firms of small scale that out-license their findings into a supply chain that ends with pharmaceutical companies marketing and distributing the final product on a global scale. In what follows, three themes of significance to the ‘globalization of bioregions’ are highlighted. These are first, the practices of globalization in biosciences, second the vehicles for its implantation, and third, the evolving, networked global hierarchy in biotechnology5.

THE PRACTICES OF GLOBALIZATION IN BIOSCIENCES We have already seen that biotechnology underpins a range of industries, that its broad categorization into biopharmaceuticals, agro-food and environment/energy biotechnology produces distinctive economic geographies, at least in the first two. Briefly, we can say that there is a third distinctive pattern in regard to the main environmental/energy segment in that it is normally found in older industrial regions such as the Ruhr region in Germany or even established chemicals complexes like Grangemouth in Scotland. The latter was a product of ‘growth pole’ planning involving

Global bioregions

153

the major chemicals companies in the UK in the 1960s. It currently hosts the aforementioned Avecia and Syngenta (both divested former divisions of ICI), British Petroleum (BP), Polyami (artificial fibres), GE Plastics, Rohm & Haas (paints and personal care products). Avecia sells both energy and environmental effluent services from its Grangemouth site, both to co-inhabitants and the wider market. This is in addition to its core products of fine chemicals, pharmaceutical biotechnology and biochemical/biologics manufacturing. Returning to the Ruhr region, the origin of its bioenvironmental industry, consisting of some 600 firms, lies in the coal and steel industry, whose companies diversified from their traditional home base during the 1980s, utilizing ventilation and filtration technologies integral to their former capabilities as applications to meet new demand from regulators for higher-standard emissions and environmental clean-up controls. Hence one segment of biotechnology is based on ‘open innovation’ (Chesbrough, 2003) among research institutes, pharmaceuticals companies and small, dedicated biotechnology firms (DBFs). The second one in agro-food is organized around direct basic knowledge transfer between large public research institutes and large agro-food biotechnology firms with little or no DBF involvement (although Wageningen has around ten agro-food DBFs and a consortium, the Centre for Biosystems Genomics, a network of four universities, two research institutes and 15 firms in agrofood biotechnology across the Netherlands in its ‘Food Valley’ science park). The third model is inter-industry trade by direct spin-out, divisional divestiture, or in-house evolution of innovation capabilities from coal, steel and chemicals corporates entering and selling products and services in new environmental or energy markets. The last-named model is more engineering- (synthetic) than science-driven (analytical) and thus unaffected in locational terms by the presence or absence of research institutes, but rather remains embedded in its original locational trajectory (Coenen et al., 2004).

THE VEHICLES FOR IMPLANTATION OF GLOBALIZED BIOSCIENCE In the biopharmaceuticals case, globalization is driven through knowledge networks, as inspection of Figure 5.1 shows. Leading-edge research centres and their star scientists interact through co-publishing. Such knowledge and the research on which it rests is of great value to the pharmaceuticals industry. Hence ‘big pharma’ not only funds and licenses such research and its results; it formerly also invested substantial equity stakes in spin-out

154

The capitalization of knowledge Paris

Sydney

Stockholm

PU

SU

Uppsala

INS

UNSW

RIT KI

Copenhengen

Lund Ucop

UL UU

San Diego

UG

SanFranc

Toronto

UCSD

Tokyo

Grenoble

CBSP

SUAS

UT

UCSF

TML

Salk

UTo

SU

TIT

SRI UBer BI

Juresalem HeU

Boston

Montreal

HaH

HHHaH

MU

HMS

New York

Munich

Cambridge(MA)

UM

GH

NYU

Singapore

MIPS MIT

Zurich

Cam(UK)

RU

ZU

DSI

HU

CamULondon MSR

Geneva

NUS MI

ColmU

Oxford

BPRC UG

UCL

ICL

OU JRH

3 7–8

London

4–6 >10

NIMR

NIMR

Notes: Represented are co-publications between ‘star’ bioscientists in leading research institutes in eight of the top ten Science Citation Index journals for recent periods: Cell (2002–04); Science (1998–2004); Proceedings of the National Academy of Sciences (2002–04); Genes and Development (2000–04); Nature (1998–2004); Nature Biotechnology (2000–04); Nature Genetics (1998–2004); EMBOJ (European Molecular Biology Organization Journal) (2000–04). In all, 9336 articles were checked. Abbreviations refer to research institutes.

Figure 5.1

Publishing collaborations

firms established by such stars, or acquired them outright, although nowadays it tends to make variable-term exchange-based or traded partnerships with them to access valued, preferably patented knowledge to be transformed into biotechnologically based drug treatments (‘open innovation’). Furthermore, the large corporates also establish research institutes in such science-driven clusters.6 We noted such cluster penetration for ‘open innovation’ in regard to Syngenta for genomic decoding in San Diego. But as Zeller (2002) has

Global bioregions

155

shown, San Diego and now Cambridge, Massachusetts genomics research capabilities are deeply penetrated by Novartis, in particular. This began with the signing of a deal for first refusal to access 50 per cent of the research output in immunology, neurological science, and cardiovascular diseases of The Scripps Research Institute in San Diego in 1992. Thereafter, in 2002 Novartis announced a new $250 million genomics research institute in San Diego, named the Genomics Institute of the Novartis Research Foundation (GNF). The 200 staff complemented inhouse research teams at institutes in Basel and New Jersey (more recently also Cambridge, Massachusetts; see below). Several GNF scientists also have faculty appointments at Scripps, and 17 per cent of postdoctoral researchers work with GNF scientists. GNF also gave rise to the Joint Center for Structural Genomics (JCSG) and the Institute for Childhood and Neglected Diseases (ICND) funded as consortia in San Diego by the US National Institutes of Health. In 2004 a new $350 million investment opened its doors, sustained by a further long-term commitment of $4 billion investment beyond that which was announced for Cambridge, Massachusetts by Novartis through its Novartis Institutes for Biomedical Research (NIBR). NIBR constitutes the primary pharmaceutical research arm in the company’s strategy of post-genomic drug discovery, concentrating on the key therapeutic areas of cardiovascular disease, diabetes, infectious diseases, functional genomics and oncology. From such globalized research and commercialization nodes or clusters, linked through networks of star research programmes that are largely resourced by public research funds, DBFs form to exploit the findings in early-stage commercialization activities. From these interactions licensing deals on patented knowledge are conducted such that DBFs receive milestone payments from big pharma, which then market and distribute biotechnologically derived drugs globally through their intra-corporate channels. This is a vivid example of the manner in which globalization is managed through extensive and intensive linkages unifying multinationals and clusters. In the papers by Lawton Smith (2004), Bagchi-Sen and Scully (2004) and particularly Finegold et al. (2004), these processes are anatomized in considerable detail.

THE NETWORKED HIERARCHY IN BIOTECHNOLOGY Three key things have been shown with implications for understanding of knowledge management, knowledge spillovers and the roles of collaboration and competition in bioregions. The first is that two kinds of proximity

156


are important to the functioning of knowledge complexes like biosciences in Boston and the northern and southern Californian clusters. These are geographical but also functional proximity (Rallet and Torre, 1998). The first involves, in particular, medical research infrastructure for exploration knowledge as well as venture capital for exploitation knowledge, that is for research on the one hand, and commercialization on the other. The second point is that where exploration knowledge infrastructure is strong, that nexus leads the knowledge management process, pulling more distant ‘big pharma’ governance elements behind it. Where, by contrast, exploitation knowledge institutions are stronger than exploration, they may, either as venture capital or ‘big pharma’, play a more prominent role. But in either case the key animator is the R&D and exploitation-intensive DBF. DBFs are key ‘makers’ as well as ‘takers’ of local and global spillovers; research institutions are more ‘makers’ than ‘takers’ locally and globally; while ‘big pharma’ is nowadays principally a ‘taker’ of localized spillovers from different innovative DBF clusters. It is then a global marketer of these and proprietorial (licensed or acquired) knowledge, generated with a large element of public financing but appropriated privately. For obvious reasons to do with scale, especially of varieties of financing of DBFs from big pharma on the one hand, and venture capitalists, on the other, we conclude that Boston, San Francisco and San Diego are the top US bioregions that also have the greater cluster characterization of prominent spinout from key knowledge centres, an institutional support set-up like Boston’s Massachusetts Biotechnology Council, San Diego’s Connect network and San Francisco’s California Healthcare Institute, and major investment from both main pillars of the private investment sector. We have attempted to access comparable data from many and diverse statistical sources that justify and represent the successful or potentially successful clusters from outside the USA, including the four lesser or unclustered of the seven US bioregions. Global cities like New York, London and even Tokyo have relatively large numbers of DBFs, but they are dotted around in isolation and have no established bioregional promotional bodies (such as BioCom in San Diego or the Massachusetts Biotechnology Council in Cambridge) rather than clustered close to key universities as in the listed bioregions. We can see from Table 5.1 the predominance of Boston and San Francisco on all indicators, and the differences between the former (also New York) and the Californian centres are strikingly revealed by these data. Boston’s life scientists generate on the order of $285 000 each per annum in National Institutes of Health research funding (New York’s generate some $288 000). San Diego’s considerably smaller number of life scientists generates $480 000 per capita, substantially more than in Northern California, where it is some $226 000. North Carolina, with

Global bioregions

Table 5.1

157

Core biotechnology firms, 2000: comparative global performance indicators

Location Boston San Francisco New York Munich Lund–Medicon San Diego Stockholm–Upp. Washington, DC Toronto Montreal Ral-Dur NC Zurich Cambridge Oxford Singapore Jerusalem Seattle

DBFs

Life scientists

VC($m)

Big pharma funding($m)

141 152 127 120 104 94 87 83 73 72 72 70 54 46 38 38 30

4980 3090 4790 8000 5950 1430 2998 6670 1149 822 910 1236 2650 3250 1063 1015 1810

601.5 1063.5 1730 400.0 80.0 432.8 90.0 49.5 120.0 60.0 192 57.0 250.0 120.0 200.0 300.0 49.5

800 (1996–2001) 400 (1996–2001) 151.6 (2000) 54 (2001) 300 (2002) 320 (1996–2001) 250 (2002) 360 (2000) n.a. n.a. 190 (2000) n.a. 105 (2000) 70 (2000) 88 (2001) n.a. 580 (2000)

Sources: NIH; NRC; BioM, Munich; VINNOVA, Sweden; Dorey (2003); Kettler and Casper (2000); ERBI, UK; Lawton Smith (2004); Kaufmann et al. (2003).

the second smallest number of life scientists in Table 5.1, scores highest at $510 000 per capita, although Seattle, at $276 000, is comparable to Boston, New York and San Francisco. How should we interpret these statistics? One way is to note the very large amounts of funding from ‘big pharma’ going especially to the Boston, and to a lesser extent New York and both Californian centres. Canada’s bioregional clusters challenge many elsewhere in the world regarding cluster development. Some of Europe’s are rather large on the input side (for example life scientists) but less so on the output side (for example funding, firms). The process of bioregional cluster evolution has occurred mainly through academic entrepreneurship supported by well-funded research infrastructure and local venture capital capabilities. In Israel, there is a highly promising group of bioregions including Rehovot and Tel Aviv as well as the main concentration in Jerusalem, where patents are highest although firm numbers are of lesser significance. The most striking feature of the global network of bioscientific knowledge transfer through exploration collaboration as measured by research

158

The capitalization of knowledge HMS

HMS HMS

HMS

Stanford Uni UCSF

HMS UT Scripps

Stanford Uni

Cam Uni Stanford Uni

HMS

HMS

Cam Un

Zurich Uni

Scripps

Scripps

RU

Stanford Uni

Cam Uni NYU

MIT

UCSD

UCSF

Stanford Uni UCSF

UBerkley

RU MIT Salk

RU UCSF

MIT

Stanford MIT

Cam Uni

NYU RU Stanford Uni

UBerkley

KI

HMS RU KI UCSD Salk

UCSF

Stanford Uni Cam Uni

KI RU UCSD UCSF Salk MIT

Scripps

Hebrew Uni UBerkley UT UCSF UCL

Scripps

Salk

UCSD UT

MIT

UCL UCSD

MIT UCSD

Salk

UBerkley

NYU UCL UCSF

Cam Uni KI KI RU

UT Zurich Uni Hebrew Uni

Zurich U

UCSD UBerkley

NYU

Cam Uni

Hebrew Uni UCSD

Salk

Notes: Results (log scale) are based on articles from the top two or three journals (with the highest impact factor provided by Web of Knowledge) of each sub-field, which are as follows: immunology: Annual Review of Immunology; Nature Immunology; Immunity; molecular biology & genetics: Cell Molecular; Journal of Cell Biology; microbiology: Microbiology and Molecular Biology Review; Annual Review of Microbiology; Clinical Microbiology Review; neuroscience: Annual Reviews Neuroscience; Trends Neuroscience; biotechnology & applied microbiology: Drug Discovery; Nature Biotechnology; cell & developmental biology: Annual Review of Cell and Developmental Biology; Advances in Anatomy, Embryology and Cell Biology; Anatomy and Embryology; biophysics & biochemistry: Annual Review of Biophysics; Current Opinions in Biophysics.

Figure 5.2

Publication shares of Bioscience Research Institutes in the seven main fields of bioscientific research

publication collaborations is, as we have seen, the dominance of US megacentres in such collaborations. However, within that hierarchy, which taken in the round by including examination and exploitation knowledge processes, puts Cambridge–Boston at the peak of the global bioscientific network hierarchy, specific institutional dominance of knowledge production is remarkable. This is shown in Figure 5.2, which ranks specific research institutions from among those benchmarked in Table 5.1 according to their share of publications in the seven main bioscientific fields. The

Global bioregions

159

predominance of a single, large and complex institution, Harvard Medical School (HMS), is noteworthy to say the least. Regarding immunology, HMS published 10 per cent of the articles in the three most cited journals, with Stanford at 4.5 per cent and UC San Francisco (UCSF) at 3.5 per cent. Karolinska Institute led Europe (tied with Salk Institute) at 1.2 per cent, followed by MIT at 0.8 per cent and Cambridge University at 0.5 per cent. HMS topped the share of molecular biology articles in the top three journals in 1998–2004 with 6 per cent of the relevant share, with MIT well behind at 1.7 per cent, UCSF at 1.4 per cent next, then Cambridge and Stanford Universities tied at 0.9 per cent. Of those scoring below 1 per cent, Salk Institute, Rockefeller University and UC San Diego (UCSD) follow. Next, in microbiology, HMS is again first at 6 per cent, followed by The Scripps Research Institute at 3.4 per cent, Stanford University at 3 per cent and Karolinska Institute at 2.4 per cent. For neuroscience, UCSF is first at 4.5 per cent but HMS is second at 4.1 per cent, Stanford is third at 3.8 per cent, Rockefeller follows at 3.3 per cent while Europe’s highest entrant is Cambridge University at 3 per cent. Karolinska Institute is tenth at 1.3 per cent. In biotechnology HMS is fifth (4.1 per cent) after University of Toronto (5.5 per cent) and The Scripps Research Institute (5.4 per cent) with Stanford and Cambridge universities equal at 4.5 per cent. These are followed by Zurich University (3.7 per cent), UCSF and Rockefeller tied at 2.8 per cent and UCSD at 1 per cent. In cell and development biology, HMS is the clear global leader with over 7 per cent leading journal share by citation and MIT trailing second at under 3 per cent. Finally, a relative weakness is biophysics and biochemistry, where HMS scores only 2.5 per cent and trails Cambridge University (UK), Scripps, Stanford, MIT, NYU and UC Berkeley. Thus in these seven key bioscientific fields alone, HMS is first four times, second once, fifth once and seventh once. Clearly, with or without control of house journals (which, it can be argued, significantly favours ‘home’ contributions), HMS is the leading quality publishing centre for biosciences in the world. Added to which, for Cambridge–Boston is the fact that MIT is second for molecular biology and cell biology, fourth in biophysics and biochemistry, and in the top ten for most of the rest. Thus we have seen inside the black box of global research excellence, gaining a detailed understanding of the broader network hierarchy of global knowledge transfer through co-publishing revealing the nature and extent of the global bioregions network. Recall from Figure 5.1 that it is clear that a global hierarchy exists in which there are five ‘megacentres’ (Cooke, 2003; Niosi and Bas, 2003). Measured by co-publishing activity, these are Boston–Cambridge, New York, San Francisco and San Diego. At the next level are London, Cambridge (UK) and Stockholm. Surrounding these,

160


at lower levels of co-publication activity, are global city locations like Paris, Tokyo and Toronto, but interspersed are other university towns like Lund, Uppsala and Oxford as well as lesser cities like Geneva, Zurich and Grenoble. Biotechnology thus does not fit the ‘global cities’ thesis (Sassen, 1994) particularly well. Other important but scarcely global cities like Jerusalem, Munich, Montreal and Singapore also enter at this third level of co-publishing intensity. Notice, though, that in the seven core fields of bioscientific publishing relatively few of the main publishing centres extend deeply into these lesser network nodes. They divide most asymmetrically between 11 globally key publishing institutes in North America, all bar University of Toronto in the USA, four in Europe and one, Hebrew University, Jerusalem in Asia. All the US top institutes are in the five co-publishing megacentres depicted in Figure 5.1. Despite dubious claims that San Diego7 is the top global cluster, it is clear that with respect to generation of new exploration knowledge it is Boston (including Cambridge MA) that in fact scores as the strongest biotechnology cluster. Moving to the opposite end of the global network hierarchy, as Finegold et al. (2004) make clear, Singapore’s case is one that runs strongly against the grain of the ‘clusters cannot be built’ thesis since it is almost entirely the product of state intervention. Four new research institutes in bioinformatics, genomics, bioprocessing and nanobiotechnology now exist at a cost of $150 million up to 2006. Public venture capital of $200 million has been committed to three bioscience investment funds to fund startups and attract FDI. A further $100 million is earmarked for attracting up to five globally leading corporate research centres. The Biopolis is Singapore’s intended world-class R&D hub for the georegion. It is dedicated to biomedical R&D activities and designed to foster a collaborative culture among the institutions present and with the nearby National University of Singapore, the National University Hospital and Singapore’s science parks. Internationally celebrated scientists have also been attracted, such as Nobel laureate Sidney Brenner, Alan Colman, leading transgenic animal cloning scientist from Scotland’s Roslin Institute, Edison Liu, former head of the US National Cancer Institute, and leading Japanese cancer researcher Yoshaki Ito. Next, for example, Tartu in Estonia is an ‘aspirational’ cluster in a rather less-developed form than that of Singapore. Noticeably, it does not for example yet appear in the co-publishing graphic presented in Figure 5.1. But as Raagmaa and Tamm (2004) show, it has certain historic bioscientific strengths to build upon as well as a burgeoning biomedical devices industry and emergent bioscientific and biotechnological research, if not yet DBFs that have exploited it. Nevertheless, as everywhere, government

Global bioregions

161

is keen to build up a biotechnology industry; Tartu is the leading university centre in Estonia for this kind of activity and new markets are often identified in relation to combinations of local expertise and even local environmental assets. The process leading to emergence of a leading East European industry cluster and the speed of its appearance, owing something partly to hype, no doubt, reveals an interesting case of possible failure but equally possible success at the point when its trajectory is really yet to be decided. Finally, in exploring both the cluster formation and globalization processes they imply, this analysis throws some light on the complexities of interactions between economic geography and science-based industry. This is so particularly regarding aspects of the networked interactions involved in knitting together the public and private spheres, basic research and commercial exploitation, and tacit and codified knowledge. There are good grounds now for looking upon biotechnology as a pioneer both of ‘open innovation’ and ‘Globalization 2’, in which ‘knowledgeable clusters’ rather than multinationals drive development more than hitherto. As well as being of intrinsic interest, these papers may offer hypotheses for other researchers investigating in other related or unrelated fields these challenging processes and the problems and opportunities for regional economic development policy that they entail.

CONCLUSIONS Thus biotechnology contains much of interest for academic and policy analysts interested in the distinctive ways in which clustering specific to biotechnology has occurred in different contexts. It also poses intriguing puzzles for academic specialists interested in biotechnology per se or more broadly in the clustering phenomenon. We see that local networks are common but that they may not be as large or as important as distant, even global, ones. The industry within which biotechnology of the type mainly discussed here is embedded, pharmaceuticals, does not appear to follow a Schumpeterian innovation or entrepreneurship pathway, being characterized by a double dependence of large corporations on SMEs (for R&D) and the inverse for financing, marketing and distribution. Large corporates merge but do not become out-competed and thus disappear entirely. Hence Novartis, formed from the merger of Ciba and Sandoz, retains the former as its opthalmics brand, with Sandoz its brand for generic drugs. Moreover, both are heavily dependent on public funding for basic research – the exploration knowledge that is the resource to which exploitation knowledge is applied. The latter includes knowledge of the kind

162


possessed by venture capitalists to facilitate commercialization of the few drug candidates that reach the market. Biotechnology thus seems to be more a Penrosian sector in that firm, including small-firm, capabilities determine industry organization (Penrose, 1959). Finally, to some extent echoing points made by Zucker et al. (1998), biotechnology is not a prime case of milieu effects dominating market effects. As they argue, scientific expertise has meant that since the beginning academics have wielded greater technological influence over firms than was the case with other industries. Firms leaned on academics’ terms and often, as DBFs, their ownership. As such, untraded interdependencies and even dense local network relationships, at the individual firm level, are less pronounced than might be expected given the obvious cluster concentrations in which biotechnology firms exist. Noticeable too are the ways in which public funding, policy and expertise can often be much more central to biotechnology clustering than has been recorded for other sectors. But of greatest theoretical interest is the way ‘open innovation’ hints at ‘Globalization 2’ where knowledgeable clusters shift the economic geography of industry organization towards themselves. Moreover, as the case of Cambridge–Boston and Harvard Medical School showed, some such spaces have begun to exert spatial knowledge monopoly effects that are already bringing ‘increasing returns’ as multinationals establish R&D facilities to capture ‘knowledge spillover’ advantages from cluster conventions of both ‘open science’ and ‘open innovation’.

ACKNOWLEDGEMENTS Thanks to Ann Yaolu for assistance with the scientometrics. Thanks also to Lennart Stenberg, Vinnova, for discussing various concepts and hypotheses. Finally, thanks to colleagues in the CIND, Circle and ISRN networks who heard these observations in preliminary form and encouraged their development into what I hope is a coherent analysis.

NOTES 1. The ‘exploration’ and ‘exploitation’ distinction was, of course, first made in organization science by March (1991). It is nowadays necessary to introduce the intermediary ‘examination’ form of knowledge to capture the major stages of the move from research to commercialization in biotechnology because of the long and intensive trialling process in both agro-food and pharmaceutical biotechnology. However, a moment’s reflection about other sectors suggests that this kind of knowledge and its organizational processes have been rather overlooked as it applies outside biotechnology (see Cooke, 2005).

Global bioregions

163

2. San Francisco’s megacentre capability in ICT gives its biotechnology competitive advantage in bioinformatics, screening, gene sequencing and genetic engineering software and technologies. 3. As shown below, a strong public role in cluster-building is also evident in Finegold et al.’s (2004) analysis of biopharmaceuticals developments in Singapore. 4. But disappointment has grown since the downturn at the only modest growth and drying pipeline dynamics associated with Germany’s policy-induced startup biotechnology businesses. Moreover, German high-tech entrepreneurship in general suffered a heavy blow with the demise of the Neuer Markt, its now defunct take on NASDAQ. 5. Unusually, the role of ‘big pharma’ is rather under-emphasized in this analysis of its relation to health biotechnology. This is not because large pharmaceuticals firms are unimportant in this context, for they clearly are. However, for exploration and even, increasingly, examination and some exploitation knowledge production they practise ‘open innovation’, as Chesbrough (2003) demonstrates for the case of Millennium Pharmaceuticals, a leading bioinformatics supplier that redesigned itself as a biotechnological drugs manufacturer through investment of ‘open innovation’ contract earnings from the likes of Monsanto and Eli Lilly. These practices are now emulated by specialist suppliers in industries like ICT, automotives and household care, according to the same author. This chimes with a more general hypothesis we can call ‘Globalization 2’, in which in a ‘knowledge economy’ the drivers of globalization become ‘knowledgeable clusters’ of various kinds. These exert an irresistible attraction for large corporates, who become ‘knowledge supplicants’ as their in-house R&D becomes ineffective and inefficient. They pay for, but no longer generate, leading-edge analytical knowledge for innovation. 6. As Owen-Smith and Powell (2004) show, ‘open science’ conventions in such clusters as Cambridge–Boston ‘irrigate’ the milieu with knowledge spillovers, giving to some clusters an element of ‘increasing returns’ from ‘spatial knowledge monopoly’ to a significant degree. 7. National Institute of Health (NIH) funding for medical and bioscientific research in Cambridge–Boston was in excess of $1.1 billion by 2000, $1.5 billion by 2002 and $2.1 billion in 2003. Cooke (2004) shows that it exceeded all of California by 2002, and by 2003 the gap widened to $476 billion ($2021 as against $1545 billion). Interestingly, this is a recent turnaround since the 1999 total of $770 million was marginally less than the amount of NIH funding passing through the northern California cluster in 1999, a statistic that only increased to $893 million in 2000. Thus Greater Boston’s supremacy is recent but definitive. San Diego’s NIH income includes that earned by Science Applications International Corporation. This firm is based in San Diego but performs most of its NIH research outside its home base as a research agent for US-wide clients. Thus it warrants mention but is excluded from totals calculated by this author. However, it is included in the Milken International report ‘America’s Biotech and Life Science Cluster’ (June 2004), which ranks San Diego the top US cluster. This oversight seriously weakens its claims for San Diego’s top US cluster position. Further reasons for rejecting the Milken Institute’s ranking of San Diego first as well as inclusion of questionable research funds are that the Institute deploys a spurious methodology based on research dollars per metropolitan inhabitant to promote San Diego’s ranking. Finally, the research was commissioned by local San Diego interests (Deloitte’s San Diego) and excludes ‘big pharma’ funding, on which San Diego performs less than half as well as Boston (Table 5.1).

REFERENCES Bagchi-Sen, S. and J. Scully (2004), ‘The Canadian environment for innovation and business development in the biotechnology industry; a firm-level analysis’, European Planning Studies, 12, 961–84.

164


Bazdarich, M. and J. Hurd (2003), Anderson Forecast: Inland Empire & Bay Area, Los Angeles, CA: Anderson Business School. Carlsson, B. (ed.) (2001) New Technological Systems in the BioIndustries: An International Study, London: Kluwer. Casper, S. and A. Karamanos (2003), ‘Commercialising science in Europe: the Cambridge biotechnology cluster’, European Planning Studies, 11, 805–22. Chesbrough, H. (2003), Open Innovation, Boston, MA: Harvard Business School Press. Coenen, L., J. Moodysson and B. Asheim (2004), ‘Nodes, networks and proximities: on the knowledge dynamics of the Medicon Valley biotech cluster’, European Planning Studies, 12, 1003–18. Cooke, P. (2002), ‘Rational drug design and the rise of bioscience megacentres’, presented at the Fourth Triple Helix Conference, ‘Breaking Boundaries – Building Bridges’, Copenhagen and Lund, 6–9 November. Cooke, P. (2003), ‘The evolution of biotechnology in three continents: Schumpeterian or Penrosian?’, European Planning Studies, 11, 757–64. Cooke, P. (2004), ‘Globalization of bioregions: the rise of knowledge capability, receptivity and diversity’, Regional Industrial Research Report 44, Cardiff: Centre for Advanced Studies. Cooke, P. (2005), ‘Rational drug design, the knowledge value chain and bioscience megacentres’, Cambridge Journal of Economics, 29 (3), 325–42. De la Mothe, J. and J. Niosi (eds) (2000), The Economics & Spatial Dynamics of Biotechnology, London: Kluwer. Dorey, E. (2003), ‘Emerging market Medicon Valley: a hotspot for biotech affairs’, BioResource, March, www.investintech.com, accessed 1 March 2004. DTI (1999), Biotechnology Clusters, London: Department of Trade & Industry. Feldman, M. and J. Francis (2003), ‘Fortune favours the prepared region: the case of entrepreneurship and the Capitol Region biotechnology cluster’, European Planning Studies, 11, 757–64. Finegold, D., P. Wong and T. Cheah (2004), ‘Adapting a foreign direct investment strategy to the knowledge economy: the case of Singapore’s emerging biotechnology cluster’, European Planning Studies, 12, 921–42. Freeman, C. (1987), Technology Policy & Economic Performance: Lessons from Japan, London: Pinter. Fuchs, G. (ed.) (2003), Biotechnology in Comparative Perspective, London: Routledge. Gregersen, B. (1992), ‘The public sector as a pacer in National Systems of Innovation’, in B.A. Lundvall (ed.), National Systems of Innovation, London: Pinter, pp. 129–45. Hall, P. and A. Markusen (eds) (1985), Silicon Landscapes, London: Allen & Unwin. Henderson, R., L. Orsenigo and G. Pisano (1999), ‘The pharmaceutical industry and the revolution in molecular biology: interactions among scientific, institutional and organisational change’, in D. Mowery and R. Nelson (eds), Sources of Industrial Leadership, Cambridge: Cambridge University Press, pp. 99–115. Johnson, J. (2002), ‘Valley in the Alps’, Financial Times, 27 February, p. 10. Kaiser, R. (2003), ‘Multi-level science policy and regional innovation: the case of the Munich cluster for pharmaceutical biotechnology’, European Planning Studies, 11, 841–58. Kaufmann D., D. Schwartz, A. Frenkel and D. Shefer (2003), ‘The role of location

Global bioregions

165

and regional networks for biotechnology firms in Israel’, European Planning Studies, 11, 823–40. Kettler, H. and S. Casper (2000), The Road to Sustainability in the UK & German Biotechnology Industries, London: Office of Health Economics. Lawton Smith, H. (2004), ‘The biotechnology industry in Oxfordshire: enterprise and innovation’, European Planning Studies, 12, 985–1002. Lawton Smith, H. and S. Bagchi-Sen (2004), ‘Guest editorial: innovation geographies; international perspectives on research, product development, and commercialisation of biotechnologies’, Environment & Planning C: Government & Policy, 22, 159–60. Lee, C. and M. Walshok (2001), Making Connections, Report to University of California, Office of the President. Lundvall, B.-Å. (ed.) (1992), National Systems of Innovation, London: Pinter. Maillat, D. (1998), ‘Interactions between urban systems and localised productive systems: an approach to endogenous regional development in terms of innovative milieu’, European Planning Studies, 6, 117–30. March, J. (1991), ‘Exploration and exploitation in organizational learning’, Organization Science, 2, 71–87. McKelvey, M. (1996), Evolutionary Innovation: The Business of Biotechnology, Oxford: Oxford University Press. Nelson, R. (ed.) (1993), National Innovation Systems, Oxford: Oxford University Press. Niosi, J. and T. Bas (2003), ‘Biotechnology megacentres: Montreal and Toronto regional systems of innovation’, European Planning Studies, 11, 789–804. OECD (2000), Health Data, Paris: OECD. Orsenigo, L. (1989), The Emergence of Biotechnology, New York: St Martin’s Press. Orsenigo, L., F. Pammolli and M. Riccaboni (2001), ‘Technological change and network dynamics: lessons from the pharmaceutical industry’, Research Policy, 30, 485–508. Owen-Smith, J. and W. Powell (2004), ‘Knowledge networks as channels and conduits: the effects spillovers in the Boston biotechnology community’, Organization Science, 15, 5–21. Parkes, C. (2004), ‘Job losses in Silicon Valley worse than first feared’, Financial Times, 25 March, p. 9. Penrose, E. (1959), The Theory of the Growth of the Firm, Oxford: Oxford University Press. Powell, W., K. Koput, J. Bowie and L. Smith-Doerr (2002), ‘The spatial clustering of science and capital: accounting for biotech firm-venture capital relationships’, Regional Studies, 36, 291–305. Raagmaa, G. and P. Tamm (2004), ‘An emerging biomedical business in a low capitalised country’, European Planning Studies, 12, 943–60. Rallet, A. and A. Torre (1998), ‘On geography and technology: proximity relations on localised innovation networks’, in M. Steiner (ed.), Clusters & Regional Specialisation, London: Pion, pp. 41–56. Ryan, C. and P. Phillips (2004), ‘Knowledge management in advanced technology industries: an examination of international agricultural biotechnology clusters’, Environment & Planning C: Government & Policy, 22, 217–32. Sassen, S. (1994), Cities in a World Economy, Thousand Oaks, CA: Pine Forge Press.

166


Schumpeter, J. (1975), Capitalism, Socialism & Democracy, New York: Harper & Row. Small Business Economics (2001), Special Issue: ‘Biotechnology in Comparative Perspective – Regional Concentration and Industry Dynamics’ (guest editors: Gerhard Fuchs and Gerhard Krauss), 17, 1–153. Wickelgren, I. (2002), The Gene Masters, New York: Times Books. Wolter, K. (2002), ‘Can the US experience be repeated? The evolution of biotechnology in three European regions’, Mimeo, Duisburg: Duisburg University. Zeller, C. (2002), ‘Regional and North Atlantic knowledge production in the pharmaceutical industry’, in V. Lo and E. Schamp (eds), Knowledge – the Spatial Dimension, Münster: Lit-Verlag, pp. 86–102. Zucker, L., M. Darby and J. Armstrong (1998), ‘Geographically localised knowledge: spillovers or markets?’, Economic Inquiry, 36, 65–86.

6.

Proprietary versus public domain licensing of software and research products Alfonso Gambardella and Bronwyn H. Hall

1.

INTRODUCTION

In the modern academic research setting, many disciplines produce software and databases as a by-product of their own activities, and also use the software and data generated by others. As Dalle (2003) and Maurer (2002) have documented, many of these research products are distributed and transferred to others using institutions that range from commercial exploitation to ‘free’ forms of open source. Many of the structures used in the latter case resemble the traditional ways in which the ‘Republic of Science’ has ensured that research spillovers are available at low cost to all. But in some cases, moves toward closing the source code and commercial development take place, often resulting either in the disappearance of open source versions or in ‘forking’, where an open source solution survives simultaneously with the provision of a closed commercial version of the same product. This has also created tensions between the reward systems of the ‘Republic of Science’ and the private sector, especially when the production of research software or the creation of scientific databases is carried out in academic and scientific research environments (see also Hall, 2004). As these inputs to scientific research have become more important and their value has grown, a number of questions and problems have arisen surrounding their provision. How do we ensure that incentives are in place to encourage their supply? How do market and non-market production of these knowledge inputs interact? In this chapter, we address some of these questions. We develop a framework that highlights the difficulties in sustaining the production of knowledge when it is the outcome of a collective enterprise. Since the lack of coordination among the individual knowledge producers is typically seen as the culprit in the underprovision of public knowledge, the latter can be sustained by institutional devices 167

168


that encourage such a coordination. A key idea of the chapter is that the Generalized Public License (GPL) used in the provision of open source software is one such mechanism. We then discuss another limitation in the production of this type of knowledge. To make it useful for commercial or other goals, one needs complementary investments (e.g. development costs). If the knowledge is freely available, there could be too many potential producers of such investments, which reduces the incentives of all of them to make the investments in the first place. Paradoxically, if the knowledge were protected, its access would be more costly, which might produce the necessary rents to enhance the complementary investments. But protecting upstream knowledge has many drawbacks, and we argue that a more effective solution is to protect the downstream industry products. Finally, we discuss how our framework and predictions apply to the provision of scientific software and databases. An example of the difference between free and commercial software solutions that should be familiar to most economists and scientific researchers is the scientific typesetting and word processing package TeX.1 This system and its associated set of fonts was originally the elegant invention of the Stanford computer scientist Donald Knuth, also famous as the author of the Art of Computer Programming, the first volume of which was published in 1969. Initially available on mainframes, and now widely distributed on UNIX and personal computer systems, TeX facilitated the creation of complex mathematical formulae in a word-processed manuscript and the subsequent production of typeset camera-ready output. It uses a set of text-based computer commands to accomplish this task rather than enabling users to enter their equations via the graphical WYSIWYG interface now familiar on the personal computer.2 Although straightforward in concept, the command language is complex and not easily learned, especially if the user does not use it on a regular basis. Although many academic users still write in raw TeX but work on a system with a graphical interface such as Windows, there now exists a commercial program, Scientific Word, which provides a WYSIWYG environment for generating TeX documents, albeit at a considerable price when compared to the freely distributed original. This example illustrates several features of the academic provision of software that we shall discuss in this chapter. First, it shows that there is willingness to pay for ease of software use even in the academic world and even if the software itself can be obtained for free. Second, the most common way in which software and databases are supplied to the academic market is a kind of hybrid between academic and commercial, where they are sold in a price-discriminatory way that preserves access for the majority of scientific users. Such products often begin as open source

Proprietary versus public domain licensing

169

projects directed by a ‘lead’ user, because the culture of open science is quite strong in the developers and participants. Nevertheless, they are eventually forced into the private sector as the market grows and nondeveloper users demand support, documentation, and enhancements to the ease of use. In the next section we discuss some basic aspects of the problem of creating incentives for the production of knowledge when many producers are involved. Section 3 discusses our analytic framework, which shows that without some kind of coordination, production of the public knowledge good (science or research software or database) is suboptimal, and that the GPL can solve the problem at least in part. Section 4 focuses on complementary investments. Sections 5 and 6 apply our framework to the specific setting where the knowledge being produced is software or a database that will be used by academic researchers and possibly also by private firms, using as an example a product familiar to economists, econometric software. We conclude by discussing some of the ways in which pricing can ameliorate the problem of providing these products to academic researchers. The Appendix develops the technical details of our model in Section 3.

2.

INCENTIVES FOR KNOWLEDGE PRODUCTION WITH MANY PRODUCERS

The design of incentive systems that reward inventors and knowledge producers and encourage dissemination of their output has been a familiar issue to economists and other scholars for a long time (e.g. Nelson, 1959; Arrow, 1962; Scotchmer, 1991). If anything, the issue has become more important today with the advent of the Internet and other computer networking methods. The principal effect of the increase in computer networking and Internet use is that it lowers the marginal cost of distributing codified knowledge to the point where it is essentially zero. This in turn has the potential to reduce incentives for production of such knowledge or to increase the demands of the producers for protection of their property rights to the knowledge. Hence there is a felt need to undertake additional efforts to understand the production of knowledge, and to think about new approaches to policy. To address these issues, we must first ask what motivates the producers of knowledge. Key factors identified in the literature are curiosity and a taste for science, money, the desire for fame and reputation, and, as a secondary goal, promotion or tenure (Stephan, 1996). The latter two goals are usually achieved via priority in publication, that is, being the first to get

170


a discovery into print. Although monetary income is clearly a partial motivation in the search for reputation and promotion, considerable evidence exists that for researchers in universities and public research organizations with some level of guaranteed income, the first motive − intellectual curiosity − is of overriding importance (e.g. Isabelle, 2004). For this type of researcher, the desire for financial rewards is often driven by the desire to fund their own scientific research (Lee, 2000) rather than by consumption per se. Scientists’ motivations also are coloured by the culture in which they are embedded, with traditional norms giving way to a more marketoriented view among some younger scientists today (Isabelle, 2004; Owen-Smith and Powell, 2001). Several scholars (e.g. Merton, 1957, 1968; David, 1993) have described the two regimes that allocate resources for the creation of new knowledge: one is the system of granting intellectual property rights (IPR), as exemplified by modern patent and copyright systems; the other is the ‘open science’ regime, as often found in the realm of ‘pure’ scientific research and sometimes in the realm of commercial technological innovation, often in infant industries (Allen, 1983). Today we also see this system to a certain extent in the production of free and open source software. The first system assigns clear property rights to newly created knowledge that allow the exclusion of others from using that knowledge, as well as the trading and licensing of the knowledge. As is well known, such a system provides powerful incentives for the creation of knowledge, at the cost of creating temporary monopolies that will tend to restrict output and raise price. Additionally, in such systems, the transaction costs of combining pieces of knowledge or building on another’s knowledge may be rather high, and in some cases achieving first- or even second-best incentives via ex post licensing may be impossible (Scotchmer, 1991). The use of other firms’ knowledge output will often require payment or reciprocal cross-licensing, which means that negotiation costs have to be incurred. Finally, obtaining IPR usually requires publication, but only of codified knowledge, and trade secrecy protection is often used in addition. The second set of institutional arrangements, sometimes referred to as the norms governing the ‘Republic of Science’, generates incentives and rewards indirectly: the creation of new knowledge is rewarded by increased reputation, further access to research resources, and possible subsequent financial returns in the form of increased salary, prizes, and the like (Merton, 1957, 1968). This system relies to some extent on the fact that individuals often invent or create for non-pecuniary reasons such as curiosity. Dissemination of research results and knowledge is achieved at relatively low cost, because assigning the ‘moral rights’ to the first publisher of an addition to the body of knowledge gives creators an incentive


171

to disseminate rapidly and broadly. Therefore, in this system the use of others’ output is encouraged and relatively cheap, with the cost being appropriate citation and possibly some reciprocity in sharing knowledge. But it is evident that this system cannot capture the same level of private economic returns for the creation of knowledge. Inventors must either donate their work or receive compensation as clients of public or private patrons.3 Hall (2004) highlights the tension that arises when these two systems come up against each other. For example, it is common for the difference in norms and lack of understanding of the potential partner’s needs and goals to produce breakdowns in negotiations between industry and academia. These breakdowns can have an economic as well as a cultural cause, as shown by Anton and Yao (2002) in a study of contracting under asymmetric information about the value of the knowledge to be exchanged. In addition, there is the simple fact that both systems rely on reciprocal behavior between the parties to a knowledge exchange, so that contracting between participants in the two difference systems becomes subject to misunderstanding or worse. This is illustrated by the reaction of the genomic industry in the USA when asked to take out licenses to university-generated technology: once the university starts acting like a private sector firm, there is a temptation to start charging it for the use of the outputs of industry research, and consequent negative effects on researchers who still believed themselves part of the ‘open science’ regime. In fact, notice should also be taken of an important variation of the ‘open science’ regime for the sharing of knowledge production outputs, one that has arisen many times in the development of industry throughout history: the free exchange and spillover of knowledge via personnel contact and movement, as well as reverse engineering, without resort to intellectual property protection. This has become known as the system for ‘collective invention’. Examples include the collective invention in the steel and iron industry described by Allen (1983) (see also Von Hippel, 1987), the development of the semiconductor industry in Silicon Valley (Hall and Ziedonis, 2001), the silk industry in Lyons during the ancien regime, described by Foray and Hilaire-Perez (2005), and the collective activities of communities of users who freely distribute information to the manufacturers (Harhoff et al., 2003). In these environments, most of which are geographically localized innovation areas with social as well as business relationships that build trust (or at least knowledge of whom to trust), the incentive system for the production and exchange of knowledge is somewhat different from that in either of the other two systems. The first and most obvious difference is that the production of ‘research’

172


in the industry setting is supported not by public or private patronage but by commercial firms that finance it by the sale of end products that incorporate their discovery. Because rewards come from the sale of products rather than information itself, as they do in the conventional IP-based system, the sharing of information about incremental innovations is motivated by different considerations than in the case of the open source regime. Although priority is not per se valuable except in the sense that it may confer lead time for production, shared knowledge, especially about incremental improvements to a complex product, is perceived to be useful and essential for the progress of the entire industry, including the firm that shares the knowledge. When an industry is advancing and growing rapidly, the desire to exclude competitors from the marketplace is not as strong as when an industry reaches maturity. An implication is that this form of free exchange of knowledge tends to collapse, or is unstable over time, as has happened in many of the historical examples. In the next section we try to capture this idea and discuss some conditions under which the academic or industry-based open source regime might break down.

3.

‘PUBLIC DOMAIN’ VERSUS ‘PROPRIETARY’ RESEARCH

Configuration of the Open Source Equilibrium When do the different systems of knowledge generation and sharing discussed in the previous section develop, and when they might be expected to break down? In this section we address these questions. To make our argument more precise we provide a simple formalization in the Appendix. Below we discuss the intuitions and the implications of our model. As discussed in the previous section, many researchers face a tradeoff. They can put a given research outcome in the public domain or seek private profits from it. As a stylized representation, in the former case they enjoy no economic rents, while in the latter they restrict public diffusion of their findings, seek property rights on them, and gain monetary income. We label the first mode ‘public domain’ (PD), and the second ‘proprietary research’ (PR). As also noted earlier, this framework encompasses many situations, such as academic scientists who could publish their research findings vis-à-vis holding patents or other property rights on them (Dasgupta and David, 1994); software developers who contribute to open source software as opposed to patenting their programs (Lerner and Tirole, 2002); user–inventors who transfer their inventions to the producers rather than protecting them as intellectual property and then selling


173

them (Von Hippel, 1988; Von Hippel and Von Krogh, 2003; Harhoff et al., 2003); communities of technologists who coordinate to share their ‘collective’ inventions, as opposed to keeping their knowledge secret (Allen, 1983; Nuvolari, 2004; Foray and Hilaire-Perez, 2005). Like any individuals, researchers gain utility from monetary income, but their utility also increases with the stock of public domain (PD) knowledge. Their benefits from this knowledge are from two sources: their own contributions and other contributions. First, they enjoy utility from the fact that they contribute to public knowledge. This is because they ‘like’ contributing to PD knowledge per se, or because they enjoy utility from a larger stock of public knowledge and hence they wish to contribute to its increase. There could also be instrumental reasons. Contribution to public knowledge makes their research visible, which provides fame, glory or potential future monetary incomes in the form of increased salary, funding for their research, or consultancy. Second, the researchers gain utility from the fact that others contribute to PD knowledge. Again this could be because they care about the state of public knowledge. In addition, a greater stock of public knowledge provides a larger basis for their own research, which implies that, other things being equal, they would like others contribute to it. We assume that the benefits from the contributions of other researchers to public knowledge will be enjoyed whether one works under PD or in the proprietary research (PR) regime. This implies that a researcher will operate under PD if the benefit that she enjoys from her public contribution is higher than the foregone monetary income from not privatizing her findings. In the Appendix we show that in equilibrium this is true of all the researchers who operate under PD, while the opposite is true of the researchers who operate under PR. In general, the equilibrium will involve a share of researchers operating under PD or PR that is between 0 and 1. The first prediction of our analysis is, then, that the two regimes can coexist, as we shall also see with some examples in the following sections. Our model also predicts that new profit opportunities common to all the researchers in a field reduce the share of PD researchers in equilibrium, while a stronger taste for research (e.g. because of particular systems of academic values) raises it. There is fairly widespread evidence that in fields like software or biotechnology there are pressures on academic researchers to place their findings in a proprietary regime. Also, our examples in the later sections show that shifts from academic to commercial software are more prominent when the market demand for the products increases, which raises the profitability of the programming efforts. Finally, there are several accounts of the fact that tension between industrial research and academic norms becomes higher if university access to IPRs is increased

174


(Cohen et al., 1998; Hall et al., 2001; Hertzfeld et al., 2006; Cohen et al., 2006). As these authors report, such tension has already been observed in the USA, as the latter country has pioneered the trend towards stronger IPRs and the use of intellectual property protection by universities, but it is becoming more pronounced in Europe as well, as European universities follow the path opened up by the US system (Geuna and Nesta, 2006). Collins and Wakoh (1999) describe similar changes in Japan, and show how the regime shift to patenting by universities is inconsistent with the previous system of collaborative research with industry in that country, implying increasing stress for the system. Instability of Open Source Production Our model also shows that the only way to get a stable equilibrium configuration with individuals operating under open sharing rules is when there is coordination among them. Otherwise, the sharing (cooperative) equilibrium tends to break down because some individuals find it in their interest to defect. The instability of the open sharing equilibrium is just an application of the famous principle by Mancur Olson (1971) that without coordination collective action is hard to sustain. Our contribution is simply to highlight that Olson’s insight finds application to the analysis of the instability of open systems. When many researchers contribute to PD knowledge, an individual deviation to PR is typically negligible compared to the (discrete) jump in income offered by a proprietary regime. Thus, individually, the researchers always have an incentive to deviate. Another way to see this point is to note that some of the tensions that are created in the open research systems can be attributed to the asymmetry between the open and the proprietary mode. The researchers shift to proprietary research only if it is individually profitable. By contrast, in the collective production of knowledge, a desirable individual outcome depends on the actions of others. In our framework this is because the individuals care about the fact that others contribute to the stock of knowledge, and because this may affect their benefits from their own contribution as well. As we show in the Appendix, this creates situations in which the lack of coordination produces individual incentives to deviate in spite of the fact that collectively the researchers would like to produce under PD. The intuition is that a group of individuals can produce a sizable increase in the stock of public knowledge if they jointly deviate from the PR regime. Thus, if there were commitment among them to stay within the PD rules, they could be better off than with private profits. In turn, this is because the larger the group of people who deviate in a coordinated fashion, the higher the impact on the public knowledge good, while the private profits,


175

which do not depend so heavily on the collective action, are not affected substantially by the joint movement of researchers from one regime to the other. But even if they all prefer to stay with the PD system, because of the larger impact of their PD contributions as a group, individually they have an incentive to deviate because if the others stay with PD, the individual deviation does not subtract that much from public knowledge, while it does produce a discrete jump in the individual’s private income. Since everyone knows that everyone else faces this tension, and could deviate, it will be difficult to keep the researchers under the PD system unless some explicit coordination or other mechanism is in place. Ultimately, this asymmetry in the stability of the two configurations suggests why there may be a tendency to move from public to private production of knowledge, while it is much harder to move back from private to public. The implication is that there is little need for policy if more proprietary research is desirable, as the latter is likely to arise naturally from the individual actions. By contrast, policy or institutional devices that could sustain the right amount of coordination is crucial if the system underinvests in knowledge that is placed in the public domain. Generalized Public License (Copyleft) as a Coordination Device The Generalized Public License (GPL) used in open source software can be an effective mechanism for obtaining the required coordination. As discussed by Lerner and Tirole (2002), inter alia, with a GPL the producer of an open source program requires that all modifications and improvements of the program be subject to the same rules of openness; most notably the source code of all the modifications ought to be made publicly available like the original program.4 To see how a GPL provides the coordination to solve the Mancur Olson problem, imagine the following situation. There is one researcher who considers whether to launch a new project or not. We call her the ‘originator’. She knows that if she launches the project, others may follow with additional contributions. The latter are the ‘contributors’. If the originator attaches a GPL to the project, the contributors can join only under PD. If no GPL is attached, they have the option to privatize their contribution. Of course, once (and if) the project is launched, the contributors always have the option not to join the project and work on some alternative activities. Given the expected behavior of the contributors, the originator will choose whether to launch the project or not. She also has potential alternatives. If she decides to launch it, she will choose whether to put her contribution under PD or PR, and if the former, she considers whether to attach a GPL to the project. We can safely rule out the possibility that the originator operates under PR and attaches a GPL

176


to the project. It will be odd to think that she can enforce open source behavior given that she does not abide by the same rules. The key implication of a GPL is that it increases the number of contributors operating under PD. The intuition, which we formalize in the Appendix, is simple. Without a GPL the contributors have three choices: work on the project under PD, or under PR, or not join because they have better alternatives. PD contributors to the project will still choose PD if a GPL is imposed. If they preferred PD over both PR and other alternatives, they will still prefer PD if the PR option is ruled out. Those who did not join the project will not join with a GPL either. They preferred their alternatives over PD and PR, and will still prefer them if PR is not an option. Finally, some of those who joined under PR will join under PD instead, while others who joined under PR will no longer join the project. As a result, a GPL reduces the total number of researchers who join the project, but raises the number of researchers working under PD. The reduced number of participants is consistent with the fact that the GPL is a restriction on the behavior of the researchers. However, this is a small cost to the public diffusion of knowledge because those who no longer participate would have not joined under PD. By contrast, the GPL encourages some researchers who would not have published their results without the GPL to do so. Given the behavior of the contributors, will the originator launch the project and issue a GPL? We know that the originator, like any other researcher, enjoys greater utility from a larger size of the public knowledge stock. At the same time, she enjoys utility from monetary income or, as we noted, from alternative projects. Here we want to compare her choice when she can employ a GPL vis-à-vis a world in which there is no GPL. With a GPL she knows that the number of contributors to public knowledge increases, which in turn increases the size of the expected public knowledge stock when compared to a no-GPL case. As a result, when choosing whether to launch the project under PD with a GPL, under PD and no GPL, under PR, or work on alternative projects, she knows that the GPL choice raises the future public knowledge stock in the area while not raising her monetary income from the project or her utility from alternatives. This makes it more likely that the originator will choose to work on the project under PD cum GPL. More generally, a GPL will increase the number of projects launched under PD and the size of the public knowledge contributions. To summarize, the way the GPL works is by giving rise to an implicit coordination among a larger number of researchers to work on PD. The originator knows that there will be researchers who would prefer PR but choose PD if the former opportunity is not available, while all those who


177

would choose PD will stick to it in any case. This enlarges the number of expected PD researchers, thereby placing greater advantages on the PD choice. Our intuition is that those with a strong taste for PD research will always work under PD, whether there is a GPL or not. By contrast, those with a high opportunity cost will never join the project. But those who have a small opportunity cost, and a weak taste for PD research, might contribute via PD if a GPL is introduced. The GPL then lures people who are on the border between doing PD research on the project or not. For example, a GPL may be crucial to enhance the participation under PD of young researchers, who do not have significant opportunity costs (e.g. because they do not yet have high external visibility), but who do not have a strong taste for PD research either, and hence would privatize their findings if it were profitable to do so. There might also be dynamic implications, for example the GPL helps young researchers to ‘acquire’ a taste for PD research. This might help create a system of norms and values for public research that could sustain the collective action. We leave a more thorough assessment of such dynamic implications to future research. Nature and Consequences of the GPL Coordination A GPL is most effective as a coordination device when the opportunity cost of the individual researchers, and the private profits from contributing to the project, are not positively correlated. Suppose that they were. This could arise because there is some common element between the two factors. For example, an individual researcher could be effective in commercializing knowledge in any field because he belongs to institutions (university or other) that encourage the commercialization of knowledge. In this case, the contributors to the project, who have low opportunity costs, also have low private profits from contributing to the project via PR. A GPL would not make a big difference because a very large fraction of the contributors to the project will do so under PD since their private rewards are low in any case. Hence a GPL induces few researchers to switch from PR to PD. In turn, this has a small effect on the choice of the originator to launch the project under PD vis-à-vis PR because the number of additional PD contributors with a GPL is small. By contrast, if they are not positively correlated, some of the contributors to the project, who have low opportunity cost, will have high private rewards from PR. They could be encouraged by a GPL to switch to PD. As a result, the number of PD contributors could be sizably different with a GPL, with implied greater opportunities for PD rather than PR research. The independence between the opportunity cost and the private rewards, as opposed to positive correlation, may be associated with the novelty of

178


the project. When the projects are in new areas, the opportunities of the individuals may change substantially, and the researchers who might profit the most from the new projects can be different from those who benefited in the old projects. New skills, or new forms of learning are necessary in the new fields, and the people who have made substantial investments in the old projects may have greater difficulties in the new areas (see, e.g. Levinthal and March, 1993). In these cases, researchers with low opportunity costs may instead find that they have great opportunities to commercialize knowledge in the new fields (high private rewards). Thus the GPL is more likely to be a useful coordination device when the project is in a new field rather than an incrementally different one from previous projects, and when it is socially desirable to run these projects under PD. Our mechanism relies on the fact that there is enforcement of the GPL. But can the copyleft system be enforced? In some settings people seem to abide by the copyleft rules, as Lerner and Tirole (2002) have noted, in spite of the lack of legal enforcement. In many situations, there may be a reputation effect involved when the copyleft agreement is not complied with. In this respect, the reason why a copyleft license may be useful is that without it, it may not be clear to the additional contributors whether the intention of the initial developers of the project was to keep it under PD or not. But if the will is made explicit, deviations may be seen as an obvious and explicit challenge to the social norms, and this may be sanctioned by the community. The GPL then acts as a signal that clears the stage of potential ambiguities about individual behavior and the respect of social norms. Even in science, if a researcher develops a certain result, others may build on it, and privatize their contributions. This might be seen as a deviation from the social norms. While this behavior could be sanctioned, according to the strength with which the norms of open science are embedded in and pursued by the community, with no explicit indication that the original contributor did not want future results from her discoveries to be used for private purposes, the justification for the sanctions or the need for them can be more ambiguous. A GPL removes ambiguity about the original intentions of the developers, and any behavior that contradicts the GPL is more clearly seen as not proper. This reduces privatization of future contributions compared to a situation with no GPL, increases the expectations that more researchers will make their knowledge public, and, other things being equal, creates greater incentives to make projects public in the first place. It is in this respect that we think that explicit indications of the norms may be a stronger signal than the mere reliance on the unwritten norms of open science or open source software. A related point is that the literature has typically been concerned with


179

the need to protect the private property of knowledge when this is necessary to enhance the incentives to innovate. The inherent assumption is that when it is not privately protected, the knowledge is by default public, and it enriches the public domain. Yet our model points out that this is not really true. The public nature of knowledge needs itself to be protected when commitments to the production of knowledge in the public domain are socially desirable. In other words, there is a need for making it explicit that the knowledge has to remain public, and this calls for positive actions and institutions to protect it. Not allowing for private property rights on some body of knowledge is not equivalent to assuming that the knowledge will be in the public domain. One may then need to assign property rights not just to private agents, but also to the public. For example, the IPRs are typically thought of as being property rights to private agents. But we also need to have institutions that preserve the public character of knowledge. The copyleft license is a beautiful example of this institutional device. A natural policy suggestion is therefore to make it legal and enforceable as copyright, patents and other private-based IPRs.

4.

COMPLEMENTARY INVESTMENTS IN OPEN SOURCE PRODUCTION

Another feature of traditional open source or academic software production that we alluded to in the introduction is that it normally requires additional investments that enhance the usefulness and value of the scattered individual contributions, or it simply requires investments to combine them. For example, while several individuals can contribute to the development of a whole body of scientific knowledge, there must be some stage at which the ‘pieces’ are combined into useful products, systems, or transferable knowledge. Some scientists, or most likely some specialized agents, i.e. academic licensing offices or firms, normally perform this function. A typical example is when scientific knowledge needs substantial downstream investments to become economically useful technologies or commercializable products. Thursby et al. (2001) report that this is often the case for university research outputs. The latter activities are normally performed by firms. In software, additional investments are often required to enhance the usability of the software for those who did not develop it, and to produce documentation and support. The need for additional investments in open source production, or more generally in tasks that rely on public domain knowledge, has some specific implications that we want to discuss in this section. The problem is that the (downstream) ‘assembling’ agent needs some

180


profits in order to carry out the investments that are necessary to produce the complementary downstream assets of the good. Since the downstream assembling agents are typically firms, we now refer to them as such. There are two issues. First, the firm needs to obtain some economic returns to finance its investment. Clearly, there are many ways to moderate its potential monopoly power so that the magnitude of the rents will be sufficient to make the necessary investments but not high enough to produce serious extra-normal profits. However, it would be difficult for the firm to obtain such rents if it operated under perfect competition, or if it operated under an open, public domain system itself. The second issue is more subtle. The firm uses the public domain contributions of the individual agents (software programmers, scientists etc.) as inputs in its production process. If these contributions are freely available in the public domain, and particularly if they are not available on an exclusive basis, many downstream firms can make use of them. As a result, the downstream production can easily become a free entry, perfectly competitive world, with many firms having access to the widely available knowledge inputs. If so, each firm could not make enough rents to carry out the complementary investments. This would be even harder for the individual knowledge producers, who are normally scattered and have no resources to cover the fixed set-up costs for the downstream investments. The final implication is that the downstream investments will not be undertaken, or they will be insufficient. Of course, there can be other factors that would provide the firms with barriers to entry, thereby ensuring that they can enjoy some rents to make their investments. However, in productions where the knowledge inputs are crucial (e.g. software), the inability to use them somewhat exclusively can generate enough threats of widespread entry and excessive competition to discourage the complementary investments. Paradoxically, if the knowledge inputs were produced under proprietary rules, the producers of them could charge monopoly prices (e.g. because they could obtain an exclusive license), or at least enjoy some positive price cost margins. This raises the costs of the inputs. In turn, this heightens barriers to entry in the downstream sector, and adjusts the level of downstream investment upward. In other words, if the inputs are freely available, there could be excessive downstream competition, which may limit the complementary investments. If they are offered under proprietary rules, the costs of acquiring the inputs are higher, which curbs entry and competition, and allows the downstream firms to make enough rents to carry out such investments.5 But the privatization of the upstream inputs has several limitations. For one, as Heller and Eisenberg (1998) have noted, the complementarity


181

among the ‘pieces’ of upstream knowledge produced by the different individuals can give rise to the so-called problem of the anti-commons. That is, after all the other rights have been collected under a unique proprietorship, the final owner of a set of complementary inputs can enjoy enormous monopoly power. This is because by withholding his own contribution, he can forestall the realization of the whole technology, especially when the complementarity is so tight that each individual contribution is crucial to make the whole system work. The possibility of ex post hold-up can discourage the effort to collect all the complementary rights ex ante, and therefore prevent the development of the technology. Another limitation of the privatization of the upstream inputs is the one discussed in the previous section. With copyleft agreements, more people can contribute to the public good. The decentralized nature of the process by which scientists or open source software producers operate has typically implied that the network of public contributors to a given field can be so large that the overall improvements can be higher than what can be obtained within individual organizations, including quite large ones. Some evidence that open source projects also increase the quality of software output has been supplied by Kuan (2002). One solution to the problem of paying for complementary downstream investment is allowing for property rights, and particularly intellectual property rights, on the innovations of the downstream producer. This would of course raise its monopoly power and therefore curb excessive competition. At the same time, it avoids attaching IPRs to pieces of upstream knowledge, thereby giving rise to the problems of the anticommons, or to reduced quality of the upstream knowledge. In addition, the downstream producer would enjoy rights on features of the innovation that are closer to his own real contribution to the project, that is the development of specific downstream investments. Clearly, this also implies that the IPRs thus offered are likely to be more narrow, as they apply to downstream innovations as opposed to potentially general pieces of knowledge upstream. At the same time, they are not likely to be as narrow as in the case of small individual contributions to an open software module or a minor contribution to a scientific field, which can give rise to the fragmentation and hold-up problems discussed earlier.

5.

ACADEMIC SOFTWARE AND DATABASES

In this section we draw some implications for the provision of scientific software and databases from the model and discussion in the previous two sections and then go on to discuss the possible modes in which they could

182


be provided. First, this type of activity is more likely to be privatized than scientific research itself because there is greater and more focused market demand for the product, because norms are weaker due to weaker reputation effects, and because there are more potential users who are not inventors (and do not participate in the production of the good). Second, there could easily be both public and private provision at the same time, because such an equilibrium can be sustained when there are different communities of researchers with different norms. Third, as the market for a particular product grows, privatization is likely simply because the individual’s discrete return to privatization has increased. Finally, when the components to a valuable good are produced under public domain rules, free entry in the downstream industry producing a final good based on those components implies too few profits for those undertaking investments that will enhance the value of the good. The final producers have to earn some rents to be able to make improvements beyond the mere availability of research inputs. The privatization of scientific databases and software has both advantages and disadvantages. With respect to the latter, David (2002) has emphasized the negative consequences of the privatization of scientific and technical data and information. One of the most important drawbacks is the increase in cost, sometimes substantial, to other scientists, researchers or software developers for use of the data in ways that might considerably enhance public domain knowledge. A second is that the value of such databases for scientific research is frequently enhanced by combining them or using them in their entirety for large-scale statistical analysis, both of which activities are frequently limited when they are commercially provided.6 Maurer (2002) gives a number of examples of privatized databases that have somewhat restricted access for academic researchers via their pricing structure or limitations on reuse of the data, such as Swiss-PROT, Space Imaging Corporation, Incyte and Celera. In this issue, David (2006) cites the case of the privatization of Landsat images under the Reagan Administration, which led to a tenfold increase in the price of an image. In terms of our model, the potential to privatize scientific and technical data and information implies that a smaller number of researchers will contribute to the public good, with implied smaller stock of public knowledge being produced, which frustrates the launch of projects undertaken under public diffusion rules. At the same time, a common argument in favor of the privatization of databases is that it helps in the development of a database-producing industry, and more generally of an industry that employs these data as inputs. A similar argument can be used more broadly for software. For example, the recent European Directive that defines the terms for the patenting of


183

software in Europe (European Commission, 2002) was largely justified by the argument that it would encourage the formation of a software industry in niches and specialized fields. Although it is sometimes true that exclusivity can have positive effects on the provision of information products, it is also true that there can be drawbacks like those suggested earlier (fragmentation of IPRs, little contribution to public domain knowledge, restricted access when welfare would be enhanced with unlimited access) to the privatization of knowledge inputs. At times, one can obtain similar advantages by allowing for the privatization of the outputs that can be generated using the database or software in question. That is, discovery of a useful application associated with a particular gene that is obtained by use of a genomic database is patentable in most countries. Or, in the case of the econometric software example used later in the chapter, consulting firms such as Data Resources, Inc. or Chase Econometrics marketed the results of estimating econometric models using software whose origins were in the public domain. Following our earlier argument, by allowing for the privatization of the downstream output we make it possible for the industry to obtain enough rents to make the necessary complementary investments, while avoiding the limitations of privatizations in the upstream knowledge. There are, however, limits to this particular strategy for ensuring that scientific databases and software remain in the public domain while downstream industries based on these freely available discoveries can earn enough profit to cover their necessary investments. The difficulty of course is that in the case of generally useful information products, a firm selling a particular product, one whose inputs is an upstream academic product, has no reason to undertake the enhancements to the upstream product that would make it useful to others, unless the firm can sell the enhanced product in the marketplace. But this is what we were trying to avoid, and what is ruled out by a GPL. In fact, we now turn to a discussion of an alternative way in which such goods can be provided. The production of information products including software and databases has always been characterized by large fixed costs relative to marginal cost, but the cost disparity has grown since the advent of the Internet. In practice, the only real marginal costs of distribution arise from two sources: the support offered to individual users (which in many cases has been converted into a fixed cost by requiring users to browse knowledge bases on the Web) and the congestion costs that can occur on web servers if demand is too great.7 Standard economic theory tells us that when the production function for a good is characterized by high fixed costs and low marginal costs, higher welfare can often be achieved by using discriminatory pricing, charging those with high willingness to pay

184


more in order to offer the good to others at lower prices, thus increasing the overall quantity supplied. The problem with applying this mechanism generally is the difficulty of segmenting the markets successfully and of preventing resale. In the case of academic software and databases, however, it is quite common for successful price-discriminating strategies to be pursued.8 There are several reasons for this: (1) segmentation is fairly easy because academics can be identified via addresses and institutional web information; (2) resale is difficult in the case of an information product that requires signing on to use it and also probably not very profitable; (3) the two markets (academic and commercial) have rather different tastes and attitudes toward technical support (especially towards the speed with which it is provided), so the necessary price discrimination is partly cost-based.

6.

CASE STUDY: ECONOMETRIC SOFTWARE PACKAGES

As an illustration of the pattern of software development in the academic arena, we present some evidence about a type of product familiar to economists that has largely been developed in a university research environment but is now widely available from commercial firms: packaged econometric software. Our data are drawn primarily from the excellent surveys on the topic by Charles Renfro (2003, 2004). We have supplemented it in places from the personal experience of one of the co-authors, who participated in the activity almost from its inception. The evidence supplied here can be considered illustrative rather than a formal statistical test of our model, since the sample is relatively small. To form a complete picture of the phenomenon of software and database commercialization in academia, it would be necessary to augment our study with other case studies. For example, see Maurer (2002) for a good review of methods of database provision in scientific research. Econometric software is very much a by-product of the empirical economic research activity, which is conducted largely at universities and nonprofit research institutions and to a lesser extent in the research departments of banks and brokerage houses. It is an essential tool for the implementation of statistical methods developed by econometric theorists, at least if these methods are to be used by more than a very few specialists. To a great extent, this type of software originated during the 1960s, when economists began to use computers rather than calculating machines for estimation, and for the first time had access to more data than could comfortably be manipulated by hand. The typical such package is implemented using a simple command


185

language and enables the use of a variety of modeling, estimating and forecasting methods on datasets of varying magnitudes. Most of these packages are now available for use on personal computers, although their origins are often a mainframe computer implementation. For a complete history of the development of this software, see Renfro (2003). Like most software, econometric software can be protected via various IP measures. The most important is a combination of copyright (for the specific implementation in source code of the methods provided) and trade secrecy (whereby only the ‘object’ code, or machine-readable version of the code, is released to the public). This combination of IP protection has always been available but has only become widely used during the personal computer era. Before that time, distributors of academic software usually provided some form of copyrighted source code for local installation on mainframes, and relied on the fact that acquisition and maintenance were performed by institutions rather than a single individual to protect the code. This meant that the source code could be modified for local use, but because the size of the potential market for ‘bootleg’ copies of the source was rather small, piracy posed no serious competitive threat. The advent of the personal computer, which meant that in many cases software was being supplied to individuals rather than institutions, changed this situation, and today the copyright-trade secrecy model is paramount.9 Thus it is possible to argue that developments in computing have made the available IP protection in the academic software sector stronger at the same time that the potential market size grew, which our model implies will lead to more defection from public domain to proprietary rules. In Table 6.1, we show some statistics for the 30 packages identified by Renfro. The majority (20 of the 30) have their origins in academic research, either supported by grants or, in many cases, written as a by-product of thesis research on a student’s own time.10 A further five were written specifically to support the modeling or research activities of a quasigovernmental organization such as a central bank. Only five were written with a specific commercial purpose in mind. Two of those five were forks of public domain programs, and in contrast to those of academic origin (whose earliest date of introduction was 1964 and whose average date was 1979), the earliest of the commercial programs was developed in 1981/82, a date that clearly coincides with the introduction of the non-hobbyist personal computer. Notwithstanding the academic research origin of most of these packages, today no fewer than 25 out of the 30 have been commercialized, with an average commercialization lag of nine years. Reading the histories of these packages supplied in Renfro (2003), it becomes clear that although many of them had more than one contributor, normally there was a ‘lead user’ who coordinated development, the

186

Table 6.1


Econometric software packages Total number of products

Number commercialized

Average lag to commercialization

Average date of introduction

Research grants or own research Quasigovernmental organization Private (for profit)

20

16

9.4

1979

5

4

16.4

1974

5

5

0.8

1984

Total or average

30

25

9.0

1979

Type of seed funding

identity of the ‘lead user’ occasionally changing as time passed. Most of the packages had their origins in the solution of a specific research problem (e.g. the development of LIMDEP for estimation of the Nerlove and Press logit model, or the implementation of Hendry’s model development methodology in PCGive), but were developed, often through the efforts of others besides the initial inventor, into more general tools. These facts clearly reflect the development both of computing technology and of the market for these kinds of packages. As predicted by our model, growth in the market due to the availability of personal computers and the growth of the economics profession as whole has caused the early largely open source development model of the 1960s to become privatized. Nevertheless, there remain five programs that are supplied for free over the Internet; of these, three had their origins before 1980 and the other two are very recent. As our model in Section 3 suggests, not all of the individuals in the community shift to the private system, and the share of PD activities can well be between 0 and 1. Interestingly, only one of the five is explicitly provided with a GPL attached. A quote from one of the author’s websites summarizes the motivation of those who make these programs available quite well: Why is EasyReg free? EasyReg was originally designed to promote my own research. I came to realize that getting my research published in econometric journals is not enough to get it used. But writing a program that only does the Bierens’ stuff would not reach the new generation of economists and econometricians. Therefore, the program should contain more than only my econometric techniques.


187

When I taught econometrics at Southern Methodist University in Dallas in the period 1991–1996, I needed software that my graduate students could use for their exercises. The existing commercial software was not advanced enough, or too expensive, or both. Therefore, I added the econometric techniques that I taught in class first to SimplReg, and later on to EasyReg after I had bought Visual Basic 3. Meanwhile, working on EasyReg became a hobby: my favorite pastime during rainy weekends. When I moved to Penn State University, and made EasyReg downloadable from the web, people from all over the world, from developing countries in Asia and Africa as well as from western Europe and the USA, wrote me e-mails with econometric questions, suggestions for additions, or just saying ‘thank you’. It appears that a lot of students and researchers have no access, or cannot afford access, to commercial econometrics software. By making EasyReg commercial I would therefore let these people down. There are also less altruistic reasons for keeping EasyReg free: (1) By keeping EasyReg free my own econometric work incorporated in EasyReg will get the widest distribution. (2) I will never be able to make enough money with a commercial version of EasyReg to be compensated for the time I have invested in it. (3) Going commercial would leave me no time for my own research.11

Indeed, the second statement suggests that one reason to leave the software in the public domain was that the researcher’s commercial profits were not large enough. Likewise, the third statement suggests that the researcher cared about research and this was an important reason for not privatizing it. This is suggestive of the fact that the individual displayed a relatively low utility of commercial profits vis-à-vis his preference for research, which in turn affected his choice of staying public. In sum, the model’s prediction that both private and public modes of provision can coexist when at least some individuals adhere to community norms is borne out, at least for one example. We also discussed explicitly the role of complementary services or enhanced features for non-inventor users in the provision of software. This is clearly one of the motivations behind commercialization, as was illustrated by the example of TeX. Table 6.2, which is drawn from data in Renfro (2004), attempts to give an impression of the differences between commercialized and non-commercialized software, admittedly using a rather small sample. To the extent that ease of use can be characterized by the full WIMP interface, there is no difference in the average performance of the two types of software. The main difference seems to be that the commercialized packages are larger and allow both more varied and more complex methods of interaction. Note especially the provision of a macro facility to run previously prepared programs, which occurs in 84 percent of the commercial programs, but only in two out of the five free programs.

188

Table 6.2


Comparing non-commercial and commercial software

Features Full windows, icons, menus interface (WIMP) Interactive use possible Macro files can be executed Manipulate objects with icons/menus Generate interactive commands with icons/menus

Share of noncommercial (%)

Share of commercial (%)

60

60

60 40 60 20

68 84 88 60

Such programs are likely to require more user support and documentation because of their complexity, which increases the cost of remaining in the PD system. In short, as our earlier discussion suggested, a commercial operation, which is likely to imply higher profits, also provides a greater degree of additional investments beyond the mere availability of the research inputs. To summarize, the basic predictions of our model, which are that participants in an open science community will defect to the private (IPusing) sector when profit opportunities arise (e.g. the final demand for the product grows, or IP protection becomes available) are confirmed by this example. We also find some support for the hypothesis that commercial operations are likely to undertake more complementary investments than pure open source operations. We do not find widespread use of the GPL idea in this particular niche market yet, although use of such a license could evolve. In the broader academic market, Maurer (2002) reports that a great variety of open source software licenses is in use, both viral (GPL, LPL) and non-viral (BSD, Apache-CMU). Finally, our model in Section 3 does not explicitly incorporate all the factors that are clearly important in the case of software and databases. Specifically, one area seems worthy of further development. We did not model the competitive behavior of the downstream firms in the database and software industries. In practice, in some cases, there is competition to supply these goods, and in others, it is more common for the good to be supplied at prices set by a partially price-discriminating monopolist. We report the evidence on price discrimination for our sample briefly here. Table 6.3 presents some very limited data for our sample of 30 econometric software packages. Of the 30, five are distributed freely and a further eight are distributed as services, possibly bundled with consulting (such sales are essentially all commercial); this is the ‘added value’ business


Table 6.3

189

Price discrimination in econometric software

Price-discriminate? By size or complexity Academic/commercial No discrimination NA Sold as a service Free Total

No. of packages 3 10 2 2 8 5 30

model discussed earlier. Of the remaining 17, we were able to collect data from their websites for 15. Of these, only two did not price-discriminate, three discriminate by the size and complexity of the problem that can be estimated, and ten by the type of customer, academic or commercial.12 A number of these packages were also offered in ‘student’ versions at substantially lower prices, segmenting the market even further. This evidence tends to confirm that in some cases, successful price discrimination is feasible and can be used to serve the academic market while covering some of the fixed costs via the commercial market. Although price discrimination is widely used in these markets, it does have some drawbacks as a solution to the problem of software provision. The most important one is that features important to academics or even programs important to academics may fail to be provided or maintained in areas where there is either a very small commercial market or no market, because their willingness to pay for them is much lower. Obviously this is not a consequence of price discrimination per se, but simply of low willingness to pay; the solution is not to eliminate price discrimination, but to recognize that PD production of some of these goods is inevitable. For example, a database of elementary particle data has been maintained by an international consortium of particle physicists for many years. Clearly such a database has little commercial market.

7.

CONCLUSIONS

Among the activities that constitute academic research, the production of software and databases for research purposes is likely to be especially subject to underprovision and privatization. The reason is that, like most

190


research activities, the public-good nature of the output leads to freeriding, but that the usual norms and rewards of the ‘Republic of Science’ are less available to their producer and maintainers, especially the latter. In this chapter we presented a model that illustrates and formalizes these ideas and we used the model to show that the GPL can be a way to ensure provision of some of these goods, at least when the potential producers also want to consume them. Although we have emphasized the beneficial role of the GPL as a coordination device for producing the public good, in these conclusions we also want to point out that the GPL is not a panacea that works in all situations, and one of those situations may indeed be the production of scientific software and databases. One reason is that in practice it is difficult to distinguish between the ‘upstream’ activities, which, as we discussed, ought to be produced under PD, and the ‘downstream’ ones. As we noted, the latter may entail important complementary investments. Therefore they could be more effectively conducted under private rules that enable the producers to raise the rents necessary to perform such investments. But the GPL ‘forces’ the contributors to work under PD rules. If one cannot properly distinguish between upstream and downstream activities, the downstream activities, with implied complementary investments, will also be subject to PD rules. This makes it more difficult to raise the resources to make the investments, with implied lower quality of the product. To return to the example of the introduction, the TeX Users’ Group reports the following on their website in answer to the FAQ ‘If TeX is so good, how come it’s free?’: It’s free because Knuth chose to make it so. He is nevertheless apparently happy that others should earn money by selling TeX-based services and products. While several valuable TeX-related tools and packages are offered subject to restrictions imposed by the GNU General Public License (‘Copyleft’), TeX itself is not subject to Copyleft. (http://www.tug.org)

Thus part of the reason for the spread of TeX and its use by a larger number of researchers than just those who are especially computeroriented is the fact that the lead user chose not to use the GPL to enforce the public domain, enabling commercial suppliers of TeX to offer easy-touse versions and customer support. The so-called ‘lesser’ GPL (LGPL) or other similar solutions can in part solve the problem. As discussed by Lerner and Tirole (2002), among others, the LGPL and analogous arrangements make the public domain requirement less stringent. They allow for the mixing of public and private codes or modules of the program. As a result, the outcome of the process is more likely to depend on the private incentives to make things


191

private or public, and this might encourage the acquisition of rents in the downstream activities. But following the logic of our model, as we allow for some degree of privatization, the efficacy of the license as a coordination mechanism is likely to diminish. We defer to future research a more thorough assessment of this trade-off. Here, however, we want to note that when the importance of complementary investments is higher, one would expect LPGL to be socially more desirable. The benefits of having the downstream investments may offset the disadvantage of a reduced coordination in the production of the public good. By contrast, when such investments are less important, or the separation between upstream and downstream activities can be made more clearly (and hence one can focus the GPL only on the former), a full GPL system is likely to be socially better.

ACKNOWLEDGMENTS Conversations with Paul David on this topic have helped greatly in clarifying the issues and problems. Both authors acknowledge his contribution with gratitude; any remaining errors and inconsistencies are entirely our responsibility. We are also grateful to Jennifer Kuan for bringing some of the open source literature to our attention. This chapter was previously published in Research Policy, Vol. 35, No. 6. 2006, pp. 875–92.

NOTES 1. 2. 3. 4.

5. 6.

This brief history of TeX is drawn from the TeX Users’ Group website, http://www.tug. org. In giving a simplified overview, we have omitted the role played by useful programs based on TeX such as LaTeX, etc. See the website for more information. WYSIWYG is a widely used acronym in computer programming design that stands for ‘What You See Is What You Get’. We can subsume both cases as instances of ‘patronage’ – self-patronage of the donated efforts is a special case of this. See David (1993) and Dasgupta and David (1994). There are many variants of a GPL, with different possibilities of privatizing future contributions. See, for example, Lerner and Tirole (2005). However, in this chapter we want to focus on some broad features of the effect of a GPL as a coordinating device, and therefore we simply consider the extreme case in which the GPL prevents any privatization of the future contributions. This argument should be familiar as it is the same as the argument used by some to justify Bayh–Dole and the granting of exclusive licenses for development by universities. The usual commercial web-based provision of data is based on a model where the user constructs queries to access individual items in the database, like looking up a single word in the dictionary. The pricing of such access reflects this design and is ill suited (i.e.

192

7.

8.

9.

10. 11. 12.

The capitalization of knowledge very costly) for researcher use in the case where research involves studying the overall structure of the data. This can be a real cost. The US Patent Office, which provides a large patent database free to the public at large on its web server, has a notice prominently posted on the website saying that use of automated scripts to access large amounts of these data is prohibited and will be shut down, because of the negative impact this has on individuals making live queries. Another type of academic information product deserves mention here, academic journals. The private sector producers of these journals face the same type of cost structure and have pursued a price discrimination strategy for many years, discriminating between library and personal use, and also among the income levels of the purchasers in some cases, where income level is proxied by country of origin. In principle, in the aftermath of the (1981) Diamond v. Diehr decision, patent protection might also be available for some features of econometric software. In this area, as in many other software areas, there is tremendous resistance to this idea on the part of existing players, perhaps because they are well aware of the nightmare that might ensue if patent offices were unacquainted with prior art in econometrics (as is no doubt currently the case). Unfortunately, it is not possible to identify precisely the nature of the seed money support for many of the packages from the histories supplied in Renfro (2003), other than the simple fact that the development took place at a university. This quotation is from Hermann Bierens’s website at http://econ.la.psu.edu/~hbierens/ EASYREG.HTM. The average ratio of commercial to academic price was 1.7. Assuming an iso-elastic demand curve with elasticity h and letting s = share of commercial (high-demand) customers, one can perform some very rough computations using the relationship ΔQ/Q = − h ΔP/P or (1 − s) = h 0.7. If h = 1, then the implied share of academic customers is 70 percent. If the share of academic customers is only 30 percent, then the implied demand elasticity is about 0.42.

REFERENCES Allen, R.C. (1983), ‘Collective invention’, Journal of Economic Behavior and Organization 4, 1–24. Anton, J.J. and D.A. Yao (2002), ‘The sale of ideas: strategic disclosure, property rights, and contracting’, Review of Economic Studies, 67, 585–607. Arrow, K. (1962), ‘Economic welfare and the allocation of resources for invention’, in R.R. Nelson (ed.), The Rate and Direction of Inventive Activity, Princeton, NJ: Princeton University Press, pp. 609–25. Cohen, W.M., R. Florida and L. Randazzese (2006), For Knowledge and Profit: University–Industry Research Centers in the United States, Oxford: Oxford University Press. Cohen, W.M., R. Florida, L. Randazzese and J. Walsh (1998), ‘Industry and the academy: uneasy partners in the cause of technological advance’, in R. Noll (ed.), The Future of the Research University, Washington, DC: Brookings Institution Press, pp. 171–99. Collins, S. and H. Wakoh (1999), ‘Universities and technology transfer in Japan: recent reforms in historical perspective’, University of Washington and Kanagawa Industrial Technology Research Institute, Japan. Dalle, J.M. (2003), ‘Open source technology transfer’, paper presented to the Third EPIP Conference, Maastricht, The Netherlands, 22–23 November.


193

Dasgupta, P. and P.A. David (1994), ’Toward a new economics of science’, Research Policy, 23, 487–521. David, P. (1993), ‘Knowledge, property, and the system dynamics of technological change’, in Proceedings of the World Bank Annual Conference on Development Economics 1992, Washington, DC: The World Bank, pp. 215–47. David, P.A. (2002), ‘The economic logic of open science and the balance between private property rights and the public domain in scientific data and information: a primer’, National Research Council Symposium on The Role of the Public Domain in Scientific and Technical Data and Information. National Academy of Sciences, Washington, DC. David, P.A. (2006), ‘A tragedy of the public knowledge “commons”? Global science, intellectual property and the digital technology boomerang’, unpublished paper, Stanford University, Stanford CA and All Souls College, Oxford University, Oxford UK. European Commission (2002), Draft directive on the patentability of computerimplemented inventions (20 February), available at http://www.europa.eu.int/ comm/internal_market/en/indprop/comp/index.htm. Foray, D. and L. Hilaire-Perez (2005), ‘The economics of open technology: collective organization and individual claims in the Fabrique Lyonnaise during the Old Regime’, in C. Antonelli, D. Foray, B.H. Hall and W.E. Steinmueller (eds), Essays in Honor of Paul A. David, Cheltenham, UK and Northampton, MA, USA: Edward Elgar, pp. 239–54. Geuna, A. and L. Nesta (2006), ‘University patenting and its effects on academic research: the emerging European evidence’, Research Policy, 35 (6), 790–807. Hall, B.H. (2004), ‘On copyright and patent protection for software and databases: a tale of two worlds’, in O. Granstrand (ed.), Economics, Law, and Intellectual Property, Boston/Dordrecht: Kluwer Academic Publishers, pp. 259–78. Hall, B.H. and R. Ziedonis (2001), ‘The determinants of patenting in the U.S. semiconductor industry, 1980–1994’, Rand Journal of Economics, 32, 101–28. Hall, B.H., A.N. Link and J.T. Scott (2001), ‘Barriers inhibiting industry from partnering with universities’, Journal of Technology Transfer, 26, 87–98. Harhoff, D., J. Henkel and E. von Hippel (2003), ‘Profiting from voluntary information spillovers: how users benefit by freely revealing their innovations’, Research Policy, 32, 1753–69. Heller, M.A. and R.S. Eisenberg (1998), ‘Can patents deter innovation? The anticommons in biomedical research’, Science, 280, 698–701. Hertzfeld, H.R., A.N. Link and N.S. Vonortas (2006), ‘Intellectual property protection mechanisms and research partnerships’, Research Policy, 35 (6), 825–38. Isabelle, M. (2004), ‘They invent (not patent) like they breathe: what are their incentives to do so? Short tales and lessons from researchers in a public research organization’, paper presented at the Third EPIP Workshop, Pisa, Italy, 2–3 April. Kuan, J. (2002), ‘Open source software as lead user’s make or buy decision: a study of open and closed source quality’, Stanford University, CA. Lee, Y.S. (2000), ‘The sustainability of university–industry research collaboration’, Journal of Technology Transfer, 25, 111–33. Lerner, J. and J. Tirole (2002), ‘Some simple economics of open source’, Journal of Industrial Economics, 50, 197–234. Lerner, J. and J. Tirole (2005), ‘The scope of open source licensing’, Journal of Law, Economics and Organization, 21, 20–56.

194


Levinthal, D. and J.G. March (1993), ‘The myopia of learning’, Strategic Management Journal, 14, 95–112. Maurer, S.M. (2002), ‘Promoting and disseminating knowledge: the public/private interface’, available at http://www7.nationalacademies.org/biso/Maurer_ background_paper.html, accessed 20 June 2008. Merton, R.K. (1957), ‘Priorities in scientific discovery: a chapter in the sociology of science’, American Sociological Review, 22, 635–59. Merton, R.K. (1968), ‘The Matthew effect in science’, Science, 159 (3810), 56–63. Nelson, R.R. (1959), ‘The simple economics of basic scientific research’, Journal of Political Economy, 77, 297–306. Nuvolari, A. (2004), ‘Collective invention during the British Industrial Revolution: the case of the Cornish pumping engine’, Cambridge Journal of Economics, 28, 347–63. Olson, M. (1971), ‘The Logic of Collective Action: Public Goods and the Theory of Groups’, Cambridge, MA: Harvard University Press. Owen-Smith, J. and W.W. Powell (2001), ‘To patent or not: faculty decisions and institutional success at technology transfer’, Journal of Technology Transfer, 26, 99–114. Renfro, C.G. (2003), ‘Econometric software: the first fifty years in perspective’, Journal of Economics and Social Measurement, 29, 1–51. Renfro, C.G. (2004), ‘A compendium of existing econometric software packages’, Journal of Economics and Social Measurement, 29, 359–409. Scotchmer, S. (1991), ‘Standing on the shoulders of giants: cumulative research and the patent law’, Journal of Economic Perspectives, 5, 29–41. Stephan, P.E. (1996), ‘The economics of science’, Journal of Economic Literature, 34, 1199–235. Thursby, J., R. Jensen and M.C. Thursby (2001), ‘Objectives, characteristics and outcomes of university licensing: a survey of major U.S. universities’, Journal of Technology Transfer, 26, 59–72. Von Hippel, E. (1987), ‘Cooperation between rivals: informal know-how trading’, Research Policy, 16, 291–302. Von Hippel, E. (1988), The Sources of Innovation, Oxford: Oxford University Press. Von Hippel, E. and G. von Krogh (2003), ‘Open source software and the private– collective innovation model: issues for organization science’, Organization Science, 209–23.


APPENDIX

195

A MODEL OF PUBLIC DOMAIN VERSUS PROPRIETARY RESEARCH

Set-up and Equilibrium The total (indirect) utility of a researcher is U = z + q·X(n − 1), where X(n − 1) is the stock of PD knowledge when n − 1 other researchers work under PD, and q ≥ 0 is a parameter that measures how much they care about the fact that others work under PD. Also, z = x(n) if they work under PD, and z = p if they work under PR, where x(n) is the utility that the researcher gains from her public contribution (assumed to be a function of the number of PD researchers n) and p is the utility from the monetary income. We assume that x(n) ≥ 0, and we make no assumption about the impact of n on x. There could be diminishing returns, i.e. a larger n implies smaller utility from one’s own contribution (e.g. because fewer important discoveries can be made), or there could be externalities, i.e. x increases with n, or both. Note that we assume that the researchers enjoy the public contribution of the others even if they work under PR. We could make more complicated assumptions, for example the impact of X(n − 1) on utility is different according to whether the individual operates under PD or PR, but this will not affect our main results. A researcher will produce under PD if p ≤ x(n). We assume that the individuals are heterogeneous in their preferences of PD versus PR, i.e. [p – x(n)] ~ F(·|n), where the distribution function F depends on n because of x(n). In principle, the support of p − x(n) is the whole real line. The share of individuals working under PD is then F(0|n), and the equilibrium number of researcher ne working under PD is defined by F(0|ne) = ne/N, where N is the total number of researchers in the community. This condition says that in equilibrium the share of researchers working under PD is equal to the share of researchers whose utility from p is not larger than the utility of their contribution x(ne) to PD. Figure 6A.1 depicts our equilibria. Point E in Figure 6A.1 is an equilibrium because if the number of researchers working under PD increases beyond ne, the share of researchers with p − x(n) ≤ 0 increases by less than the share of researchers working under PD. But this is a contradiction because for some of the researchers who have moved to PD it was not profitable to do so. The reasoning is symmetric for the deviations from PD to PR in equilibrium. Stability of the equilibrium requires that the F(ne) curve cuts the ne/N line from above. This ensures that whenever an individual deviates from the equilibrium, moving from PD to PR, the share of individuals with p ≤ x(ne − 1), i.e., those who find it profitable to operate under PD, is higher than the actual share of individuals working

196


F(n); n/N

F(n); n/N

1

1 n/N E

n/N F(n)

ne

F(n)

E1 N

One stable equilibrium (E)

Figure 6A.1

E2

ne

ne

N

Two stable equilibria (E1 and E2)

Equilibria

under PD after the move, i.e. (ne − 1)/N. Hence the move is not profitable. Similarly, whenever an individual moves from PR to PD in equilibrium, the share of researchers with p ≤ x(ne + 1) becomes smaller than the share of researchers who now work under PD, i.e. (ne + 1)/N. The stability conditions are then F(0| ne − 1) > (ne − 1)/N and F(0|ne + 1) < (ne + 1)/N. Multiple equilibria are also possible. There may be more than one ne that satisfies (1) with F(n) cutting n/N from above, as shown by Figure 6A.1. The share of researchers working under PD decreases if the economic profitability of research increases relatively to the researchers’ utility from their public contributions. This can be thought of as a first-order stochastic downward shift in F(·) which would stem from an increase in p − x(n) for all the individuals. Likewise, a stronger taste for research would be represented by an upward shift in F as p − x(n) decreases for all the individuals. This raises ne. Instability of PD Knowledge Production To see why the production of knowledge under PD is unstable, suppose that ne is an equilibrium and v researchers working under PR coordinate to work under PD. If ne is an equilibrium, then p > x(ne + v), for at least one of these researchers, otherwise ne + v would be an equilibrium. Yet, it is possible that x(ne + v) + q∙X(ne + v − 1) > p + q·X(ne − 1) for all the v researchers; i.e. if the v researchers coordinate, they are better off than in equilibrium. To see this recall that x(n) ; X(n) − X(n − 1). Therefore x(ne + v) + q∙X(ne + v − 1) – q∙X(ne − 1) = x(ne + v) + qSvj51x (ne 1 v 2 j) . But the expression in the summation sign is non-negative because x ≥ 0. Hence this expression can be greater than p in spite of the fact that p > x(ne + v).


197

If the v researchers working under PR coordinate to work under PD, the system is unstable because at least one of them can find it profitable to deviate since he exhibits p > x(ne + v). Once he deviates, at least one of the remaining v − 1 researchers has an incentive to deviate because ne + v − 1 is not an equilibrium, and so on until all the v researchers have deviated. At that point nobody else has an incentive to deviate because ne is an equilibrium. The GPL Model Let B be an opportunity cost faced by an originator and any potential contributor to a project, with [B − x(n)] ~ G(·|n). We assume that even if the contributors do not join the project, they still enjoy qX(n) from the project if n researchers work on it under PD. That is, their indirect utility is B + qX(n). Let nG and nNG be the equilibrium number of contributors joining the project under PD if the originator launches the project, works under PD and attaches a GPL or not. Finally, let | n NG be the number of contributors under PD if the originator launches the project under PR (and cannot attach a GPL to it). If the originator launches the project, works under PD and does not issue a GPL, the contributors working under PD will exhibit p ≤ x(nNG + 1) and B ≤ x(nNG + 1). If G(·) is the joint distribution function of p − x and B − x, in equilibrium we have G(0, 0 | nNG) = nNG/N. With no GPL, the condition becomes less restrictive, because only B ≤ x is required. As a result, with a GPL, G(0|nNG) > nNG/N. This will induce some researchers to join under PD, because the share of researchers with B ≤ x is smaller than the share of researchers actually working under PD, that is, nNG/N. Provided that the stability conditions discussed above hold, the movement towards PD will stop at nG such that G(0|nG) = nG/N. This implies nG ≥ nNG; i.e. a GPL induces more researchers to work under PD. If the originator launches the project under PR, the contributors can |NG) ≤ 0 and B − x still join under PD. These will be all those with p − x(n NG | (n ) ≤ 0. The difference with the previous no-GPL case is only that the originator does not join the project under PD. More generally, the same reasoning as above applies here, and nG ≥ | n NG. As a matter of fact, if n is NG NG | large enough, n ≈ n . Given the behavior of the contributors, will the originator who launches the project working under PD issue a GPL? With PD-GPL his utility will be x(nG + 1) + qX(nG). With PD and no GPL it will be x(nNG + 1)+qX(nNG). By using the fact that x(n) ; X(n) − X(n − 1), the former will be higher than the latter if X(nG + 1) − (1 − q)X(nG) ≥ X(nNG + 1) − (1 − q)X(nNG). A sufficient condition for this inequality to hold is that q ≥ 1. This follows

198

Table 6A.1


Comparing researcher actions with and without a GPL

Researchers’ set

Action under no GPL

Action under GPL

B ≤ x; p ≤ x B ≤ x; p ≥ x (p ≥ B) B ≥ x; p ≤ x (p ≤ B) B ≥ x; p ≥ x

Join under PD Join under PR Not join Join under PR if p ≥ B Not join if p ≤ B

Join under PD Join under PD Not join Not join

from nG ≥ nNG and the fact that X(n) increases with n, which in turn follows from x(n) ≥ 0. Thus, if the originator chooses to work under PD, setting a GPL will be a dominant strategy unless q is close to zero (i.e. the impact of the others’ behavior is not that important) and some special conditions occur. For simplicity, we assume that q is large enough, and therefore choosing a GPL always dominates when the originator chooses PD. |NG). As a result, If the originator chooses PR, his utility will be p + qX(n the originator will choose to work on the project under PD (and issue a GPL) if x(nG + 1) + q ∙ X(nG) ≥ B, and x(nG + 1) + q ∙ X(nG) ≥ p + q ∙ |NG). If there were no GPL, the condition would be the same with nNG in X(n lieu of nG. Since nG ≥ nNG, with no GPL the condition becomes more restrictive. As a result, the possibility to use a GPL implies not only that more researchers will join under PD, but also that more projects will be launched under PD with a GPL. As discussed in the text, the GPL is least effective when there is a strong positive correlation between B and p. This implies that many individuals with small B tend to have a small p as well. As a result, the restriction p ≤ x associated with B ≤ x does not restrict the set of PD researchers much more than B ≤ x alone, which means that nG is close to nNG, and the additional set of PD researchers created by the GPL is not large. In turn, this implies that the GPL does not encourage a more intensive coordination than without it. The GPL raises the number of contributors working under PD in spite of the fact that the total number of contributors to the project decreases. To see this, assume for simplicity that x is roughly constant with respect to n, so that x(nNG) ≈ x(nG) ; x. Consider Table 6A.1, which shows that, with a GPL, some researchers who joined under PR switch to PD, while the opposite is not true. The researchers who no longer join the project with the GPL are only those who joined under PR. Thus they do not affect n in equilibrium.

PART II

Triple helix in the knowledge economy

7.

A company of their own: entrepreneurial scientists and the capitalization of knowledge Henry Etzkowitz

INTRODUCTION: THE ACADEMIC WORLD HAS TURNED The capitalization of knowledge is an emerging mode of production (Etzkowitz, 1983). Until the past few decades, a sceptical view of firm formation was the taken-for-granted perspective of most faculty members and administrators at research universities. Since 1980, an increasing number of academic scientists have broadened their professional interests, from a single-minded interest in contributing to the literature, to making their research the basis of a firm. Formerly largely confined to a specialized academic sector, firm formation has spread to a broad range of universities: public and private; elite and non-elite. In addition, a complex web of relationships has grown up among university-originated startups in emerging industries and older and larger firms in traditional industries. Often the same academic scientists are involved with both types of firms, managing a diversified portfolio of industrial interactions (Powell et al., 2007). An entrepreneurial science model, combining basic research and teaching with technological innovation, is displacing the ‘ivory tower’ of knowledge for its own sake. In the mid-1980s, a faculty member at Stanford University reviewed his colleagues’ activities: ‘In psychiatry there are a lot of people interested in the chemistry of the nervous system and two of them have gone off to form their own company.’ Another Stanford professor, during the same period, estimated that In electrical engineering about every third student starts his own company. In our department [computer science] it’s starting as well. That’s a change in student behaviour and faculty acceptance because the faculty are involved in companies and interacting a lot with companies and the attitude is . . . we talk to them, we teach them. Why not try it . . . this is my experience. 201

202


While still significant, the ‘barrier to entry’ to firm formation is decreasing, especially as universities develop mechanisms to assist the process. Until quite recently, pursuing the ‘endless frontier’ of basic research was the primary ideological justification of elite US academic institutions. Harvard University was the ideal, with numerous schools identifying themselves as the ‘Harvard’ of their respective regions. With an entrepreneurial mode increasingly followed at Harvard, and at academic institutions that model themselves upon it, the prediction that MIT would eventually conform to the ivory tower mode has been disconfirmed (Geiger, 1986). Instead, the reverse has occurred as universities take up the ‘land grant’ mission of regional economic development and capitalization of knowledge, MIT’s founding purpose (Etzkowitz, 2002). This chapter discusses the impetuses to firm formation arising from the nature of the US research university system and its built-in drivers of transition to an entrepreneurial academic model.

A COMPANY OF THEIR OWN The relatively new existence of regularized paths of academic entrepreneurship, as a stage in an academic career or as an alternative career, is only of interest to some faculty; others prefer to follow traditional career paths. However, for some members of the professoriate, participation in the formation of a firm has become an incipiently recognizable stage in an academic career, located after becoming an eminent academic figure in science. For others, typically at earlier stages of their career, either just before or after being granted permanent tenure, such activity may lead to a career in industry outside the university. As one faculty member put it: ‘Different people I have known have elected to go different ways . . . some back to their laboratories and some running technology companies. You can’t do both.’ This difficulty has not prevented other professors from trying. A typical starting point is recruitment to the Scientific Advisory Board of a firm. The discussions that take place in this venue typically include the business implications of the firm’s research strategy, providing the neophyte academic entrepreneur with the equivalent of Entrepreneurial Science 101. The approach of leaving it up to the technology transfer office to find a developer and marketer for a discovery precisely met the needs of many faculty members, then and now, who strictly delimit their role in putting their technology into use. A faculty member delineated this perspective on division of labour in technology transfer: ‘It would depend on the transfer office expertise and their advice. I am not looking to become a businessperson. I really am interested in seeing if this could be brought into the market.

Entrepreneurial scientists and the capitalization of knowledge

203

I think it could have an impact on people’s lives. It is an attractive idea.’ This attitude does not necessarily preclude a startup firm but it does exclude the possibility that the faculty member will be the lead entrepreneur. A stance of moderate involvement is becoming more commonplace, with scientists becoming knowledgeable and comfortable operating in a business milieu while retaining their primary interest and identity as academic scientists. A faculty member exemplifying this approach expressed the following view: In science you kind of sit down and you share ideas . . . There tends to be a very open and very detailed exchange. The business thing when you sit down with somebody, the details are usually done later and you have to be very careful about what you say with regard to details because that is what business is about: keeping your arms around your details so that you can sell them to somebody else, otherwise there is no point.

Faculty are learning to calibrate their interaction to both scientific and business needs, giving out enough information to interest business persons in their research but not so much so that a business transaction to acquire the knowledge becomes superfluous. Another researcher said, ‘I am thinking about what turns me on, in terms of scientific interest and the money is something if I can figure out how to get it then it is important but it is certainly not the most important thing to me.’ The primary objective is still scientific; business objectives are strictly secondary. There has been a significant change of attitude among many faculty members in the sciences towards the capitalization of knowledge. Three styles of participation have emerged, reflecting increasing degrees of industrial involvement. These approaches can be characterized as (1) hands off, leaving the matter entirely to the technology transfer office; (2) knowledgeable participant, aware of the potential commercial value of research and willing to play a significant role in arranging its transfer to industry; and (3) seamless web, integration of campus research group and research programme of a firm. Of course, many faculty fit in the fourth cell of ‘no interest’ or non-involvement. These researchers are often referred to under the rubric of the federal agency that is their primary source of support, as in ‘She is an NIH person’. While still a minority interest of a relatively small number of academics at most universities, the prestige of an entrepreneurial undertaking has risen dramatically. Entrepreneurial Scientists A significant number of faculty members have adopted multiple objectives: ‘To not only run a successful company . . . and start a centre here

204


[at the university] that would become internationally recognized but to retain their traditional role as individual investigator’, directing a research group. An ideal-typical entrepreneurial scientist held that the ‘interaction of constantly going back and forth from the field, to the university lab, to the industrial lab, has to happen all the time’. These relationships involve different levels of commitment (financial and otherwise) by industrial sponsors, including the involvement of industrial sponsors in problem selection and research collaboration. As industrial sectors and universities move closer together, informal relationships and knowledge flows are increasingly overlaid by more intensive, formal institutional ties that arise from centres and joint projects. Firms formed by academics have been viewed in terms of their impact on the university but they are also ‘carriers’ of academic values and practices into industry and, depending upon the arrangements agreed upon, a channel from industry back to the university. In these latter circumstances, traditional forms of academic–industry relations, such as consulting and liaison programmes that encourage ‘knowledge flows’ from academia to industry, become less important as an increasing number of large firms acquire academic startups for sources of new products (Matkin, 1990). The conduct of academic science is also affected by a heightened interest in its economic potential. As companies externalize their R&D, they want more tangible inputs from external sources such as universities. As one close observer from the academic side of the equation put it, ‘From the point of view of the company, they tend to want a lot of bang for the buck . . . [they] tend to not get involved in Affiliates programs precisely because they can’t point to anything.’ The growth of centres and the formation of firms from academic research have had unintended consequences that have since become explicit goals: the creation of an industrial penumbra surrounding the university as well as an academic ethos among firms that collaborate with each other in pre-competitive R&D through joint academic links. In 2001, for example, 626 licences from US university technology transfer offices formed the basis of 494 startup firms (AUTM, 2005). Industry–University Relations Transformed From an industrial perspective, universities have been viewed as a source of human capital – future employees – and, secondarily, as a source of knowledge useful to the firm. In this view, the academic and industrial spheres should each concentrate on their respective functions and interact across distinct, strongly defended, boundaries. The hydraulic assumptions of knowledge flows include reservoirs, dams and gateways that facilitate and regulate the transmission of information between institutional spheres

University input


Quasi-firm research

High-tech growth firm

Spin-offs

Mid- and low-tech firms

205

Industry contribution Figure 7.1


with distinctly different functions (e.g. academia: basic research; companies: product development). From this perspective, what industry needs from academic researchers is basic research knowledge; therefore universities should focus on their traditional missions of research and education, their unique function (Faulkner and Senker, 1995). However, the organization of academic research, especially in the sciences, has been transformed from a dyadic relationship between professor and research student to a research group with firm-like qualities, even when the objective is unchanged (see Figure 7.1). The emergence of research findings with recognizable commercial potential, on the one hand, but a gap between potential and demonstrated utility that typically requires a bridge in the form of a startup firm. Such a firm may develop into a full-fledged commercial firm but in many cases it is purchased by a larger firm once product viability is demonstrated. Indeed, this outcome is often the firm founders’ objective. Intensive relationships occur with a group of firms that have grown out of university research and are still closely connected to their originary source and with a third group that, given the rapid pace of innovation in their industrial sector, have externalized some of their R&D and seek to import technologies or engage in joint R&D programmes to develop them (Rahm, 1996). In a fourth group of companies with little or no R&D capacity, relations with academia, if any, will also be informal through engaging an academic consultant to test materials or trouble-shoot a specific problem.

206


THE ORIGINS OF ENTREPRENEURIAL SCIENCE The formation of firms out of research activities occurred in the late nineteenth century at Harvard, as well as at MIT, in the fields of industrial consulting and scientific instrumentation (Shimshoni, 1970). However, these commercial entities were viewed as anomalies rather than as a normal outcome of academic research. In recent decades, an increasing number of academic scientists have taken some or all of the steps necessary to start a scientific firm by writing business plans, raising funds, leasing space, recruiting staff and so on (Blumenthal et al., 1986a; Krimsky et al., 1991). These studies probably underestimate the extent of faculty involvement, especially in molecular biology. For example, in the biology department at MIT, where surveys identified half the faculty as industrially involved in the late 1980s, an informant could identify only one department member as uninvolved at the time. While the model of separate spheres and technology transfer across strongly defined boundaries is still commonplace, academic scientists are often eager and willing to marry the two activities, nominally carrying out one in their academic laboratory and the other in a firm with which they maintain a close relationship. Thus, technology transfer is a two-way flow from university to industry and vice versa, with different degrees and forms of academic involvement: ● ● ●

the product originates in the university but its development is undertaken by an existing firm; the commercial product originates outside of the university, with academic knowledge utilized to improve the product, or the university is the source of the commercial product and the academic inventor becomes directly involved in its commercialization through establishment of a firm.

Potential products are often produced as a normal part of the research process, especially as software becomes commonplace in collecting and analysing data. As a faculty member commented in the mid-1980s, ‘In universities we tend to be very good at producing software, [we] produce it incidentally. So there is a natural affiliation there. My guess is a lot of what you are going to see in university–industry interaction is going to be in the software area.’ In the 1990s this phenomenon spread well beyond the research process, with software produced in academia outside of the laboratory, and startups emerging from curriculum development and other academic activities (Kaghan and Barnet, 1997).


207

IMPETUSES TO ACADEMIC FIRM FORMATION The appearance of commercializable results in the course of the academic research process, even before scientists formulate their research programmes with the intention of seeking such results or universities reorder their administrative processes in order to capture them, is the necessary cause for the emergence of entrepreneurial science. In addition to the opportunity presented by research results with commercial potential, entrepreneurial science has several sufficient causes, both proximate and long term, that encourage academic scientists to utilize these opportunities themselves rather than leaving them to others. Entrepreneurial impetuses in US academic science include the stringency in federal research funding, a culture of academic entrepreneurship originating in the seeking of government and foundation funds to support research, examples of colleagues’ successful firms, and government policies and programmes to translate academic research into industrial innovation (Etzkowitz and Stevens, 1995). Research Funding Difficulties Although federal investment in academic R&D increased during the 1990s, academic researchers strongly perceived a shortfall of resources during this period (National Science Board, 1996). The explanation of this paradox lies in the expansionary dynamic inherent in an academic research structure, based upon a PhD training system that produces research as a by-product. The expansionary dynamic is driven by the ever-increasing number of professors and their universities who wish them to engage in research. Formerly this pressure was largely impelled by the wish to conform to the prevailing academic prestige mode associated with basic research. In recent years, expansionary pressure has intensified due to increased attention to the economic outcomes of basic research that drew less research-intensive areas of the country into the competition to expand the research efforts of local universities as an economic development strategy. With the notable exception of a relatively brief wartime and early postwar era, characterized by rapidly expanding public resources for academic research, US universities have always lived with the exigencies of scarce resources. The increasing scale and costliness of research is also a factor. As traditional sources of research funding were unable to meet ever-expanding needs, academics sought alternative sources such as industrial sponsors. A faculty member discussed his involvement with industry: ‘In some areas we have found it necessary to go after that money. As the

208


experimental needs in computer science [increase], equipment needs build up. People realize that a small NSF grant just doesn’t hack it anymore.’ Nevertheless, there are cultural and other barriers to overcome before a smooth working relationship can be established. A professor described the dilemma: ‘It’s harder work with industry funding than federal funding, harder to go through the procurement process, to negotiate the terms of the contract.’ Industry’s expectations for secrecy, for example, are sometimes unreasonable. Dissatisfaction with working with existing companies is another reason professors offer for starting their own company. Creating an independent financial base to fund one’s own research is a significant motivator of entrepreneurial activity. Stringency in federal research funding has led academic scientists to broaden their search for research support from basic to applied government programmes and vice versa. A possible source that has grown in recent years has been research subventions from companies, including firms founded by academics themselves for this purpose, driving industrial support of academic research up from a low of 4 per cent to a still modest 7 per cent during the 1980s. Concomitant with a shift from military to commercial criteria, new sources of funding for academic research have opened up in some fields that have experienced a rise in practical significance, such as the biological sciences, making the notion of stringency specific to others, such as nuclear physics, which has experienced a decline (Blumenthal et al., 1986a). Nevertheless, federal research agencies are still the most important external interlocutors for academic researchers. A department chair at Cal Tech noted: The amount of money from industry is a pittance in the total budget, therefore everybody’s wasting their time to try to improve it . . . it’s still a drop in the bucket . . . we were running about 3 per cent total in our dept. We do value our industrial ties . . . have good friends, interact strongly with them in all kinds of respects and . . . the unrestricted money is invaluable, you wouldn’t want to lose a penny of it and would like to increase it a lot, but its impact vis-à-vis federal funding is almost non-existent.

In contrast to this view, based upon the current source of resources, others look towards realizing greater value from the commercial potential of research. Even without creating a firm themselves, academic scientists can earn funds to support their research by making commercializable results available for sale to existing firms. As a Stanford faculty member described the process: It’s also motivating for us to try to identify things that we do that may be licensable or patentable and to make OTL [Office of Technology and Licensing]


209

aware of that because according to University policy, 30 per cent of the money comes back to the scientist, 30 per cent comes back to the Department as well as 30 per cent for the University. So, almost all the computing equipment and money for my post docs have been funded by the work that we did. So there’s motivation.

Earlier in the twentieth century, experiencing difficulties in the French research funding system, the Curies considered exploring the commercial possibilities of their radium discovery for just this purpose (Quinn, 1995). Tapping capital markets through a public offering of stock is an additional source of research funding, especially in biotechnology-related fields, although one not yet recognized in Science Indicators volumes! In response to the increasingly time-consuming task of applying for federal research grants, a faculty member said that ‘another way to get a whole bunch of money . . . is to start a company’. After resisting the idea of starting a firm in favour of establishing an independent non-profit research institute, largely supported by corporate and governmental research funds, two academic scientists returned to an idea they had earlier rejected of seeking venture capital funding. As one of the founders explained their motivation for firm formation, ‘Post docs who are really good will want to have some place that at least guarantees their salary. And that we were not able to do. It was for that reason we decided to start the company.’ Other scientists realized that they could combine doing good science with making money by starting a company. As they enhanced their academic salaries through earnings from entrepreneurial ventures, and continued to publish at a high rate, they lost any previous aversion to the capitalization of knowledge (Blumenthal et al., 1986b). The Industrial Penumbra of the University The success of the strategy to create a penumbra of companies surrounding the university has given rise to an industrial pull upon faculty members. For example, a faculty member reported: The relationship with Collaborative is ongoing daily. We are always talking about what project we are going to do next. What the priority is, who is involved, there are probably six projects, a dozen staff members and maybe close to a dozen people scattered around three or four different departments on campus that are doing things with them.

Geographical proximity makes a difference in encouraging appropriate interaction. Such intensive interaction sheds new light on the question of industrial influence on faculty research direction and whether this is good,

210


bad or irrelevant. Thus the ‘issue of investigator initiation is much more complicated because I am bringing my investigator initiated technology to their company initiated product. It is a partnership in which each partner brings his own special thing. That is the only reason they are talking. Do your thing on our stuff.’ Previous conflicts based on an assumption of a dividing line between the academic and industrial sides of a relationship are superseded as divisions disappear. A more integrated model of academic–industry relations is emerging along with a diversified network of transfer institutions. Indeed, the very notion of technology transfer, or at least transfer at a distance, is superseded as universities develop their own industrial sector. Not surprisingly, a receptive academic environment is an incentive to entrepreneurship, while a negative one is a disincentive. MIT and Stanford University are the exemplars of firm formation as an academic mission. In the 1930s the president of MIT persuaded the leadership of the New England region to make the creation of companies from academic research the centrepiece of their regional economic development strategy (Etzkowitz, 2002). At Stanford, Frederick Terman, dean of engineering, provided some of the funds to help two of his former students, Hewlett and Packard, to form their firm just before World War II. A faculty member commented: ‘Because it has been encouraged here from its inception, it makes it easy to become involved [in firm formation]. [there are] . . . more opportunities . . . people come in expecting it.’ On the other hand, at a university noted for its opposition to entrepreneurial activity, an administrator noted that despite the disfavour in which it is held, ‘There have been some [firms founded] . . . it’s frowned upon. It takes a lot of time and the faculty are limited . . . in the amount of time they have.’ Under these conditions, procedures that could ameliorate conflicts are not instituted and faculty who feel constricted by the environment leave. A surrounding region filled with firms that have grown out of the university is also a significant impetus to future entrepreneurial activity. The existence of a previous generation of university-originated startups provides consulting opportunities, even for faculty at other area universities. A Stanford professor noted: ‘in the area there’s a lot of activity and that tends to promote the involvement of people’. From their contact with such companies, faculty become more knowledgeable about the firm formation process and thus more likely to become involved. Faculty who have started their own firms also become advisers to those newly embarking on a venture. An aspiring faculty entrepreneur recalled that a departmental colleague who had formed a firm ‘gave me a lot of advice . . . he was the role model’. The availability of such role models makes it more likely that other faculty members will form a firm out of their research results when the opportunity appears.


211

Once a university has established an entrepreneurial tradition, and a number of successful companies, fellow faculty members can offer material, in addition to moral, support to their colleagues who are trying to establish a company of their own. A previous stratum of universityoriginated firms and professors who have made money from founding their own firms creates a potential cadre of ‘angels’ that prospective academic firm founders can look to in raising funds to start their firms. Early faculty firm founders at MIT were known on campus for their willingness to supply capital to help younger colleagues. Normative Impetus to Firm Formation In an era when results are often embodied in software, sharing research results takes on a dimension of complexity well beyond reproducing and mailing a preprint or reprint of an article. Software must be debugged, maintained, enhanced, translated to different platforms to be useful. These activities require organizational and financial resources well beyond the capacity of an academic lab and its traditional research supporters, especially if the demand is great and the software complex. As one of the researchers described the dilemma of success, ‘We had an NSF [National Science Foundation] Grant that supported [our research] and many people wanted us to convert our programs to run on other machines. We couldn’t get support (on our grant) to do that and our programs were very popular. We were sending them out to every place that had machines available that could run them.’ The demand grew beyond the ability of the academic laboratory to meet it. Firm formation is also driven by the norms of academic collegiality, mandating sharing of research results. When the federal research support funding system was not able or willing to expand the capabilities of a laboratory to meet the demand for the software that its research support had helped create, the researchers reluctantly turned to the private sector. They decided that, ‘Since we couldn’t get support, we thought perhaps the commercial area was the best way to get the technology that we developed here at Stanford out into the commercial domain.’ This was a step taken only if they failed to receive support from NSF and NIH [National Institutes of Health] to distribute the software. ‘The demonstration at NIH was successful, but they didn’t have the funds to develop this resource.’ The researchers also tried and failed to find an existing company to develop and market the software. As one of the researchers described their efforts, ‘We initially looked for companies that might license it from us . . . none were really prompted to maintain or develop the software further.’ Failure to identify an existing firm to market a product is a traditional impetus

212


to inventors, who strongly believe in their innovation, to form a firm themselves to bring their product to market. Chemists involved with molecular modelling, previously a highly theoretical topic, have had to face the exigencies of software distribution as their research tools increasingly became embodied in software. Since the interest in the software is not only from academic labs but from companies that can afford to pay large sums, the possibility opens up of building a company around a programme or group of programmes and marketing them to industry at commercial rates while distributing them to academia at a nominal cost. Academic firm founders thus learn to balance academic and commercial values. In one instance, as members of the Board, the academics were able to influence the firm to find a way to make a research tool available to the academic community at modest cost. An academic described the initial reaction to the idea: ‘The rest of the board were venture capitalists, you can imagine how they felt! They required we make a profit.’ On the other hand, ‘It was only because we were very academically oriented and we said, “Look, it doesn’t matter if this company doesn’t grow very strongly at first. We want to grow slow and do it right and provide the facilities to academics”.’ The outcome was a compromise between the two sides, meeting academic and business objectives at the same time, through the support of government research agency to partially subsidize academic access to the firm’s product. There is some evidence that firms spin out of interdisciplinary research or, at least, that some such collaborations are a significant precursor to firm formation. As one academic firm founder described the origins of his firm, ‘If it had not been for the collaboration between the two departments [biochemistry, computer science], intimate, day-to-day working [together], it never would have happened. Intelligenetics and Intellicorp grew out of this type of collaboration. We had GSB [the Graduate School of Business], the Medical School and Computer Science all working together.’ In this model, the various schools of the university contributed to the ability of the university to spin out firms, providing specialized expertise well beyond the original intellectual property. Academic Entrepreneurial Culture An entrepreneurial culture within the university encourages faculty to look at their research results from a dual perspective: (1) a traditional research perspective in which publishable contributions to the literature are entered into the ‘cycle of credibility’ (Latour and Woolgar, 1979) and (2) an entrepreneurial perspective in which results are scanned for their commercial as well as their intellectual potential. A public research


213

university that we studied experienced a dramatic change from a single to a dual mode of research salience. A faculty member who lived through the change described the process: ‘When I first came here the thought of a professor trying to make money was anathema . . . really bad form. That changed when biotech happened.’ Several examples of firm formation encouraged by overtures from venture capitalists led other faculty, at least in disciplines with similar opportunities, to conclude: ‘Gosh, these biochemists get to do this company thing, that’s kind of neat, maybe it’s not so bad after all.’ Although some academics working in the humanities and science policy experts remain concerned that the research direction of academic science will be distorted (Krimsky et al., 1991), serious opposition dissipated as leading opponents of entrepreneurial ventures from academia, such as Nobel Laureate Joshua Lederberg of Rockefeller University, soon became involved with firms themselves, in his case, Cetus. A research group within an academic department and a startup firm outside are quite similar despite apparent differences represented by the ideology of basic research, on the one hand, and a corporate legal form on the other. As an academic firm founder summed up the comparison, ‘the way [the company] is running now it’s almost like being a professor because it’s all proposals and soft money’. There is an entrepreneurial dynamic built into the US research funding system, based on the premise that faculty have the primary responsibility for obtaining their own research funds. As a faculty member described the system, ‘It’s amazing how much being a professor is like running a small business. The system forces you to be very entrepreneurial because everything is driven by financing your group.’ At least until a startup markets its product or is able to attract funding from conventional financial sources, the focus of funding efforts is typically on a panoply of federal and state programmes that are themselves derived from the research funding model and its peer review procedures. As a faculty entrepreneur viewed the situation, ‘What is the difference between financing a research group on campus and financing a research group off campus? You have a lot more options off campus but if you go the federal government proposal route, it’s really very similar.’ The entrepreneurial nature of the US academic research system helps explain why faculty entrepreneurs typically feel it is not a great leap from an oncampus research group to an off-campus firm. A typical trajectory of firm formation is the transition from an individual consulting practice, conducted within the parameters of the one-fifth rule, to a more extensive involvement, leading to the development of tangible products. A faculty member described his transition from consulting to

214


firm formation: ‘It got to the point where I was making money consulting and needed some sort of corporate structure and liability insurance; so I started [the company] a couple of years ago. From me [alone, it has grown] to eight people. We’re still 70 per cent service oriented, but we do produce better growth media for bacteria and kits for detecting bacteria.’ The firm was built, in part, on the university’s reputation but was symbiotic in that its services to clients brought them into closer contact with on-campus research projects. In another instance, an attempt was made to reconcile the various conflicting interests in firm formation and make them complementary with each other by the university having some equity in the company and holding the initial intellectual property rights. Despite the integrated mode arrived at, some separation, worked out on technical grounds, was still necessary to avoid conflicts. There is no line. It’s just a complete continuum. It is true that I have a notebook that says [university name] and a notebook that says [firm name] and if I make an invention in the [company] notebook then the assignment and the exclusive license goes to [the firm] and if I make an invention in the university notebook then the government has rights to the invention because they are funding the work. [Interviewer: How do you decide which notebook you are going to write in?] We have ways of dividing it up by compound class. In the proposals that I write to the government I propose certain compound classes. There is no overlap between the compound classes that we work on campus and the compound classes that we work on off campus so there is a nice objective way of distinguishing that.

The technical mode of separation chosen, by compound classes, suggests that while boundaries have eroded as firm and university cooperate closely to mount a joint research effort, a clear division of labour persists. Once the university accepted firm formation and assistance to the local economy as an academic objective, the issue of boundary maintenance was seen in a new light. An informant noted: ‘When the university changed its attitude toward entrepreneurial ventures, one consequence was that the administration renegotiated its contract with the patent management firm that dealt with the school’s intellectual property.’ A new sentence said, ‘If the university chooses to start an entrepreneurial new venture based upon the invention then the university can keep the assignment and do whatever it wants. [Interviewer: Why did the university make that change?] Because the university decided that it wanted to encourage faculty to spin off these companies.’ Organizational and ideological boundaries between academia and industry were redrawn, with faculty encouraged to utilize leave procedures to take time to form a firm, and entrepreneurial ventures noted as contributing to research excellence in university promotional literature.


215

CONCLUSION: THE CAPITALIZATION OF KNOWLEDGE A co-evolution of academic scientists and universities may be identified. When the transition towards capitalization of knowledge is uneven, conflicts of interest arise; when the transition is in parallel, confluence of interests is the likely result. A more direct role in the economy has become an accepted academic function, and this is reflected in the way universities interact with industry. There has been a shift in emphasis from traditional modes of academic–industry relations oriented to supplying academic ‘inputs’ to existing firms either in the form of information flows or through licensing patent rights to technology in exchange for royalties. Utilizing academic knowledge to establish a new firm, usually located in the vicinity of the university, has become a more important objective. Indeed, the firm may initially be established on or near the campus in an incubator facility sponsored by the university to contribute to the local economy. Defining and maintaining an organization’s relationship to the external environment through ‘boundary work’ has different purposes, depending upon whether goals are static or undergoing change. A defensive posture is usually taken, and buttressed by arguments supporting institutional integrity, in affirming a traditional role (Gieryn, 1983). The reworking of boundaries around institutions undergoing changes in their mission occurs through a ‘game of legitimation’ that can take various forms. One strategy is to conflate new purposes with old ones to show that they are in accord. For example, universities legitimize entrepreneurial activities by aligning them with accepted functions such as research and service. In addition, new organizational roles are posited, in this instance contribution to economic and social development as a third academic mission in its own right. Similarly, faculty members find that their entrepreneurial activities provide vivid examples for their teaching practice as well as a source of research ideas. University–industry relations are increasingly led by opportunities discerned in academic research that is funded to achieve long-term utilitarian objectives as well as theoretical advance, recognizing that the two goals are compatible and, indeed, mutually reinforcing. In a recursion from science studies research to practice, a University of New Orleans professor requested a copy of a study of entrepreneurial activities at State University of New York at Stony Brook to encourage his colleagues in the marine research centre to found a firm. Regional economic development is superseding the sale of intellectual property rights to the highest bidder, even as the translation of commercializable research results into economic activity is becoming an accepted academic mission (Etzkowitz, 2008).

216


NOTE Data on university–industry relations in the USA are drawn from studies conducted by the author with the support of the US National Science Foundation. More than 100 in-depth interviews were conducted with faculty and administrators at universities, both public and private, with long-standing and newly emerging industrial ties.

REFERENCES AUTM (2005) www.autm.net/surveys/ accessed 10 July 2007. Blumenthal, D. et al. (1986a), ‘Industrial support of university research in biotechnology’, Science, 231, 242–46. Blumenthal, D. et al. (1986b), ‘University–industry research relations in biotechnology’, Science, 232, 1361–66. Etzkowitz, Henry (1983), ‘Entrepreneurial scientists and entrepreneurial universities in American academic science’, Minerva, 21 (Autumn), 198–233. Etzkowitz, Henry (2002), MIT and the Rise of Entrepreneurial Science, London: Routledge. Etzkowitz, Henry (2008), The Triple Helix: University–Industry–Government Innovation in Action, London: Routledge. Etzkowitz, H. and Ashley Stevens (1995), ‘Inching toward industrial policy: the university’s role in government initiatives to assist small, innovative companies in the U.S.’, Science Studies, 8 (2), 13–31. Faulkner, Wendy and Jacqueline Senker (1995), Knowledge Frontiers: Public Sector Research and Industrial Innovation in Biotechnology, Engineering Ceramics, and Parallel Computing, Oxford: Clarendon Press. Geiger, Roger (1986), To Advance Knowledge: The Growth of American Research Universities, 1900–1940, New York: Oxford University Press. Gieryn, T. (1983), ‘Boundary-work and the demarcation of science from nonscience: strains and interests in professional ideologies of scientists’, American Sociological Review, 48, 781–95. Kaghan, William and Gerald Barnet (1997), ‘The desktop model of innovation in digital media’, in Henry Etzkowitz and Loet Leydesdorff (eds), Universities and the Global Knowledge Economy: A Triple Helix of Academic–Industry– Government Relations, London: Cassell. Krimsky, Sheldon, James Ennis and Robert Weissman (1991), ‘Academic– corporate ties in biotechnology: a quantitative study’, Science, Technology and Human Values, 16 (3), 275–87. Latour, Bruno and Steve Woolgar (1979), Laboratory Life, Beverly Hills, CA: Sage. Matkin, Gary (1990), Technology Transfer and the University, New York: Macmillan. National Science Board (1996), Science and Engineering Indicators, Washington, DC: National Science Foundation. Powell W., Jason Owen Smith and Jeanette Colyvas (2007), ‘Innovation and emulation: lessons from American universities in selling private rights to public knowledge’, Minerva, June, 121–42. Quinn, Swan (1995), Marie Curie: A Life, New York: Simon and Schuster.


217

Rahm, Diane (1996), R&D Partnering and the Environment of U.S. Research Universities, Proceedings of the International Conference on Technology Management: University/Industry/Government Collaboration, Istanbul: Bogazici University. Shimshoni, D. (1970), ‘The mobile scientist in the American instrument industry’, Minerva, 8 (1), 59–89.

8.

Multi-level perspectives: a comparative analysis of national R&D policies Caroline Lanciano-Morandat and Eric Verdier

In the era of globalization, the quality of science–industry relations is presented as a key source of industrial innovation and, more generally, of economic competitiveness. The supranational orientations of the European Union and the recommendations of the OECD converge to promote a ‘knowledge society’. These political references recycle the results of research in economics and in the sociology of innovation, and together inspire European states’ public R&D and innovation (RDI) policies. Yet these convergences between political action, international expertise and the social sciences do not mean that there is necessarily a ‘one best way’ of articulating science and industry. Shifts in public intervention have to compromise with institutions inherited from the past,1 and with the practices of firms and other private sector actors. This chapter develops an approach that integrates these different levels of analysis. The aim is to explain the specific national characteristics of policy-making in a context in which the references of that action tend to be standardized. In this respect, there is some convergence with the analysis in terms of ‘Varieties of Capitalism’ (Hall and Soskice, 2001), although our approach differs as regards the national coherence highlighted by the latter in a deterministic perspective: ‘In any national economy, firms will gravitate towards the mode of coordination for which there is institutional support’ (Hall and Soskice, 2001, p. 9). Our approach consists in explicitly taking into account the arrangements, negotiations and institutional bricolages that the actors use to coordinate their actions in the form of conventions. It highlights the devices that support the science–industry relations between different public and private sector actors, from firms to political authorities at different levels (regional, national and European). It therefore examines the rules, contractual devices and organizational forms equipping interactions between actors, but also the different political and ethical principles 218

A comparative analysis of national R&D policies

219

orienting their choices. ‘The state, like other institutions, is essentially a convention between persons, but unlike other institutions, in western democracies all state conventions are based on representations of the “common good” for their societies’ (Storper, 1998, p. 10). Four types of conventions of policy-making concerning RDI and innovation can thus be identified. It emerges that national specificities do not stem from a structural distinction between what would ‘ontologically’ be German science–industry relations and French or British science–industry relations. An action regime peculiar to each country, and that we therefore qualify as ‘societal’, is in fact the outcome of a compromise between these four legitimate patterns concerning RDI policy. But this compromise is only temporary. It evolves under the impulse of dynamics that may be endogenous – for instance the emergence of new technological districts – or exogenous – ‘best practices’ imposed at international level as new standards. Consequently, the legitimacy of different conceptions of the common good varies and the work of the actors (firms, government agencies, universities etc.) generates new rules and therefore new compromises. This approach is briefly illustrated in a comparative analysis of the trajectories of the British, French and German RDI policies based on the classical ‘societal approach’ (Maurice et al., 1986) that we combine with a ‘conventionalist’ analysis in order to deal with issues of coordination (Eymard-Duvernay, 2002).

AN ANALYTICAL FRAMEWORK: THE CONSTRUCTION OF POLICY-MAKING CONVENTIONS These conventions as representations of the common good are the basis for the legitimacy of rules. They encompass different conceptions of what is ‘efficient and fair’ in collective action concerning RDI. Each of them corresponds to a specific mode of justification, in other words to a particular research ethic (see ‘The orders of worth’ – Economies de la grandeur – Boltanski and Thévenot, 1991 and Lamont and Thévenot, 2000): scientific progress; state services and the national interest; the market, i.e., the creation of shareholder value; and, lastly, the project embodying technological creativity. These referents are based on conceptualizations from the economics and the sociology of innovation. They are mobilized by national and international experts, who promote them in the national2 and supranational arenas (in the OECD and the EU). By explicitly referring to the state’s position in collective action, we

220


distinguish four patterns of conventions: the ‘Republic of Science’, the ‘state as an entrepreneur’, the ‘state as a regulator’ and, finally, the ‘state as a facilitator’ (of technological projects). These conventions cannot be dissociated from the following dimensions which constitute the RDI policy-making regime in which they are set: 1.

2.

3.

4.

5. 6. 7. 8.

The positioning of the public authorities in a multi-level perspective (identify which level predominates: local, national, supranational or European?), considering that many analysts today perceive a distinct weakening of the national level (Larédo and Mustar, 2001); consequently, the diversity of practices within the same country is increased, at least potentially. The boundary between the public and private sectors; for instance, the university could be considered either essentially as a public actor or as a private entrepreneur (Etzkowitz and Leydesdorff, 2000). The predominant organizational frame that connects different actors: from complete independence in the pure academic form to interdependence in the network form. The mediators in charge of networking the different worlds, from occasional contact to integration when the state is the entrepreneur of science–industry relations. The definition of the competencies that shapes the legitimacy of the actors, the criteria of success and the goals to achieve. The rules that frame, drive and evaluate researchers’ work, in a perspective of innovation. The modes of financing, from public funds (for basic research) to a web of public and private financing in the case of a network. The rules governing the circulation and employment of people (the type of labour market).

These conventions claim to account for the ideal-types of research ethic underlying collective policy-making, rather than directly for the structural coherences of a particular country or region. There are common points with the approach developed by Storper and Salais (1997) in terms of conventions of the ‘absent state’ (particularly marked in the USA) and the ‘external’ or ‘overarching state’ (strong in France) and the ‘situated state’ or ‘subsidiary state’ (marked in Germany). Historical societal constructions stem from the arrangements – that vary in time and space – between these different conceptions of policymaking. It is therefore necessary to highlight the tensions, conflicts and compromises that trigger changes in public systems and are spurred by attempts at reform. Their degree of success is the result of interactions


221

between inherited historical constructions and the projects of (new) actors involved in the definition of the common good.

FOUR PATTERNS OF RDI POLICY-MAKING CONVENTIONS The characteristics of the four conventions of policy-making in science and innovation can be synthesized in terms of their insertion in a general regime of action (see Table 8.1). The Republic of Science The ‘Republic of Science’ is based on a convention similar to the model of Merton, the founder of the sociology of science. It highlights the positive role of science in society and aims for ‘the development of codified knowledge’ (Merton, 1973). It implies a strict separation between scientific institutions and those governing the rest of society. In this model, public intervention can acquire legitimacy only by adhering to guidelines and priorities defined independently by scientists whose reputation sets the standards for competencies. This conception of the ‘academic state’ limits public intervention to the financing of the pure public good that scientific knowledge is supposed to be. These characteristics imply the complete application of a ‘disclosure norm’ for scientific progress (the ‘open science’ model), after peer validation. Government has to ensure that ‘generic’ resources are made available to society. It is up to firms to ‘endogenize’ them, i.e., to appropriate them efficiently, in a specific way. The other side to the ‘Republic of Science’ is therefore the ‘Kingdom of Technology’ (Polanyi, 1962), founded on a private appropriation by each agent of this general, abstract knowledge, for the purpose of generating comparative advantages from the efficient application of new knowledge. This radical distinction between pure research and the pursuit of industrial and economic objectives causes relations between universities and industry to depend on personalities in academia who act as advisers on the efficient application of knowledge. These relations remain occasional and informal and even tend to be hidden, for the purpose of maintaining science’s original purity. The ‘priority norm’, which rewards only the first discovery and thus grants the creator a sort of ‘moral ownership’ of a product, is an incentive to produce original knowledge. More generally, the competencies produced under the aegis of this ‘Republic’ are above all academic. The evaluation of scientists, at the time of recruitment and throughout their

222

Table 8.1


The characteristics of conventions of policy-making concerning R&D and innovation

Relevant dimensions

I Republic of Science

II The state as an entrepreneur

Overriding principle: research ethic Level of state regulation

The progress of State service science and national interest Discipline National (local faculty)

III The state as a regulator

IV The state as a facilitator

Market: shareholder value Regional integration (‘Europe’) Codetermination of the entrepreneurial university and firms Contract (negotiation between individuals or organizations)

Project: technological creativity Multi-level

Governance Independence of the of academic public–private communities relationship

Control by central state: ministry or agency

Organizational architecture

Academia (faculties)

Category of mediating actors

Renowned scientific personalities

Key competencies

Disciplinary knowledge

Large programme (hierarchical management and organization) Managerial and Individual political elites mobility of scientists between the private and public spheres Meritocratic Operational excellence versatility of individuals

Incentive institution

Peer evaluation Power over (disclosure and scientific and priority norms) industrial development Public grants Public and individual subsidies and fees government orders Occupational Public and labour markets private internal markets

Funding institution

Labour institution

Property rights, patents and profit-sharing Joint contribution of higher education and firms External labour markets

Delegation of responsibility for technicoscientific coordination Network (interaction and alignment within the network) Diversity of actors as intermediaries between university and firms Interdisciplinary and ability to cooperate Salary increases and stock options Multiplicity of sources and levels of financing Labour markets peculiar to networks


223

careers, is based on judgements of the scientific community. The training of engineers is embodied in academic disciplines that are the foundations of occupational labour markets. Peer evaluation is important throughout an individual’s career, by way of a labour market in which professional societies guarantee the reliability and standardization of skills. The ‘State as an Entrepreneur’ This convention underlies a ‘mission-oriented’ public policy (Ergas, 1992) corresponding to ‘radically innovative projects which are necessary for the pursuit of national interests’. The mission concerns technological domains of strategic importance to the state. Its main features are the centralization of decision-making, the definition of objectives in government programmes, the concentration of the number of firms involved, and the creation of a specific government agency with a high level of discretionary power, responsible for operational coordination, under the supervision of a national or federal administration. The science/innovation relationship is then explicitly built (unlike the preceding convention) in a framework of planning, on the basis of a model often referred to as ‘Colbertist’ (Barré and Papon, 1998). This schema organizes a science/innovation twosome guided by a ‘higher’ socioeconomic order since technological policy is legitimized by its contribution to a national interest that, in this case, is confused with the state service. This convention appeared and was conceptualized in the period after World War II and was promoted for two main reasons: 1. 2.

To accelerate the country’s productive and technological modernization in order to catch up with competitors; To guarantee the availability of technologies essential to the quest for national independence.

The literature highlights the fact that it is a ‘top-down’ innovation model, ‘suited to complex technological objects used for large public infrastructures’ (Barré and Papon, 1998, p. 227). This convention has proved particularly effective for producing high-tech objects in public sector markets (aeronautics, space, military, nuclear, telecommunications etc.). Its organization is based on the model of the ‘large technological programme’ that involves a public agency, a research institution and a large industrial group (or several privileged operators) supported by a set of subcontractors. The objectives of the programme, the actors who have to participate in it, the operations and their scheduling are strictly defined ex ante. As part of a state-led and modernizing approach, this ‘industrial’ and

224


managerial conception is based to a large extent on coordination by wellidentified professions or academic elites (e.g. graduates of leading research universities or French ‘Grandes écoles’) and by applied research laboratories administered directly to help to implement government policies. It resembles an ‘external state’ convention, in the sense of Storper and Salais (1997): ‘The state has devised a methodology to evaluate differences with the common good (that it has defined indisputably, ex ante) and intervenes beforehand to correct those differences as far as possible. Everyone relies on it, conventionally’ (Salais, 1998, p. 62). Meritocratic excellence is based on selection for admission to the best schools and universities, which regulates access to the typically French ‘Grands corps de l’Etat’3. These combine the technical skills and organizational capacities that lie at the interface between government administration and large firms, with a view to running the large technological programmes. The large programmes are almost exclusively financed and managed in the framework of state contracts and public markets, with the aim of producing technological progress as a source of competitive advantage for the country’s industry. The development of competencies is produced by internal markets in the public and private sectors. The ‘State as a Regulator’ The state as a regulator promotes the transfer of scientific results to the private sector. It also ensures that the objectives of basic research are inspired or structured by the expectations of the ‘market’ and the corporate world. Whereas the preceding convention (i.e. the state as an entrepreneur) was limited to a national scale, here there is an openness onto supranational horizons due to the increasing weight of multinationals in technological dynamics. The quality of public research and its partnerships with the private sector are becoming key arguments to attest to the attractiveness of the country on a national and local level. The role of this convention is therefore to guarantee an efficient balance between the use of public research resources and market dynamics. This balance implies that the governance of the public–private relationship is co-determined by partnerships between firms and entrepreneurial universities, via contracts negotiated between the partners. As a regulator, the state has to guarantee the balance of commitments, even if that means promoting the establishment of private R&D resources within the province of its political authority, through targeted aid. This predominantly market orientation is attested by the importance granted to the definition of the ‘property rights’ that frame and stimulate two types of initiative emblematic of this convention: the creation of high-


225

tech startups by academic scientists, and the development of contractual relations between universities and firms. The first type of initiative requires the availability of adequate funding, via access to venture capital and the support of incubators for the first steps of innovative startups, which public agencies have to promote. The second helps to compensate for the financial difficulties experienced by both types of actor: government budgetary restrictions for universities, and the necessity for firms, due to heightened competition, to outsource a part of their R&D. This construction of the common good is justified theoretically in the ‘Mode 2’ of knowledge production (Gibbons, 1994). This ‘new’ Mode 2, focused on the problems to solve, as defined by industry, differs from ‘Mode 1’ in which the problems are posed and solved in a context governed by the interests of an independent scientific community with strictly academic and disciplinary aims. Mode 2 is based on a repeated reconfiguration of human resources, in flexible forms of organization of R&D, in order to be able to adjust to market trends – an ability based on the creation of knowledge of a transdisciplinary nature (Lam, 2001)4. Thus the formalization of Gibbons’s Mode relates more to a change of the ideological norm and ‘beliefs’ than to the empirical validity of a change of scientific practices. It thus aims politically to legitimize a narrowing of the gap between the academic world and enterprise, and is clearly open to a merchandization of science (Shinn, 2002). Efficient competencies stem from a co-production of firms and public research laboratories. This co-production is supported institutionally by a contract that organizes the collaboration between public and private sector researchers. The generally short-term mobility of scientists between the two worlds and the co-production of PhDs help to build up the trust needed to meet contractual objectives. A hybrid labour market is thus created between the higher education system and the industrial system, based on a joint construction of generic knowledge that can be used both commercially and industrially. Strong tension exists between this form of policy-making and the preceding one, based on a hierarchy organized around public programmes that structure private practices. It promotes the withdrawal of the state that refuses to define the common good a priori. The ‘State as a Facilitator’ (of Technological Projects) In the past ten years the literature on the economics of science and innovation has emphasized the importance of interactions between the different partners in scientific and technical production: government higher education or research institutions, firms with their own R&D capacities, and

226


organizations involved in funding and intermediation between these different ‘worlds’. This articulation has been systematized and popularized by a current of thinking involving scientists, managers and public authorities, known as the ‘triple helix’ (Etzkowitz and Leydesdorff, 1997, 2000). In this model the ‘co-production’ of knowledge is situated at the intersection of three interacting institutional spheres: university and research organizations; industry; and public authorities, especially through their specialized agencies. This articulation is said to generate trilateral networks through the overlapping of the different institutional spheres and the emergence of hybrid organizations at the interfaces. The objective is to create an innovative environment consisting of firms that are university or research organization spin-offs, tripartite initiatives for knowledge-based economic development, strategic alliances between firms of different sizes and technological levels, public laboratories, and university research teams. By promoting the establishment of R&D organizations that transcend traditional institutional boundaries (public/private, academic/applied etc.), and the creation of scientific and industrial parks at local level (Porter, 1998), these public interventions seem to correspond to a logic of organized accumulation of knowledge and the creation of innovative capacities at the micro-, meso- and macroeconomic levels. The dynamics of this model implies organizational transformations in each of the three spheres, as well as the intensification of their interrelations. This conception of the common good calls for the creation of cooperative research networks that group together the institutionally diverse partners (Callon, 1991). In terms of public intervention, ‘the convention is . . . that of a situated State that is expected to promote initiatives and their subsequent deployment, but not to dictate to them’ (Salais, 1998, p. 78). The collective construction of the common good can be concretized in two ways, depending on the degree of state involvement. The first relates to ‘more or less spontaneous creations, gradually resulting from local interactions. They do not correspond to clearly defined identities and rarely have clearly identified boundaries’ (Vinck, 1999, p. 389). The second results from state initiatives that, in the name of the proclaimed efficiency of cooperative scientific networks, are designed to catch up with the level of rival technological clusters. Without being exclusive, the local (or regional) dimension is strongly present in the form of the science or technology district (Saxenian, 1996). In his modelling of Silicon Valley, Aoki (2001) describes how this construction of the common good is based on ‘local institutional arrangements’ between independent entrepreneurs and venture


227

capitalists, outside state regulation. The information generated by invention work is ephemeral and rapidly depreciated due to the speed with which technology evolves. This tends to reduce the protective role of contracts and industrial property, so crucial in the context of the ‘state as a regulator’ convention. Aoki (2001) notes that intellectual property clauses that limit the mobility of experts between rival firms have become inapplicable in California. Specific institutions are therefore being created for technological networks. The same applies in the more classical configuration of the professional network (see Meyer-Krahmer and Schmoch, 1998), based on branch-specific techno-scientific societies that circulate knowledge between the various types of firm, mainly by promoting cooperative relations between large firms and subcontractors. Here again, the specific dimension of collective action is based not on the withdrawal of the state but on a delegation of public responsibility to private actors. While technology districts are turned more towards radical innovation, professional networks generate mainly incremental innovations (Hall and Soskice, 2001). Despite its lack of precision, the concept of a network is relevant ‘as a passage between the micro-economic behaviours of firms and the meso- or macro-economic levels’ (Amable et al. 1997, p. 94). Moreover, this form of collective action underscores the limits of the codification of knowledge and the importance of the ‘absorptive capacities’ of each protagonist to develop effective cooperation in technological research (Lundvall and Johnson, 1994). The literature logically questions the separation between basic and applied research, as well as the existence of a causal link, in the linear schema, between scientific discoveries, industrial R&D, and market applications. Finally, criteria of separation between public and private sector research have been criticized by several authors. Callon (1992) shows that state investments in scientific production are justified not by an intrinsic characteristic of science as a common good, but by the maintenance of a degree of diversity and flexibility in science, so that a wider range of research options is left open. These analytical constructions and their translations in terms of public policy illustrate the development of a network organization based on the initiatives of public and private actors around common projects (Boltanski and Chiappello, 1999). It is symptomatic that the individual competencies of ‘new research workers’ are formulated in terms of abilities to cooperate, to work in networks and to combine different types of knowledge (Lam, 2001). These competencies underlie the constitution of specific labour markets, peculiar to the network or industrial field concerned (LancianoMorandat and Nohara, 2002).

228


MAIN TRENDS IN PUBLIC REGIMES OF ACTION IN THREE EUROPEAN COUNTRIES Relations maintained between firms and universities, with a view to producing innovations, are supported by configurations of actors and rules of the game that draw on different patterns of conventions. The resulting compromise peculiar to each country outlines a regime of action that is a particular societal construction. However, these national regimes are increasingly subject to the critical evaluation of international benchmarks. For instance, the OECD regularly formulates normative recommendations inspired by ‘good practice’ (OECD, 1998), many of which originate in the USA (OECD, 2000). The conventions of the state as a regulator and the state as a project facilitator are strongly favoured by these recommendations, which not only inspire reforms but also contribute towards the setting of policy standards in different countries. The UK: Risks of Short-sightedness and Merchandization Traditionally, the British RDI regime has been characterized by a dual position: 1.

A very strong influence of the Republic of Science, especially in the medical research and biological fields (see Table 8A.4 in the Appendix); that is why many large US firms, like Pfizer (pharmaceuticals industry) or Hewlett-Packard (IT industry),5 set up some leading centres devoted to basic research as early as the 1960s, in order to exploit that scientific potential in a ‘Technology Kingdom’ perspective (see, in Table 8A.2 of the Appendix, the share of foreign funding in the national R&D expenditure); 2. A strong engagement of the ‘state as an entrepreneur’ in the defence industry and as a decisive factor in technological independence as regards computer technology in the 1970s. In the early 1980s this characteristic was translated into the weight of public expenditures on R&D. Although these were proportionally lower than in France, they were substantially higher than in Germany (see Table 8A.3 in the Appendix). In 1990 the share of military expenditures still accounted for nearly half of public R&D budgets (49 per cent), against 15 per cent in Germany. This explains why firms such as Racal, specialized in defence electronics, were able to constitute a technological competency base of a very high quality. Without too much difficulty they were then able to redeploy towards civil markets when defence research budgets were cut in the latter half of the 1990s.


229

Like others, Racal developed its portfolio of contracts with academic research, thus contributing to overall change towards science–industry relations based on the pattern of the ‘state as regulator’. Through successive reforms6 the UK governments have tried to redirect the policy-making towards: 1.

2.

The commercialization of scientific results, spurred, throughout the 1980s and 1990s, by the withdrawal of the state, which resulted in drastic cuts in funding, the cancellation of large national programmes and the privatization of public laboratories devoted to applied research. The national R&D effort declined from the mid-1980s (see Table 8A.1 in the Appendix), under the effect of slashed government funding (from 48.1 per cent of overall funding in 1981 to 35 per cent in 1991), while the share of foreign funding strongly increased to reach an unequalled level in Europe (see Table 8A.2 in Appendix). The share of R&D financed by firms and not-for-profit institutions grew (ten points during the 1980s for the former), primarily owing to contracts with academic laboratories, although these were not enough to offset the state’s withdrawal. With a view to promoting relations between scientific research and industry, the LINK programme was launched in 1986. It was designed to develop generic research within universities in order to meet the needs of enterprises more fully than did purely academic research. In parallel, many firms, especially in the information and communication sector such as ICL, Nortel and Signal, which wanted to control their internal R&D expenditures, developed their contractual relations with universities. This trend towards a more market-oriented conception of science–industry relations also involved financial support to academic entrepreneurs, from three programmes launched in the late 1990s: University Challenge, Science Enterprise Challenge, and Higher Education Innovation Fund. This was how the Imperial College was able to generate 14 academic spin-offs in the healthcare and biotechnology areas in 2001. The emergence of techno-scientific networks in the form of university–industry consortia and technology districts (clusters) from 1997 (Georghiou, 2001). Various public programmes accompanied and stimulated this trend. This was how the ‘virtual centres of excellence’ (VCE) were formed, comprising universities and firms, especially multinationals, in areas identified as priorities by the Foresight Communications Panel (for example the ‘Mobile VCE’). Science parks, less directly supported by government intervention and closer to ‘classic’ clusters, are business and technology transfer initiatives that encourage the startup and knowledge-base business, provide an

230


environment where international businesses can develop specific and close interactions with a particular centre of knowledge creation, and have formal and operational links with higher education institutes (see Lam, 2002). We can also cite the clusters developed in the ‘Oxbridge’ framework. These networks provide incentives for firms and universities to alter the organization of their internal R&D activities in order to be able to collaborate more effectively. This movement towards the ‘state as a project facilitator’ is based on undeniable resources. The excellence of the leading scientific universities (‘Oxbridge’) unquestionably places the UK closest to the ‘standard’ recommended by the OECD (1998). The fact remains that it is increasingly difficult to reach compromises between the various institutional settings, for several reasons: a growing risk of underinvestment in R&D by both the public and private sectors, due to the predominance of such shortterm commercial objectives that traditional fields of excellence and the capacity to produce generic know-how are being eroded (see, in Table 8A.4 in the Appendix, the decrease in the UK world share of scientific publications); increasing extraversion of the UK scientific and technical system, the potential of which is locked more and more into the strategies of large multinationals (see the continual increase in foreign investments in R&D); and excessive concentration of public and private resources on a few universities (Oxbridge and London), to the detriment of the creation of a sufficiently large capacity in RDI according to the ‘knowledge society’; in 1997, seven universities received 33 per cent of R&D funds from industry. These lower academic performances and the downward trend of patents of UK origin (see Table 8A.3 in the Appendix), prompted the New Labour government to increase public investments in R&D from the late 1990s, especially with the Joint Infrastructure Fund, thus departing from 20 years of negligence motivated, in a sort of paradox, by the very high ‘productivity’ of UK science (Georghiou, 2001). Germany: The Strength of Professional Networks and the Problems of University Entrepreneurship Traditionally, the assets of German industry stem from the proximity between higher education, especially the Fachhochschulen (technical universities), and industries, via research companies situated at the interface between these two worlds. This is clearly reflected in two main indicators: first, the share of R&D financed by industry, which is structurally much higher than in the UK and France (see Table 8A.2 in the Appendix), give


231

evidence of a strong comparative advantage; and second, the performance of German firms as regards patents, especially in the chemicals, mechanics and transport sectors (see Table 8A.6 in the Appendix). Although large firms account for 90 per cent of this effort, the 10 per cent share of SMEs in this industrial research is greater than that of the UK and France, owing to the science–industry interfaces. The efficiency of these professional networks has been proven in all capital-good industries, through the regular production of incremental innovations, which explain the high quality of products and their ability to meet customers’ needs. These networks are organized around various applied research institutes that have a private status but are financed by the federal government: for instance, the Fraunhofer Society or the Gottfreid-Wilhelm-Leibniz Association of Research Institutes (MeyerKrahmer, 2001). Many firms in the pharmaceutical sector (HMR-Aventis, Merck KgeA) and the computing–telecommunications sector (Siemens) are traditionally part of these networks to construct the science–industry common good and circulate it in industry. But both state authorities and the business community no longer believe that this situation is enough to maintain the competitive positions of the German economy. Alone it no longer meets some of the main scientific, technical and industrial challenges facing the German RDI regime, especially in high-tech sectors (telecommunications, biotechnology), in which technological performances are not as good (see Table 8A.6 in the Appendix). The first problem concerns the possibility of commercializing academic scientific work in the ‘high-tech’ field (information, communication, biotechnologies etc.). Following the US example, the Federal Ministry of Education and Research has therefore tried to structure collective action around a ‘state as a regulator’ convention, by making property rights and statuses more favourable to researchers and thus creating incentives. Many large firms have used this policy to externalize a part of their high-tech research by setting up spin-offs on university campuses (e.g. Merck KgeA around Munich). In this way they have helped to promote new clusters. The results of new incentives seem to be most tangible in the biotechnology field since the number of startups and the use of venture capital have increased substantially – so much so that Germany is the leader in this respect (Ernst & Young, 2000), with an increase in the number of small research-oriented biotech firms, from 75 in 1995 to 279 at the end of 1999. Although many of these startups are fragile, real success has been witnessed in the area of technology platforms where the traditional qualities of the prevailing regime of action in Germany are at play,

232


i.e., incremental innovations that combine existing technologies (Casper et al., 1999). Although limited, the same type of dynamic has developed in the ICT sector (linked to firms such as Agilent Technology, Lucent Technologies, Alcatel Research Centre). In the software field a spin-off dynamic from the Fraunhofer Gesellschaft has developed, in line with the policy initiated by the Federal Ministry of Education and Research (BMBF). To support the networks involving SMEs in applied research, specific measures have been taken in the new Länder (Meyer-Krahmer, 2001). France: The Tricky Exit from the ‘State as an Entrepreneur’ Convention The French higher education and research system is confronted with a profound challenge to the ‘state as an entrepreneur’ convention that has prevailed until now (Larédo and Mustar, 2001) – as evidenced in the structural importance of public R&D funding (see Table 8A.2 in the Appendix), especially funds allocated to large organizations directly under the authority of the state: CNRS (National Centre for Scientific Research), Inserm (National Institute for Medical Research) and so on. Policy-making is traditionally structured around big programmes, which have proved to be highly effective in producing complex technological objects used for large public infrastructure (telecommunications, nuclear energy, air and rail transport, etc.). The results have been far less convincing in the private sector, in computer technology – as the difficult industrial trajectory of Bull (previously the ‘national champion’) attests – and in the pharmaceutical industry. The ‘top-down’ model that prevailed until quite recently in France in the biotech field (see, e.g. the publicly managed Bio-Avenir programme and its relations with Rhône Poulenc-Rorer in the SESI project study) does not generally lend itself well to the spin-off forms of innovation which abound in biotechnology and ICT (Foray, 2000). This ‘state as an entrepreneur’ institutional setting has also been altered internally, which has strongly undermined its efficiency. For instance, the existence of several institutional channels for allocating aid and support for technology transfer to industry has resulted at best in a juxtaposition but more often in sterile competition between public institutions. The system as a whole is consequently largely illegible, especially for SMEs. Since 1982 incentives for researchers to transfer their results have proved inadequate. This applies both to high-tech startups and to the reduction of the cultural gap – probably more marked in France than elsewhere – between science and industry. It explains the severe diagnosis of the French scientific and technological scene (Guillaume, 1998) – ‘Honourable scientific research,


233

weak technology’ (Barré and Papon, 1998) – which inspired the 12 July 1999 blueprint law on research. With its related measures concerning innovation, this law was designed to ease statutory constraints, to develop incubators and to facilitate access to venture capital, in order to promote the development of high-tech companies based on public research results. In this perspective, INRIA (National Institute of Computer Science) and its 436 spin-offs are presented as a model to follow7. It is significant that at the end of 2003 the director of INRIA was appointed as General Director of CNRS, an institution focusing first and foremost on basic research, with a staff of 25 000. But the new law also aims to move towards a ‘state as a facilitator’ institutional setting. The idea is no longer to set up large programmes but to encourage the creation of precise industry–research cooperative projects. This goal is reflected in a semantic and institutional change: it refers to ‘network’ rather than ‘programme’, in a logic similar to that of ‘consortia’ advocated by the Guillaume Report, which inspired the law to a large extent. Two areas in which France lags behind other leading industrialized countries are priorities: the life sciences, and information and communication technologies (see Table 8A.5 in the Appendix). In the first field case the aim is to promote activities in the field of genomics – by supporting genopoles and startups developing computer technology applications for use in the biotech field – and health-related technology – by setting up a national health technology research and innovation network (the RNTS). The example of genomics is relevant since it shows how narrow this French strategy for integrating models and institutions developed in other societal contexts is in its conception. It dates back to before the 1999 law, since the Evry genopole was launched in February 1998. This experiment was intended to make up for the fact that France had fallen behind, by encouraging ‘interpenetration’ between technological and scientific advances. It was to promote collaboration between public and private laboratories and firms, while attempting to avoid the pitfalls associated with orienting academic research too strongly towards short-term objectives (Branciard, 2001). The problem of ‘catching up’ with the most competitive countries has been an incentive to develop state-led policy-making in which coordination by the hierarchy takes precedence over cooperation. Although the latter is indispensable for producing collective learning, the time taken to establish it is not necessarily compatible with the need to catch up with competitors. Reaching a compromise between the different institutional settings is thus difficult. In the meantime the French RDI regime is witnessing its structural position, strongly supported by the ‘state as an entrepreneur’, gradually being undermined (Branciard and Verdier, 2003).

234


Recently the French government took steps to extend the ‘good practice’ of the Grenoble technological district. With the support of national and regional public agencies and bodies, this innovative district is now becoming a key player in the nanotechnologies field with the creation of a new cluster ‘Minatec’. Under the initiative of researchers of the public Commissariat général à l’énergie atomique, the cluster has attracted MNCs as its main private stakeholders, including firms from Europe (e.g. ST Microelectronics; Philips) and the USA (e.g. Motorola, Atmel). Moreover, based on the main conclusions of the ‘Beffa report’ (Beffa, 2005)8, state and regional governments decided to support 60 projects for competitiveness clusters, after a selective process. This new generation of public programmes expresses the search for compromises between ‘the state as an entrepreneur’ and ‘the state as a facilitator’. It may also meet another challenge facing the French research system, the ‘underspecialization’ of public funding devoted to basic research (LancianoMorandat and Nohara, 2005).

CONCLUSION State reforms are made and implemented at national level but are based on the recommendations of supranational authorities (the OECD and the European Commission), which are themselves influenced by the ideas produced by the sociology and the economics of innovation (Lundvall and Borras, 1997). The resulting compromises and arrangements define rapidly changing national regimes that are symptomatic of specific compromises with the international and scientific references mentioned above. The increasing weight of the ‘state as a regulator’ and the ‘state as a facilitator’ institutional settings, at the expense of the ‘state as an entrepreneur’ and, to a lesser degree, the ‘Republic of Science’, is generating increasing diversification of collective action. This action is less dependent on national institutional frames than previously. It is more and more the result of the initiatives of cooperative networks or the local configurations in which multinational firms, among others, develop practices that could not be explained only in terms of a ‘global’ strategy. If we want to continue referring to a ‘national system’, we will have to conceive of it more and more as the outcome of a ‘set’ of networks and configurations whose coherence stems only partially from the direct influence of national institutions. The approach in terms of conventions of policy-making enables us to define regimes of action for each country. These regimes are compromises between different patterns and are continually moving. They


235

are constantly subject to critique, to reinterpretations in changing contexts, and to failures in coordination that motivate attempts at reform – themselves interpreted and adjusted by individual and collective, public and private actors. As far as RDI is concerned, all countries implement measures aimed at promoting a better diffusion and commercialization of the results of public research, in order to stimulate private innovation. This ‘market’ reference (Jobert, 1994) has undeniably impacted strongly on the course of policy-making. The shared risk is of favouring ‘short-term’ behaviours by research institutions and firms, to the detriment of the accumulation of generic knowledge on which the Republic of Science and the technological creativity targeted by innovation network projects are based. Powell and Owen-Smith (1998) have drawn attention to the fact that the predominance of market criteria for assessing the ‘merits’ of academic research has helped to corrode their mission and consequently has dangerously undermined the public’s trust in these institutions and in science. Reaching a compromise between the different research ethics is thus asserted as a condition for sustainable development of the legitimacy and efficiency of policy-making concerning R&D.

NOTES 1. In one sentence, ‘state reform is about building new precedents that would lead to new conventions; to do this, they need to involve the actors, which requires talk among the actors so that they might ultimately build confidence in new patterns of mutual interaction, which is a prerequisite of new sets of mutual expectations which are, in effect, convention’ (Storper, 1998, p. 13). 2. See Branciard and Verdier (2003) on the French case and the influence of the OECD’s expertise. 3. The role of the senior levels of the French civil service, which are staffed by graduates of the elite engineering schools and the civil-service college (ENA), in conducting a state-led economic policy and controlling France’s largest firms (nationalized in 1945 and 1981), has often been emphasized (Suleiman, 1995). 4. This thesis of one regime of production of science supplanting another is criticized by Pestre (1997), who sees the two modes as having functioned in parallel for the past few centuries in the West. 5. Firms studied during our European research project: see a presentation of the methodology in the Appendix. 6. With the advent of New Labour, official reports on these issues proliferated: see competitive White Paper (DTI, 1999a), special reports on biotechnology clusters (1999b) and Genome Valley (1999c), White Paper on enterprise, skills and innovation (DTI/DfEE, 2001). 7. It is worth noting that this particular configuration is a hybrid between the French and American models (Lanciano-Morandat and Nohara, 2002). Those who created and managed it had previously visited the USA, where they learned how to handle applied research and to launch private entrepreneurial initiatives. This mind-set has since been handed on to the younger generations.

236


8. Jean-Louis Beffa, former chairman of the multinational Saint-Gobain, was the leader of an expert commission created by President Chirac.

REFERENCES Amable, B., R. Barré and R. Boyer (1997), Les systèmes d’innovation à l’ère de la globalisation, Paris: Economica. Aoki, M. (2001), Toward a Comparative Institutional Analysis, Cambridge, MA: MIT Press. Barré, R. and P. Papon (1998), ‘La compétitivité technologique de la France’, in H. Guillaume (1998), ‘Rapport de mission sur la technologie et l’innovation’, submitted to the Ministry of National Education, Research and Technology, the Ministry of the Economy, Finances and Industry, and the State Secretary for Industry, Paris, mimeo, pp. 216–27. Beffa, J.-L. (2005), ‘Pour une nouvelle politique industrielle’, Rapport remis au Président de la République Française, Paris, mimeo. Boltanski, L. and E. Chiappello (1999), Le nouvel esprit du capitalisme, Paris: Gallimard. Boltanski, L. and L. Thévenot (1991), De la Justification. Les Economies de la Grandeur, Paris: Gallimard. Branciard, A. (2001), ‘Le génopole d’Evry: une action publique territorialisée’, Journées du Lest, avril, mimeo, Aix en Provence. Branciard, A. and E. Verdier (2003), ‘La réforme de la politique scientifique française face à la mondialisation: l’émergence incertaine d’un nouveau référentiel d’action publique’, Politiques et Management Public, 21 (2), 61–81. Callon, M. (1991), ‘Réseaux technico-économiques et flexibilité’, in R. Boyer and B. Chavance (eds), Figures de l’irréversibilité, Editions de l’EHESS. Callon, M. (1992), ‘Variété et irréversibilité dans les réseaux de conception et d’adoption des techniques’, in D. Foray and C. Freeman (eds), Technologie et richesse des nations, Paris: Economica, 275–324. Casper, S., M. Lehrer and D. Soskice (1999), ‘Can high-technology industries prosper in Germany? Institutional frameworks and the evolution of the German software and biotechnology industries’, Industry and Innovation, 6 (1), 6–26. DTI (Department of Trade and Industry) (1999a), Our Competitive Future: UK Competitive Indicators 1999, London: Department of Trade and Industry. DTI (Department of Trade and Industry) (1999b), Biotechnology Clusters, London: Department of Trade and Industry. DTI (Department of Trade and Industry) (1999c), Genome Valley, London: Department of Trade and Industry. DTI/DfEE (Department of Trade and Industry/Department for Education and Employment) (2001), Opportunity for All in a World of Change: A White Paper on Enterprise, Skills and Innovation, London: HMSO. Ergas, H. (1992), A Future for Mission-oriented Industrial Policies? A Critical Review of Developments in Europe, Paris: OECD. Ernst & Young (2000), Gründerzeit. Zweiter Deutscher Biotechnologie-Report 2000, Stuttgart: Ernst & Young. Etzkowitz, H. and L. Leydesdorff (1997), Universities and the Global Knowledge


237

Economy. A Triple Helix of University–Industry–Government Relations, London and Washington, DC: Pinter. Etzkowitz, H. and L. Leydesdorff (2000), ‘The dynamics of innovation: from National Systems and “Mode 2” to a triple helix of university–industry–government relations’, Research Policy, 29, 109–23. Eymard-Duvernay, F. (2002), ‘Conventionalist approaches to enterprise’, in O. Favereau and E. Lazega (eds), Conventions and Structures in Economic Organization, New Horizons in Institutional and Evolutionary Economics, Cheltenham, UK and Northampton, MA, USA: Edward Elgar, pp. 60–78. Foray, D. (2000), ‘Inerties institutionnelles et performances technologiques dans la dynamique des systèmes d’innovation: l’exemple français’, in Michèle Tallard, Bruno Théret and Didier Uri, Innovations institutionnelles et territoires, coll. Logiques Politiques, Paris: L’Harmattan, pp. 81–100. Georghiou, K. (2001), ‘The United Kingdom national system of research, technology and innovation’, in P. Larédo and P. Mustar (eds), Research and Innovation Policies in the New Global Economy: An International Comparative Analysis, Cheltenham, UK and Northampton, MA, USA: Edward Elgar, pp. 253–96. Gibbons, M. (ed.) (1994), The New Production of Knowledge, The Dynamics of Science and Research in Contemporary Societies, London: Sage. Guillaume, H. (1998), ‘Rapport de mission sur la technologie et l’innovation’, submitted to the Ministry of National Education, Research and Technology, to the Ministry of the Economy, Finances and Industry, and to the Secretary of State for Industry, Paris, mimeo. Hall, P. and D. Soskice (eds) (2001), Varieties of Capitalism: The Institutional Foundations of Comparative Advantage, Oxford: Oxford University Press. Jobert, B. (ed.) (1994), Le tournant néo-libéral en Europe, coll. Logiques Politiques, Paris: L’Harmattan. Lam, A. (2001), ‘Changing R&D organization and innovation: developing the next generation of R&D knowledge workers’, Benchmarking of RTD Policies in Europe: A Conference organized by the European Commission, Directorate General for Research, 15–16 March. Lam, A. (2002), ‘Alternative societal models for learning and innovation in the knowledge economy’, International Journal of Social Science, 171, 67–82. Lamont, M. and L. Thévenot (eds) (2000), Rethinking Comparative Cultural Sociology: Repertoires of Evaluation in France and the United States, Cambridge: Cambridge University Press. Lanciano-Morandat, C. and H. Nohara (2002), ‘Analyse sociétale des marchés du travail des scientifiques: premières réflexions sur la forme professionnelle d’hybridation entre la science et l’industrie’, Economies et Sociétés, Série ‘Economie du travail’, AB, 8 (22), 1315–47. Lanciano-Morandat, C. and H. Nohara (2005), ‘Comparaison des régimes de Recherche et Développement (R/D) en France et au Japon: changements récents analysés à travers les trajectoires historiques’, Revue française d’Administration Publique, 112, 765–76. Larédo, P. and P. Mustar (2001), ‘French research and innovation policy: two decades of transformation’, in P. Larédo and P. Mustar (eds), Research and Innovation Policies in the New Global Economy: An International Comparative Analysis, Cheltenham, UK and Northampton, MA, USA: Edward Elgar, pp. 447–95.

238


Lundvall, B.-Å. and S. Borras (1997), The Globalising Learning Economy: Implications for Innovation Policy, DG XII, EC, Brussels. Lundvall, B.-Å. and B. Johnson (1994), ‘The learning economy’, Journal of Industry Studies, 1 (2), 23–42. Maurice, M., F. Sellier and J.-J. Silvestre (1986), The Social Foundations of Industrial Power. A Comparison of France and Germany, Cambridge, MA: MIT Press. Merton, R. (1973), The Sociology of Science: Theoretical and Empirical Investigations, Chicago, IL: Chicago University Press. Meyer-Krahmer, F. (2001), ‘The German innovation system’, in P. Larédo, and P. Mustar (eds), Research and Innovation Policies in the New Global Economy: An International Comparative Analysis, Cheltenham, UK and Northampton, MA, USA: Edward Elgar, pp. 205–51. Meyer-Krahmer, F. and U. Schmoch (1998), ‘Science-based technologies; university–industry interactions in four fields’, Research Policy, 27 (8), 835–52. OECD (1998), Technology, Productivity and Job Creation – Best Policy Practices, Paris: OECD. OECD (2000), Science, Technology and Industry Outlook 2000, Paris: OECD. Polanyi, M. (1962), ‘The Republic of Science: its political and economic theory’, Minerva, 1 (1), 54–73. Pestre, D. (1997), ‘La production des saviors entre économie et marché’, Revue d’Economie Industrielle, 79, 163–74. Porter, M. (1998), ‘Clusters and the new economics of competition’, Harvard Business Review, 76, 77–90. Powell, W.W. and J. Owen-Smith (1998), ‘Universities and the market for intellectual property in the life sciences’, Journal of Policy Analysis and Management, 17 (2), 253. Salais, R. (1998), ‘Action publique et conventions: état des lieux’, in J. Commaille and B. Jobert (eds), Les métamorphoses de la régulation politique, Paris: LGDJ, pp. 55–82. Saxenian, A.L. (1996), Regional Advantage: Culture and competition in Silicon Valley and Route 128, Cambridge, MA: Harvard University Press. Shinn, T. (2002), ‘Nouvelle production du savoir et triple hélice: tendances du prêt-à-penser des sciences’, Actes de la Recherche en Sciences Sociales, N. spécial Science, 141–2, 21–30. Storper, M. (1998), ‘Conventions and the genesis of institutions’, http://216.239.59. 104/search?q=cache:TnNqv5cKKbQJ:www.upmf-grenoble.fr/irepd/regulation/ Journees_d_etude/Journee_1998/Storper.htm+Michael+Storper&hl=fr. Storper, M. and R. Salais (1997), Worlds of Production: The Action Frameworks of the Economy, Cambridge, MA: Harvard University Press. Suleiman, E. (1995), Les ressorts cachés de la réussite française, Paris: Le Seuil. Vinck, D. (1999), ‘Les objets intermédiaires dans les réseaux de coopération scientifique, contribution à la prise en compte des objets dans les dynamiques sociales’, Revue Française de Sociologie, XL (2), 385–414.


239

METHODOLOGICAL APPENDIX Within the SESI project (Higher Education Systems and Industrial Innovation, funded by the EC), from 1998 to July 2001, national public action trajectories concerning RDI were studied from two angles: that of public policies at national and local level, and that of firms’ practices, especially from the point of view of relations with public research, both basic and applied. Three sectors were chosen in each country as being representative of the new challenges emerging for the relationship between higher education and industry in key areas where generic technologies are tending to develop, albeit in different ways. The information technology sector, whose growth has been very rapid, is of interest because it brings together industrial production activities and customer service activities, in ways specific to individual countries. The telecommunications sector, which has undergone a huge amount of technological and organizational innovation, was having its links with the public sector challenged by deregulation in various EU countries just as the project was being launched. The pharmaceutical sector, whose links with higher education and research date back further, was facing the biotechnology revolution. This project was based on the results of investigations carried out in firms. Three companies per sector and per country were studied, making a total of about 40. The initial idea was to take one ‘foreign’ multinational, one large ‘national’ company and one SME from each sector, in an attempt to have a comparable sub-sample for at least two countries. On the company side, the interviewees were R&D managers, project managers, research workers and engineers, HR managers and those responsible for related fields such as contracts and patents. Among their academic partners, interviews were conducted with heads of laboratories, departments and projects, sometimes with research workers. Semistructured interviewing techniques were used with both types of partner, based on a standardized interview guide devised for all firms in the various countries. Each case study is divided into two parts. The first deals with the firm’s trajectory and strategy in respect of innovation, competencies and knowledge. The second contains a presentation of some actual cases of collaboration between the firm and the higher education system in the two fields of research and training (competencies).

240


APPENDIX Table 8A.1

R&D expenditure: share in GDP in Germany, the UK and France (in %) 1981

G

UK

1991 F

G

UK

1995 F

G

UK

2001 F

G

UK

2003 F

G

UK

F

Share in GDP

2.43 2.38 1.93 2.52 2.07 2.37 2.25 1.95 2.31 2.51 1.86 2.23 2.52 1.88 2.18

Source:

OECD, PIST, May 2004, OST 2006.

Table 8A.2

R&D expenditure: breakdown by sources of funds in Germany, the UK and France Germany

(1) (2) (3) (4)

UK

France

1995

1999

2001

2003

1995

1999

2001

2003

1995

1999

2001

2003

60.0 37.9 0.3 1.8

65.4 32.1 0.4 2.1

65.7 31.4 0.4 2.5

66.1 31.1 0.4 2.3

48.2 32.8 4.5 14.5

48.5 29.9 5.0 17.3

46.9 29.1 5.7 18.2

43.9 31.3 5.4 19.4

48.3 41.9 1.7 8.0

54.1 36.8 1.9 7.0

54.2 36.9 1.7 7.2

50.8 40.8a – 8.4

Notes: (1) = Industry; (2) = Public funding; (3) = Other national sources; (4) = Foreign funding. Source:

OECD, Data MSTI, May 2005.

Table 8A.3

World shares (European patents in % )

Countries

1990

1996

2000

2004

Germany UK France EU-15a USA

20.9 7.0 8.3 48.1 26.4

17.7 5.8 7.1 43.0 33.1

18.1 5.3 6.3 42.6 32.3

16.4 4.8 5.6 39.7 30.5

Notes:

a

EU-25 in 2004.

Source: Data INPI and DEB, OST Processing.


Table 8A.4

241

World shares (scientific publications in %)

Countries

1993

1995

2000

2004

Germany UK France EU-15a US

6.7 8.1 5.2 32.0 33.9

6.7 8.3 5.4 32.9 32.7

7.2 7.7 5.3 33.6 30.1

6.4 6.7 4.7 34.2 27.1

Notes: Source:

a

EU-25 in 2004. Data ISI, OST Processing 2006.

Table 8A.5

Scientific publications: European shares by disciplines in % (2004) and evolution from 1999 to 2004

Disciplines

2004

Evolution 2004/1999 (%)

Germany

UK

Basic biology Medical research Applied biology/ ecology Chemistry Physics Space sciences Sciences for engineering Mathematics

18.7 18.7

20.8 23.4

13.8 11.9

15.9

18.3

20.6 23.1 16.7 16.8

Total Source:

UK

France

−2 0

−7 −6

−11 −12

11.9

−10

−13

−10

14.8 13.7 20.5 19.0

14.5 15.8 14.5 13.5

−14 −9 −4 −13

−7 −10 −12 −20

−8 −5 −9 −2

15.9

12.7

20.5

−18

−8

−3

18.8

19.5

13.6

−6

−9

−9

Data ISI, OST Processing 2006.

France Germany

242


Table 8A.6

European patents by technological domains: European shares in % (2001) and evolution from 1996 to 2001

Technological domains Electronics– electricity Instruments Chemistry– materials Pharmaceuticals– biotechnology Industrial processes Mechanics– transports equipments Household equipment– construction Total Source:

2001

Evolution 2001/1996 (%)

Germany

UK

France

Germany

37.7

12.9

15.2

+3

−7

−21

41.5 47.1

15.8 13.2

14.0 13.2

+11 −4

−4 −8

−22 −6

30.4

19.9

19.6

+12

−3

−4

42.2

10.1

13.0

−4

−9

+1

52.2

8.4

13.7

+14

−22

−20

38.9

12.8

14.9

−2

+11

−6

42.3

12.6

14.5

+3

−6

−12

Data ISI, OST Processing 2004.

UK

France

9.

The role of boundary organizations in maintaining separation in the triple helix Sally Davenport and Shirley Leitch

INTRODUCTION Life Sciences Network, an umbrella group of industry and scientists who support genetic engineering, wants the chance to contradict evidence given by groups opposed to GE and to put new evidence before [the New Zealand Royal Commission on Genetic Modification]. (Beston, 2001, p. 8) A new hybrid organizational representation of action that is neither purely scientific nor purely political is created. (Moore, 1996, p. 1621)

The triple helix is said to consist of co-evolving networks of communication between three institutional players: universities (and other research organizations), industry and government (Leydesdorff, 2000). This separation into three separate ‘strands’ implies the existence of only three players, with distinct boundaries between each sphere of activity. However, there are other organizations that mediate the interaction between science, industry and government, such as bioethics councils (Kelly, 2003) and environmental groups (Guston, 2001; Miller, 2001). These new boundary organizations have arisen in the triple helix, in order to manage ‘boundary work’ and ‘boundary disputes’ as a result of the ‘new demands on researchers and their organizations’ (Hellstrom and Jacob, 2003, p. 235). Boundaries are drawn based on the assumption that they describe ‘a stable order’. This implies that the ordering of human actions and interactions circumscribed by the boundary is restraining (Hernes and Paulsen, 2003, p. 6). However, boundaries can be both enabling and constraining (Hernes, 2003, 2004). In the mainstream management literature there is much talk of breaking down boundaries, as they are perceived to be a barrier to communication and information flows (Heracleous, 2004). In this literature organizational boundaries are perceived to be ‘real’, physical entities. However, for others, boundaries are ‘complex, shifting socially constructed entities’ (Heracleous, 2004), created to maintain both 243

244


internal order and external protection but also as ‘a means of establishing the organization as an actor in its relationships with other organizations’ (Hernes, 2003, p. 35). Thus the concept of a boundary is complex and there is considerable ambiguity surrounding the role that boundary organizations play in the triple helix (Kelly, 2003). This chapter will explore the role of one boundary organization, an incorporated society called the Life Sciences Network (LSN), which was a vehicle for presenting a pro-GM (genetic modification) stance in a national debate on behalf of the research and industry organizations that it represented. It was active in New Zealand during an election, the Royal Commission into GM (RCGM) and subsequent debates. In many ways, the LSN was similar to other boundary organizations described in the literature in that it was a vehicle for information provision and coordination of interests between triple-helix players. However, in other ways it was distinctly different. The LSN, whose membership included a range of scientific and other publicly funded organizations that could not all openly engage in debate and lobbying for various reasons, shielded its member organizations from significant engagement in the debate, especially with anti-GM activist groups. In this way, this new sort of boundary organization not only provided demarcation between science and non-science but also allowed deniability. In many ways, boundary organizations like the LSN are similar to front organizations in that they play a distinctive communication role that would be politically difficult for the individual members to undertake. The use of front organizations is a common subject of discussion in political science but they have not been recognized as potential boundary organizations in science studies. The role of lobby or interest groups, representing a variety of causes, as a vehicle for the provision of information and other resources to politicians and law makers, in return for influence, is well recognized in the former literature (e.g. Holyoke, 2003; Ainsworth and Itai, 1993; Baumgartner and Leech, 2001). Yet we know little about the use of these potentially very influential boundary organizations as mediators in the triple helix.

BOUNDARIES Boundaries are a popular subject in many social science subjects. For example, ‘the concept of boundaries has been at the center of influential research agendas in anthropology, history, political science, social psychology, and sociology’ (Lamont and Molnar, 2002, p. 167). Working from descriptions of boundaries in sociology, Hernes (2003) described

Boundary organizations in the triple helix

245

three types of boundaries: physical, social and mental. Physical boundaries not only consist of ‘real’ objects and structures, but also of rules and regulations governing the exchanges that may take place within the organization or between the organization and its environment. Physical boundaries can provide stability and predictability as ‘they serve to bind resources over time and space’. They also ‘serve to create and consolidate impressions of robustness’ externally, helping to ‘establish an organization as a more recognized and powerful actor in a field of organizations’ (Hernes, 2003, p. 38). Groups and organizations develop social boundaries to ‘distinguish themselves from others’. The sense of identity created by a social boundary enables the ‘creation of “otherness”’ and what is not part of the organization. ‘Social boundaries depend on social interactions for their existence’, are ‘central to the creation of behavioral norms’ and also ‘uphold patterns of social power’ (Hernes, 2003, p. 39). Symbolic boundaries become social boundaries when their characteristics have been agreed upon and ‘pattern social interaction in important ways’ (Lamont and Molnar, 2002, p. 169). They also provide protection to the group when its identity is perceived to be threatened. Strong social boundaries may span formal boundaries and can enable cooperation between members without close proximity (Hernes, 2003, p. 39). Mental boundaries are also about making distinctions and assist in how individuals make sense of the world. They form the basis for ‘action inside and outside the borders of the group’, and become embedded as collectively held ‘tacit assumptions’. They allow the segmentation between ‘us’ and ‘them’ (Lamont and Molnar, 2002). Like social boundaries, they are moderated by social interaction but are also guarded as a ‘basis for power’ and a ‘shield’ from ‘the outside world’ yet to be understood (Hernes, 2003, p. 40). To summarize, there are different types of boundaries that may not necessarily coincide with each other. For the purposes of this chapter, physical, social and mental boundaries allow not only the creation of a seemingly ‘stable’ concept of what is ‘inside’ the boundary; they also enable a demarcation of what is ‘outside’ the boundary and therefore facilitate expedient interaction, or boundary work, with the external environment.

BOUNDARY WORK The concepts of boundary organizations and boundary activities are not new in science studies (Guston, 2001) and are often couched in terms of the ‘problem of demarcation’ that has been concerning philosophers and

246


sociologists of science. In a key paper on boundary work, Geiryn (1983, p. 781) described the traditional approach to the ‘problem of demarcation’ as ‘how to identify unique and essential characteristics of science that distinguish it from other kinds of intellectual activities’. The same author likened boundary work to the literary concept of a ‘foil’ and suggested that we are better able to learn about ‘science’ through contrasts to ‘nonscience’ (ibid., p. 791). In this way, ‘the power and prestige of science is typically thought to be grounded in the ability of scientists to draw strong distinctions between scientific and non-scientific interests’ (Moore, 1996, p. 1592). Boundary work was originally conceived to explain how science defended the boundary of its communities from attack and was a problem of ‘how to maintain control over the use of material resources by keeping science autonomous from controls by government or industry’. ‘Boundary work is an effective ideological style for protecting professional autonomy . . . the goal is immunity from blame for undesirable consequences of nonscientists’ consumption of scientific knowledge’ (Geiryn, 1983, p. 789). Geiryn identified three occasions when boundary work was likely to be employed as a resource for scientific ideologists. First, ‘when the goal is expansion of authority or expertise’ into domains claimed by others, that is ‘boundary work heightens the contrast between rivals in ways flattering to the ideologist’s side’. Second, when the goal is monopolization of professional authority and resources, boundary work excludes rivals from within by defining them as outsiders with labels such as ‘pseudo’, ‘deviant’ or ‘amateur’. Third, ‘when the goal is protection of autonomy over professional activities, boundary work exempts members from responsibility for consequences of their work by putting the blame on scapegoats from outside’ (ibid., pp. 791–2). Moore (1996) studied the rise of US public interest science organizations, such as the Union for Concerned Scientists, as ‘boundary work’ activities, in that they ‘created a new form of action among scientists that was deemed neither purely scientific nor purely political’, and that they ‘permitted the preservation of organizational representations of pure, unified science, while simultaneously assuming responsibilities to serve the public good’. In the case of public interest science organizations, science and its relation to politics became the main object of ‘action’, and enabled the alignment between the interests of scientists and their patrons (Moore, 1996, pp. 1592–8). However, the fact that scientists are portrayed as ‘unified’ is problematic in descriptions of boundary work. Moore notes that ‘scientists, like other people, sometimes exhibit commitments to multiple social identities’ so that notions of unified scientists ignore ‘variation in when, how and


247

around what issues they are unified’. ‘Scientists exhibit a wide range of political opinions, religious beliefs, and income levels; these differences impinge upon the kinds of claims that scientists make about the proper relationships between science and politics as well as forming the basis for conflict among scientists. Thus, the process of setting boundaries is not simply a struggle between a unified group of scientists and non-scientists, but a process of struggle among scientists as well’ (Moore, 1996, p. 1596). Geiryn also saw this variation as a source of ambiguity in the notion of boundary work, noting that ‘demarcation is as much a practical problem for scientists as an analytical problem for sociologists and philosophers’ (1983, p. 792). He argued that ambiguity surfaces because of inconsistencies in the boundaries constructed by different groups of scientists internally or in response to different external challenges, as well as because of the conflicting goals of different scientists. Moore noted that the formation of the boundary organizations helped perpetuate the perception of unity in science by preserving ‘the professional organizations that represented “pure” science and unity among scientists’ (1996, p. 1594). The demarcation problem is still of great interest to science studies (e.g. Evans, 2005 and papers in that special issue of Science, Technology & Human Values). However, in more recent times, boundary work has evolved to mean the ‘strategic demarcation between political and scientific tasks in the advisory relationship between scientists and regulatory agencies’ (Guston, 2001, p. 399). This newer version of boundary work has taken on a less instrumental tone and the boundaries are viewed as means of communication rather than of division (Lamont and Molnar, 2002). The demarcation between politics and science is viewed as something to bridge, with boundary organizations seen as mediators between the two realms (Miller, 2001), and successful boundary organizations are said to be those that please both parties. In this new framing, boundary organizations may help manage boundary struggles over authority and control but are primarily focused on facilitating cooperation across social domains in order to achieve a shared objective. They ‘are useful to both sides’ but ‘play a distinctive role that would be difficult or impossible for organizations in either community to play’ (Guston, 2001, p. 403, citing the European Environmental Agency).

THE RESEARCH PROJECT The research reported in this chapter is derived from a much larger five-year study on the sociocultural impact of biotechnology in New Zealand funded by the New Zealand Foundation for Research Science &

248


Technology (UoWX0227). The study has involved hundreds of interviews and focus groups with organizations, groups and individuals involved with biotechnology as scientists, policy managers, entrepreneurs, consumers, members of ethnic, cultural or religious groups, environmentalists and activists. These interviews have been augmented by a thematic analysis of websites, official documents, annual reports, promotional material, media material and other documents. They have been further augmented by researcher observation of various events, such as conferences and workshops, staged by actors in the biotechnology sector. For this chapter, we drew on interviews with triple-helix players (researchers, government and industry representatives), with LSN, Royal Society of New Zealand (RSNZ) and Association of Crown Research Institutes (ACRI) representatives, as well as secondary data such as media and website material. The data were analysed and categorized according to our interest in the ‘boundary’ aspects of the strategies and action of the LSN, its members and other triple-helix players. A case study of the LSN was developed, which is now summarized to provide context for our discussion of the LSN as an interesting boundary organization.

THE LIFE AND TIMES OF THE LIFE SCIENCES NETWORK The Life Sciences Network Incorporated (LSN) was a non-profit body incorporated in 1999 with the overall objective, according to its Executive Director Francis Wevers, to advocate ‘cautious, selective and careful’ use of GM technology ‘to deliver benefits to New Zealand as a whole’. ‘While the emotional non-scientific issues were important’, Wevers said, ‘they’ve got to be considered in balance . . . we’ve got to be informed by knowledge, not by emotion.’ As indicated by the objectives laid out in the LSN Constitution (Box 9.1), LSN essentially wanted to coordinate responses to GM issues and to provide information to decision-makers and the public. In part, LSN was a successor to the Gene Technology Information Trust (more commonly known as ‘GenePool’), which had also claimed to be an independent provider of information on GM. However, GenePool had been wound up in September 1999 after concern was raised by the Green Party over its support from Monsanto (e.g. Espiner, 1999) and for an apparent conflict of interest (e.g. Samson, 1999) between a communications company that worked both for GenePool and for New Zealand King Salmon (Weaver and Motion, 2002), a company that had allegedly attempted to cover up deformities in some genetically modified salmon it had bred.


249

BOX 9.1 OBJECTIVES OF THE LSN The [LSN] is formed to promote the following objectives: 1. To promote the strategic economic opportunity available to each country to benefit from the application of biotechnology in the expanding knowledge age. 2. To assist members to effectively input into public biotechnology research and development policy, both individually and collectively. 3. To positively influence the advancement of responsible biotechnology research and development within an appropriate framework of regulatory controls based on scientific and risk management principles. 4. To positively influence the continued availability of: a. medicine derived from biotechnology or genetic engineering; b. products used in crime prevention, food safety testing and other similar applications, which involves or are derived from biotechnology; c. biotechnology applications and advances which benefit each country, including but not limited to, applications in manufacturing and production, pest control, environmental protection, production of quality food products, human and animal health and welfare. 5. To provide: a. An active voice for the creation of a positive environment for responsible use of genetic modification with appropriate caution; b. Accurate and timely information and advocacy to respond to issues related to biotechnology as they arise, and; c. Assisting and obtaining cooperation of organisations that will or may benefit from the application of genetic modification procedures. Source:

Extracted from the LSN Constitution, 2002.

250


The formation of LSN was driven by William Rolleston, a doctor who came from an established farming tradition but had also launched a very successful biopharmaceuticals firm, South Pacific Sera Ltd. He was also the founding chairman of Biotenz, an association for biotechnology businesses, which ‘use the application of science and technology to create biological-based wealth’ (Worrall, 2002). In 2000, when New Zealand environmental groups started voicing concerns about GM, Biotenz decided that there was a need for another body to voice opinions. ‘Biotenz wanted to add balance and facts so that debate on GM could be rigorous and well informed. We wanted to counteract the “mythinformation”, the scaremongering talk of things like Frankenfood science-fiction nonsense that created unnecessary fear among the public’ (Worrall, 2002). In July 2002, LSN had a membership of 22 organizations including 13 industry organizations (with one based in Australia) and producer boards, four Crown Research Institutes (CRIs – primarily government-funded research organizations), three universities, one independent research organization and one industry cooperative. Members paid between NZ$1000 and $20 000 per year depending on the size of the organization, contributing to the LSNs annual running costs of about NZ$300 000, with additional costs such as the extra $500 000 during the RCGM (Fisher, 2003). Together, the member organizations apparently represented producers of up to about 70 per cent of New Zealand’s gross domestic product through the member industry organizations, which included many farmer and manufacturer groups (Worrall, 2002). There were many questions over its membership, particularly from those who believed that public funding should not be used to fund what was, essentially, a lobby group (e.g. Hawkes, 2000; Collins 2002a, 2002b). The LSN was probably most active, and had its highest profile, around the time of the general election in 2002 (during which GM was a major election issue1), during the RCGM in 2001/2002 and when the moratorium on the release of GM crops was lifted in 2003. It was given ‘interested person’ status2 during the RCGM, even though many of its constituent groups were also granted such status independently, arguing that it could play a strong role in coordinating evidence in relation to scientific arguments that were likely to be raised by ‘multinational’ organizations such as Greenpeace and Friends of the Earth (Samson, 2000). During these hectic few years, there were, at times, daily press releases and articles in various media, providing the LSN perspective on the potential of GM or rebutting the views of those against the GM-based technology. The LSN espoused a very public strategy for its involvement in the RCGM (Box 9.2). In particular, the LSN desired that the totality of its


251

BOX 9.2 LSN CAMPAIGN STRATEGY FOR THE RCGM For the New Zealand Life Sciences Network (Inc.) to: 1. Build public, and therefore political confidence in gene technology, science and scientists through information, education and knowledge. 2. Coordinate the activities of member organisations and network participants to ensure the public, media, politicians and Royal Commission are presented with the most complete case in favour of continued responsible research, development and application of gene technology in New Zealand. 3. Work with non-member organisations to seek to achieve its desired campaign outcome. Implications of strategy 1. While the strategically important campaigns run outside, the Royal Commission is the focal point from which other activity can be leveraged. 2. The totality of submissions and evidence to the Royal Commission of Inquiry into Genetic Modification by the Network and its members should constitute a shadow Royal Commission report and recommendations. 3. Fully engage with Royal Commission to ensure a complete case is presented and heard. 4. Every misrepresentation or erroneous accusation made by the organisations of anti-GMO (that is Greenpeace, Friends of the Earth, Green Party) should be challenged and rebutted. 5. Fears expressed by ordinary New Zealanders should be listened to and acknowledged. 6. The focus of all communications, messages, submissions, evidence should be to establish the truth of the assertions as stated in our desired outcomes. 7. Expert witnesses from New Zealand and overseas should be available for media interviews; public meetings and other opportunities to build public confidence through information, education and knowledge.

252


8. All opportunities for positive media coverage of the submissions and evidence of the Network, its members and witnesses should be taken. 9. Scientists must be prepared to stand up to re-build public trust in their science and themselves. 10. Scientists must be prepared to engage and win over, through logical debate, church leaders because they are still perceived to be the arbiters of ethical standards. Source: Extracted from the LSN website education archives, http://www. lifesciencesnetwork.com/educationarchives.asp, accessed March 2005.

submissions, along with member group submissions, should ‘constitute a shadow Royal Commission report and recommendations’. The outcome of the RCGM, with its overall recommendation to ‘keep options open’ and to ‘preserve opportunities’, was then most reassuring that the LSN was accomplishing its objectives. Looking back on the results of the RCGM, Wevers stated that the LSN could have written the Royal Commission’s report . . . the Royal Commission basically accepted the moderate position we took . . . all that reflects is the fact that the people in organizations which made up the Life Sciences Network were not some radical outrageous group of people looking at this in an irresponsible way . . . [but] that the position that we had come to was one which was consistent with good policy, and consistent with the strategic interests of New Zealand, which is what the Royal Commission was about.

With the government’s decision to ‘proceed with caution’, and the subsequent lifting of the moratorium on GM in October 2003, the initial objectives of the LSN were largely satisfied, but members expressed a wish that the capabilities of the LSN not be lost and that, whilst the GM issue might have subsided, there were other issues that could be addressed, albeit not by the LSN in its present incarnation. Thus the LSN evolved into the Biosciences Policy Institute, set up to promote ‘the education of the public of New Zealand in matters relating to the biological sciences’ by providing non-governmental policy analysis and development. Wevers was the Institute’s executive director and a former prime minister chaired it. Wevers’s intentions for the ‘very, very low profile’ organization were to ‘reflect the diversity of opinions and to encourage people and decision makers to make policy decisions on the basis of the best evidence and the best scientific support’. A news service, which had been very heavily subscribed during the RCGM period, was to expand to six-weekly


253

digests, a journal was to be launched and a biennial conference was to be organized. In May 2004, after less than a year in operation, the Biosciences Policy Institute announced it was closing down due to ‘insufficient financial support’ (Collins, 2004). Of the demise of the Biosciences Policy Institute and dormancy of the LSN, Wevers argued, ‘It was a single-issue entity set up right from the beginning to achieve a very specific objective, which made it easy at the end to say “OK, we’ve done our job”.’ However, the ‘absence of an important issue is the thing we haven’t been able to overcome . . . There is no public controversy anymore . . . the public has moved on.’ The LSN still exists, although it no longer has a staff resource. However, as Rolleston stated, ‘It’s still there should industry and science need to get together to deal with other issues where fear and over-reaction arise’ (Collins, 2004). In the same month that the Biosciences Policy Institute closed, Wevers was awarded the Supreme Award by the Public Relations Institute of New Zealand for the LSN campaign on GM.

DISCUSSION There is no doubt that, for its member organizations, the LSN was a very successful boundary organization in that it achieved the outcome that was desired through being very actively involved in the political discussion and negotiation over the future of GM in New Zealand. Like any lobby group, the purpose of this type of boundary organization was to supply information from a particular perspective, in the hope of influencing decisions made. Even though the physical aspects of the LSN (staff numbers, office space etc.) were very small relative to those of other organizations, the identity of the organization and the ‘boundary’ between the LSN and ‘science’ was very effectively created by its constitution, its membership consisting of representative industry organizations (not individual companies, having learnt from the maligned participation of Monsanto in GenePool) and research organizations (rather than individual scientists), its website and prolific press releases and media articles. By responding at every opportunity and rebutting any anti-GM sentiment expressed ‘in public’, the LSN became the ‘voice’ for pro-GM. As one industry representative stated, ‘the pro-GM side would have been a mess without [the LSN]’. In a societal debate, such as that which occurred around the role of GM in New Zealand, the audience was very diffuse and LSN’s aim was to directly influence ‘public opinion’, and ultimately political decisionmaking and resource allocation, in favour of the member organizations’ objectives but without the member organizations needing to actively

254


participate or publicly ‘own’ the perspective or information supplied, as indicated above. This is particularly important for members that were publicly funded, such as the CRI and university members of the LSN. Thus this type of boundary organization works towards, and enables, the legitimation of a view, such as a pro-GM approach to New Zealand’s future, rather than legitimation of the strategies of the member organizations. In other words, the front organization is the ‘active voice’ (a phrase used in the LSN constitution), which enables a separation to be maintained between the voiced perspective and those players who wish to express it. In many ways, this type of boundary organization is true to the original concept in that organizations like the LSN can be seen to be defending scientific boundaries, in this case of GM science, from attack. In addition to providing a voice for its members, the LSN also enabled New Zealand science professional organizations, such as the RSNZ and the ACRI, space to appear more moderate and conciliatory in the range of views of their scientist members. Of the delineation between the LSN and RSNZ, Wevers stated: The [RSNZ] avoided being involved in this issue because it was schizophrenic. It had one view which was the dominant view of the physical scientists and another view called ‘traditional scientist’ which was a view espoused by a group of social scientists and the two views were in conflict. So the [RSNZ] was unable to take a leadership role in this and it was uncomfortable about that but realised its own political reality. The [LSN] was specifically set up to be a focal point and we weren’t going to beat around the bush and pretend we weren’t advocates.

Even though the ACRI had common membership with the LSN, with four of the nine CRIs also members of the LSN, it was constrained by its representative function. As explained by Anthony Scott, executive director of ACRI: [ACRI] is the voice for the nine CRIs on matters which the CRIs have in common and on which they agree to have a common voice. We’re representative when we’ve been authorised but cannot bind any member . . . Almost all of [the CRIs] were involved in GM research. Some are really leading from the front. Some were, or were likely to be, primarily researching whether GM was ‘safe’ or what was likely to happen [if GM release occurred] in the New Zealand situation. So some CRIs felt that they had to not only be, but had to be seen to be, independent of an advocacy role. Some of course, were doing both. So the ACRI role was to say to New Zealand ‘let’s talk about it, let’s understand what the issues are, let’s have an informed discussion’.

Scott continued, ‘I think it was a wise strategy for those CRIs that wanted to more strongly advocate the use of GM to join the LSN. It avoided cross-fertilisation, if you like.’


255

As illustrated by these quotes, the LSN played a second demarcation role, which Moore (1996) also observed with the public interest science organizations, in that the LSN was able to undertake the ‘trench-fighting’ role that the RSNZ and the ACRI were unable to do (even if they had so wished) because of the diversity of their membership. Thus the LSN enabled these professional organizations to preserve their positions as representative organizations of ‘pure science’ calling for more informed debate while, at the same time, the LSN was able to be more ‘assertive and aggressive’ and undertake the ‘trench-fighting’ for pro-GM science. Scott said that, in his view, the LSN was useful in that ‘in a socially and politically contentious issue such as [GM], a number of the CRIs were able to mediate their activities through that and get involved in what was a highly politicised debate’. To some extent the same could be said of the role LSN played for the representative industry organizations that were LSN members. As Wevers argued, ‘these are organizations that have got diverse memberships, in which they didn’t want to have that debate internally because of the internal disruption that would occur and so, consequently, its been important to them in a strategic sense, to have these things said in public to enable debate to develop’. Thus this demarcation of the political role into another organization facilitates unity in the member organizations by minimizing the opportunities for disruptive debate internal to the organizations, but still facilitates debate in ‘public’. This is not just a need of ‘science’ but also perhaps a need of any organization that purports to be representative of diverse members on a range of issues, to have a ‘front’ organization play the more contentious role. The demarcation maintained by the LSN did more, however, than just remove the member organizations from active participation in the debate. The separation certainly helped the member organizations privilege their seemingly distant positions, which may (perhaps necessarily for a public sector organizations) have been more ‘neutral’ in stance. More significantly, however, the separation in the relationship allowed the member organizations to deny aspects of the front organization’s message that did not suit its circumstances at a particular time, such as when it was challenged by other stakeholders. As publicly funded institutes, the CRIs’ and universities’ participation in the LSN was open to challenge by parliamentarians and the media, as happened just before the 2002 New Zealand election when the Green Party leader questioned the suitability of membership of the four publicly funded CRIs in the LSN (Hawkes, 2000). Again, during the election, the appropriateness of ‘taxpayer’ funding (from the CRIs) being used for pro-GE advertisements (Collins, 2002a; 2002b) was queried. A typical media response from one CRI CEO illustrates the deniability factor enabled by the separation:

256


[The CEO] approved the [research organization’s] contribution as part of the institute’s wider programme of public education. ‘I think it’s appropriate for us to provide information to the public, and that is what the Life Sciences Network is mostly about – factual information.’ (Collins, 2002a)

The demarcation in the relationships with the boundary organization enabled diverse players to join together in a way that they might not be able to do otherwise, in order to facilitate an action that could not be achieved by individual players. This demarcation not only allows an individual organization to distance itself from the debate but also enables a deniability factor in that such ‘front’ organizations, having effectively created a ‘boundary’, are viewed as relatively independent of their members. This intended effect is illustrated in Rolleston’s rebuttal of accusations against LSN: ‘The [LSN] did not tell its members what to say or think, and the [member organizations] would be free to act independently. The main reason for joining was to exchange information’ (Hawkes, 2000). The ‘front’ aspect of boundary organizations also allows this demarcation to happen on a temporary basis, maintaining a separation while a strategy is facilitated. Once the desired action has been implemented, the member organizations no longer require the front organization as a voice for their strategy. The ambiguous separation between a boundary organization and its members is no longer required and the temporary boundary organization is ‘retired’. In this case, the LSN tried to evolve into a more permanent independent policy organization, but was unable find enough financial support from previous or prospective members. As Moore observes, ‘the longevity of an organization (or form of organization) is largely determined by the ability to obtain resources such as members or monies . . . to provide a product to those that need it, and to fit with prevailing ideas about legitimate forms of organization’ (1996, p. 1621). We propose that, unlike other boundary organizations discussed in the literature, it is in the nature of these ‘front’ and/or ‘single-issue’ boundary organizations that they are mostly, like the LSN, short-term entities, as the conditions that necessitate their creation no longer exist. Unlike Moore’s (1996) public interest science organizations that had survived for several decades, the characteristics of demarcation and deniability that come with such ‘front’ boundary organizations as the LSN run counter to the transparency and accountability usually accepted as part of the operations of publicly funded organizations. However, the (albeit temporary) demarcation and deniability afforded by organizations such as the LSN may, in fact, enable far less disturbance of the ‘system’ than if the member organizations had more openly engaged in the public debate and ‘trench-fighting’. Relationships could feasibly be severely and


257

irreversibly disturbed if players entered ‘the trenches’ themselves in order to voice perspectives or pursue strategies to ‘protect’ science and research organizations in the traditional sense of boundary work. Thus the role of such boundary organizations in temporarily providing separation contributes to sustaining the long-term stability of the relationships between players, even though the ambiguity (and potential deniability) in the relationship blurs the ‘normal’ boundaries between science and political realms during its existence. Using this line of reasoning, the fact that the Biosciences Policy Institute did not survive is not surprising, as the conditions that instigated the need for, and tolerance of, the LSN were no longer present in the sciento-political reality after the lifting of the GM moratorium. Although some may view the ambiguity that surrounds the role of these sorts of lobby groups as potentially disruptive to triple-helix relationships, our research suggests that these organizations will only be formed, and the ambiguity tolerated, under certain conditions. Attempts to prolong the life of such boundary organizations once conditions have changed may be problematic because the very ambiguity surrounding the separation between players is not likely to be tolerated on a longer-term basis in a new political environment. Thus such temporary boundary organizations can be viewed as serving a useful short-term function for the science they represent. As one industry representative reflected, ‘there was a job needing doing. [The LSN] was an organization for its time’. Temporary boundary organizations such as the LSN may, in fact, be a pragmatic alternative to direct engagement of different scientific players in political debates, which could potentially be very disruptive of the stable interior of ‘bounded’ science.

CONCLUSION The aim of this chapter was to provide an example of an organization intermediary between science, industry and government that extends the current conceptions of what roles such boundary organizations can play. The LSN was an organization set up to provide the active pro-GM voice that representative science (and industry) organizations were unable to undertake, during a societal debate on the role of GM in New Zealand. The LSN very effectively played the communication roles suggested for boundary organizations, especially with its public relations campaign aimed at providing politicians with accessible information that supported the LSN cause. However, we argue that the LSN also performed several other important functions. It provided the demarcation between science and politics, which boundary work was originally conceived to explain. It also enabled

258


an important element of deniability for the member science organizations when stakeholders questioned the use of public funds for promoting one ‘side’ of the debate. Both of these characteristics of the ‘boundary’ created by the LSN were particularly salient for the public-funded research organizations, but were also equally useful to representative industry organization members. The particular contribution of this chapter is also to propose that the existence of such boundary organizations is almost inevitably likely to be fairly short and determined by the life-cycle of the issues they represent in societal debate. Member organizations and their stakeholders (including ‘the public’) are only likely to tolerate the ambiguity inherent in the demarcated relationship while ‘the job needs doing’. We propose that such short-term boundary organizations be viewed as a pragmatic solution to the alternative of direct engagement in the debate by science, which could be far more disruptive to ‘science’ and to relationships between triple-helix players than the activities of a boundary organization separated at arm’s length from ‘representative’ bodies and science in general. Guston (2004) argued that we should be democratizing science by creating institutions and practices that fully incorporate principles of accessibility, transparency and accountability. ‘Science advising in government is unavoidably political, but we must make a concerted effort to make sure it is democratic’, Guston (2004, p. 25) suggested, adding that we should be asking, ‘What are the appropriate institutional channels for political discourse, influence and action in science?’ It is debatable whether the LSN would, from some perspectives at least, be categorized as a ‘democratic’ boundary organization given the demarcation and deniability characteristics. Yet it was tolerated for a period of time by many players in the societal debate and viewed as very successful by its members and some parts of science. Given that science has evolved its boundaries to supposedly encompass an unproblematic unified whole, such short-term ‘undemocratic’ boundary organizations appear to be tolerated in certain political conditions in order to preserve systems and relationships that underpin the long-term stability of ‘science’.

NOTES 1. During the lead-up to the election, the government was accused of having covered up the importation of corn with a modicum of GM contamination. ‘Corngate’ began with a very public ‘ambush’ of the prime minister with the accusations in a live TV interview. 2. An organization was able to claim ‘interested person’ status if it could prove it had an interest in the RCGM, over and above that of any interest in common with the public (Davenport and Leitch, 2005).


259

REFERENCES Ainsworth, S. and S. Itai (1993), ‘The role of lobbyists: entrepreneurs with two audiences’, American Journal of Political Science, 37, 834–66. Baumgartner, F. and B. Leech (1996), ‘Issue niches and policy bandwagons: patterns of interest group involvement in national politics’, Journal of Politics, 63, 1191–213. Beston, A. (2001), ‘GE parties fight to finish’, New Zealand Herald, 26 February. Collins, S. (2002a), ‘Taxpayer cash in pro-GE adverts’, New Zealand Herald, 25 July. Collins, S. (2002b), ‘Varsity unit gives cash to pro-GE fund’, New Zealand Herald, 26 July. Collins, S. (2004), ‘Pro-GM lobby institute closes’, New Zealand Herald, 7 May. Davenport, S. and S. Leitch (2005), ‘Agora, ancient & modern and a framework for public debate’, Science & Public Policy, 32 (3), 137–53. Espiner, G. (1999), ‘Accepting Monsanto money naïve, says Kirton’, The Evening Post, 3, September, 2. Evans, R. (2005), ‘Introduction: demarcation socialized: constructing boundaries and recognizing difference’, Special Issue of Science, Technology & Human Values, 30, 3–16. Fisher, D. (2003), ‘Money talks for pro-GE spin machine’, Sunday Star Times, 16 November. Geiryn, T. (1983), ‘Boundary-work and the demarcation of science from nonscience: strains and interests in the professional ideologies of scientists’, American Sociological Review, 48, 781–95. Guston, D. (2001), ‘Boundary organizations in environmental policy and science: an introduction’, Special Issue of Science, Technology, & Human Values, 26 4), 399–408. Guston, D. (2004), ‘Forget politicizing science. Let’s democratise science!’, Issues in Science and Technology, 21, 25–8. Hawkes, J. (2000), ‘Row flares over council’s pro-GE stance’, Waikato Times 28 August, 1. Hellstrom, T. and M. Jacob (2003), ‘Boundary organizations in science: from discourse to construction’, Science & Public Policy, 30 (4), 235–8. Heracleous, L. (2004), ‘Boundaries in the study of organization’, Human Relations, 57, 95–103. Hernes, T. (2003), ‘Enabling and constraining properties of organizational boundaries’, in N. Paulsen and T. Hernes (eds), Managing Boundaries in Organizations: Multiple Perspectives, New York: Palgrave, pp. 35–54. Hernes, T. (2004), ‘Studying composite boundaries: a framework for analysis’, Human Relations, 57, 9–29. Hernes, T. and N. Paulsen (2003), ‘Introduction: boundaries and organization’, in N. Paulsen and T. Hernes (eds), Managing Boundaries in Organizations: Multiple Perspectives, New York: Palgrave, pp. 1–13. Holyoke, T. (2003), ‘Choosing battlegrounds: interest group lobbying across multiple venues’, Political Research Quarterly, 56 (3), 325–36. Kelly, S. (2003), ‘Public bioethics and publics: consensus, boundaries, and participation in biomedical science policy’, Science, Technology, & Human Values, 28 (3), 339–64.

260


Lamont, M. and V. Molnar (2002), ‘The study of boundaries in the social sciences’, Annual Review of Sociology, 28, 167–95. Leydesdorff, L. (2000), ‘The triple helix: an evolutionary model of innovations’, Research Policy, 29, 234–55. Miller, C. (2001), ‘Hybrid management: boundary organizations, science policy, and environmental governance in the climate regime’, Science, Technology & Human Values, 26 (4), 478–501. Moore, K. (1996), ‘Organizing integrity: American science and the creation of public interest organizations, 1955–1975’, American Journal of Sociology, 6, 1592–627. Samson, A. (1999), ‘Experts discuss action over deformed fish’, The Dominion, 23 April, 6. Samson, A. (2000), ‘Scientists seek active role in GE Inquiry’, The Dominion, 14 August, 8. Weaver, C.K. and J. Motion (2002), ‘Sabotage and subterfuge: public relations, democracy and genetic engineering in New Zealand’, Media, Culture & Society, 24, 325–43. Worrall, J. (2002), ‘GM man in the middle’, The Timaru Herald, 19 January.

10.

The knowledge economy: Fritz Machlup’s construction of a synthetic concept1 Benoît Godin

In 1962, the American economist Fritz Machlup published an influential study on the production and distribution of knowledge in the USA. Machlup’s study gave rise to a whole literature on the knowledge economy, its policies and its measurement. Today, the knowledge-based economy or society has become a buzzword in many writings and discourses, both academic and official. Where does Machlup’s concept of a knowledge economy come from? This chapter looks at the sources of Machlup’s insight. It discusses The Production and Distribution of Knowledge in the United States as a work that synthesizes ideas from four disciplines or fields of research – philosophy (epistemology), mathematics (cybernetics), economics (information) and national accounting – thus creating an object of study, or concept for science policy, science studies and the economics of science.

INTRODUCTION According to many authors, think tanks, governments and international organizations, we now live in a knowledge-based economy. Knowledge is reputed to be the basis for many if not all decisions, and an asset to individuals and firms. Certainly, the role of knowledge in the economy is not new, but knowledge is said to have gained increased importance in recent years, both quantitatively and qualitatively, partly because of information and communication technologies (see Foray, 2004). The knowledge-based economy (or society) is only one of many conceptual frameworks developed over the last 60 years to guide policies. In this sense, it competes for influence with other frameworks such as national innovation systems and triple helix. Certainly, all these conceptual frameworks carry knowledge as one dimension of analysis, but only 261

262


the knowledge-based economy has knowledge itself – its production and use – as its focus. The concept of a knowledge economy comes from Fritz Machlup. In 1962, the Austrian-born economist published a study that measured the production and distribution of (all kinds of) knowledge in the USA (Machlup, 1962a). The author estimated that, in 1958, the knowledge economy accounted for $136.4 million or 29 per cent of GNP. Machlup was the first to measure knowledge as a broad concept, while other measurements were concerned with the production of scientific knowledge, namely research and development (R&D), not its distribution. Machlup’s calculations gave rise to a whole literature on the knowledge economy, its policies and its measurement. The first wave, starting in the 1970s, was concerned with the so-called information economy. In fact, both information and knowledge as terms were used interchangeably in the literature. Using Machlup’s insights and the System of National Accounts as a source for data, Porat (1977) calculated that the information economy amounted to 46 per cent of GNP and 53 per cent of labour income in the USA in 1967. Porat’s study launched a series of similar analyses conducted in several countries and in the OECD (e.g. Godin, 2007). The second wave of studies on the knowledge economy started in the 1990s and continues today. The OECD, and Foray as consultant to the organization, relaunched the concept of a knowledge economy, with characteristics broadly similar to those Machlup identified (Godin, 2006a). This chapter is concerned with explaining where Machlup’s concept of the knowledge economy comes from. In fact, from the then-current literature, it can easily be seen that the term, as well as Machlup’s definition, was not entirely new. Philosophy was full of reflections on knowledge, and some economists were beginning to develop an interest in knowledge. Equally, Machlup’s method for measuring knowledge – accounting – already existed, namely in the fields of the economics of research and the economics of education. So, where does Machlup’s originality lie? The thesis of this chapter is that The Production and Distribution of Knowledge is a work of synthesis. First, the book is a synthesis of Machlup’s own work conducted before its publication. Second, and more importantly, the book is a synthesis of ideas from four disciplines or fields of research: philosophy (epistemology), mathematics (cybernetics), economics (information) and national accounting. Contrary to the work of some economists, this chapter takes Machlup’s work on knowledge seriously. Langlois, for example, has suggested that The Production and Distribution of Knowledge is ‘more a semantic exercise than an economic analysis . . . , categorizing and classifying, defining and refining, organizing and labeling’ (Langlois, 1985). Given the influence the book had on

Fritz Machlup’s construction of a synthetic concept

263

science studies (although not on the mainstream economic literature) and on policy discourses, we believe that this assertion offers a biased view of history. As we discuss, there are several methods for quantifying knowledge, and these are in competition. Machlup’s method was definitely not orthodox in mainstream economics, and Langlois’s judgement precisely illustrates this fact. The first section of this chapter discusses Machlup’s construction of the concept of knowledge and the sources of this construction. It looks at the definition of knowledge as both scientific and ordinary knowledge, and both production and distribution, and its ‘operationalization’ into four components: education, R&D, communication and information. Writings on epistemology, cybernetics and the economics of information are identified as the main intellectual inspiration for this construction. The second section analyses Machlup’s measurement of the knowledge economy based on a method of national accounting. This method is contrasted which economists’ most cherished method: growth accounting. The final section identifies the message or policy issues that Machlup associated with the knowledge economy.

MACHLUP’S CONSTRUCTION Fritz Machlup (1902–83), born in Austria, studied economics under Ludwig von Mises and Friedrich Hayek at the University of Vienna in the 1920s, and emigrated to the USA in 1933.2 His two main areas of work were industrial organization and monetary economics, but he also had a life-long interest in the methodology of economics and the ideal-typical role of assumptions in economic theory. Machlup’s work on the knowledge economy, a work of a methodological nature, grew out of five lectures he gave in 1959 and 1960. The rationale or motive Machlup offered for studying the economics of knowledge was the centrality of knowledge in society, despite the absence of theorizing in the economic literature. To Machlup, ‘knowledge has always played a part in economic analysis, or at least certain kinds of knowledge have . . . But to most economists and for most problems of economics the state of knowledge and its distribution in society are among the data assumed as given’ (Machlup, 1962a, pp. 3–4). To Machlup, ‘now, the growth of technical knowledge, and the growth of productivity that may result from it, are certainly important factors in the analysis of economic growth and other economic problems’ (ibid., p. 5). However, Machlup argued, there are other types of knowledge in addition to scientific knowledge. There is also knowledge of an ‘unproductive’ type for which society allocates ample resources: schools, books, radio

264


and television. Also, organizations rely more and more on ‘brain work’ of various sorts: ‘besides the researchers, designers, and planners, quite naturally, the executives, the secretaries, and all the transmitters of knowledge . . . come into focus’ (ibid., p. 7). To Machlup, these kinds of knowledge deserve study. Machlup (ibid., pp. 9–10) listed 11 reasons for studying the economics of knowledge, among them: 1. 2. 3. 4. 5. 6.

Knowledge’s increasing share of the nation’s budget. Knowledge’s social benefits, which exceed private benefits. Knowledge as strongly associated with increases in productivity and economic growth. Knowledge’s linkages to new information and communication technologies. Shifts of demand from physical labour to brain workers. Improving and adjusting national-income accounting.

Armed with such a rationale, Machlup suggested a definition of knowledge that had two characteristics. First, Machlup’s definition included all kinds of knowledge, scientific and ordinary knowledge: ‘we may designate as knowledge anything that is known by somebody’ (ibid., p. 7). Second, knowledge was defined as consisting of both its production and its distribution: ‘producing knowledge will mean, in this book, not only discovering, inventing, designing and planning, but also disseminating and communicating’ (ibid.). Defining Knowledge The first point about Machlup’s concept of knowledge was that it included all kinds of knowledge, not only scientific knowledge, but ordinary knowledge as well. Until then, most writings on knowledge were philosophical, and were of a positivistic nature: knowledge was ‘true’ knowledge (e.g. Ayer, 1956). As a consequence, the philosophy of practical or ordinary action ‘intellectualized’ human action. Action was defined as a matter of rationality and logic: actions start with deliberation, then intention, then decision (see Bernstein, 1971). Similarly, writings on decision-making were conducted under the assumption of strict rationality (rational choice theory) (e.g. Amadae, 2003). In 1949, the philosopher Gilbert Ryle criticized what he called the cultural primacy of intellectual work (Ryle, 1949). By this he meant understanding the primary activity of mind as theorizing, or knowledge of true propositions or facts. Such knowledge or theorizing Ryle called ‘knowing


265

that’. ‘Both philosophers and laymen tend to treat intellectual operations as the core of mental conduct (cognition) . . . The capacity for rigorous theory that lays the superiority of men over animals, of civilized men over barbarians and even of the divine mind over human mind . . . the capacity to attain knowledge of truth was the defining property of a mind’ (ibid., p. 26). To Ryle, there were other intellectual activities in addition to theorizing. To ‘knowing that’, Ryle added ‘knowing how’. Intelligence is ‘the ability, or inability, to do certain sorts of things’, the ability to ‘know how to perform tasks’ (ibid., pp. 27–8). These tasks are not preceded by intellectual theorizing. ‘Knowing how’ is a disposition, a skill, and is a matter of competence. To act intelligently is to apply rules, without necessarily theorizing about them first. To Ryle, the error comes from the old analytical separation of mind (mental) and body (physical): doing is not itself a mental operation, so performing ‘intelligent’ action must come from thinking. Ryle was one of the philosophers who were increasingly concerned with subjective knowledge.3 The chemist and philosopher Michael Polanyi drew a similar distinction to Ryle’s ten years later in Personal Knowledge, between what he called connoisseurship, as the art of knowing, and skills, as the art of doing (Polanyi, 1958). In this same book, Polanyi also brought forth the idea of inarticulate intelligence, or tacit knowledge, and this vocabulary became central to the modern conception of knowledge in science studies (see, e.g., Winter, 1987) – together with concepts such as learning-by-doing (e.g. Arrow, 1962a): ● ● ●

information (data; facts) versus knowledge (useful information); codified versus ‘uncodified’ knowledge (not generally available); tacit knowledge (individual and experiential).

Knowledge as subjective knowledge came to economics via the Austrian school of economists (see Knudsen, 2004), among them F.A. Hayek. In Hayek’s hands, the concept of knowledge was used as a criticism of the assumption of perfect information in economic theory. As is well known, information is a central concept of neoclassical economic theory: people have perfect information of the markets, and firms have perfect information of the technology of the time, or production opportunities. This is the familiar assumption of economic theory concerned with rational order, coordination and equilibrium, and its modern formulation owes its existence to John Hicks, Paul Samuelson and Gérard Debreu. As Hayek put it: ‘If we possess all the relevant information, if we can start out from a given system of preferences and if we command complete knowledge

266


of available means, the problem which remains is purely one of logic’ (Hayek, 1945, p. 519). But to Hayek, knowledge is never given for the whole society. Social knowledge is: dispersed bits of incomplete and frequently contradictory knowledge which all the separate individuals possess. The economic problem of society is thus not merely a problem of how to allocate given resources . . . It is rather a problem of how to secure the best use of resources known to any of the members of society . . . To put it briefly, it is a problem of the utilization of knowledge not given to anyone in its totality . . . Any approach, such as that of mathematical economics with its simultaneous equations, which in effect starts from the assumption that people’s knowledge corresponds with the objective facts of the situation, systematically leaves out what is our main task to explain (Ibid., pp. 519–20; 530).

To Hayek, as to Ryle, objective or ‘scientific knowledge is not the sum of all knowledge’ (ibid., p. 521). There are different kinds of knowledge: unorganized, particular, individual, practical, skill, and experience. In real life, no one has perfect information, but they have the capacity and skill to find information. This knowledge has nothing to do with a pure logic of choice, but is knowledge relevant to actions and plans. This kind of knowledge, unfortunately for mathematical economists, ‘cannot enter into statistics’: it is mostly subjective. ‘To what extent’, thus asked Hayek, ‘does formal economic analysis convey any knowledge about what happens in the real world?’ (Hayek, 1937, p. 33) To Hayek, economic equilibrium is not an (optimal) outcome (or state), but a process (activity) – the coordination of individuals’ plans and actions. In this process, individuals learn from experience and acquire knowledge about things, events and others that help them to act. In this sense, the system of prices plays the role of a signal; prices direct attention: ‘the whole reason for employing the price mechanism is to tell individuals that what they are doing, or can do, has for some reason for which they are not responsible become less or more demanded’ (Hayek, 1968 [1978], p. 187). ‘The price system is a mechanism for communicating information’ (Hayek, 1945, p. 526). Although perfect information, particularly information on prices, would continue to define economic orthodoxy in the 1960s (and after), more and more economists became interested in types of information, or knowledge different from strict rationality and prices,4 and in analysis of the economics of information itself (see Stigler, 1961, 1985; Boulding, 1966; Marschak, 1968, 1974; Lamberton, 1971; Arrow, 1973, 1974, 1979, 1984). The economics of science was one field where information took centre stage. From the start, the problem of science was defined in terms of decisions under uncertainty: how do you allocate resources to research,


267

where benefits are uncertain and long-term? Researchers from RAND (e.g. Hounshell, 1997), among them Arrow (1962b), then came to define scientific knowledge as information, with specific characteristics that made of it a public good: indivisibility, non-appropriability, uncertainty. As can be seen, an interest in studying information or knowledge differently was developing from various economic angles in the 1950s and early 1960s. The new developments shared a different understanding apart from strictly objective knowledge. This was also Machlup’s view. In line with Ryle and Hayek, Machlup argued for ‘subjective’ knowledge. To Machlup, knowledge ‘must not be limited by positivistic restrictions’ and need not be ‘true’ knowledge: ‘knowledge need not be knowledge of certified events and tested theories; it may be knowledge of statements and pronouncements, conjectures and hypotheses, no matter what their status of verification may be’ (Machlup, 1962a, p. 23). To Machlup, ‘all knowledge regardless of the strength of belief in it or warranty for it’ is knowledge (ibid.). After discussing existing classifications of knowledge and their limitations,5 Machlup identified five types of knowledge (1962a, pp. 22–3). His classification or definition of knowledge served as the basis for selecting activities and measuring the contribution of knowledge to the economy: ● ● ● ● ●

practical (professional, business, workers, political, households); intellectual; small-talk and pastime (entertainment and curiosity); spiritual; ‘unwanted’ (accidentally acquired).

‘Operationalizing’ Knowledge Defining knowledge as composed of all kinds of knowledge, scientific and ordinary, was the first aspect of Machlup’s definition of knowledge. The second was defining knowledge as both its production and its distribution. To Machlup, information is knowledge only if it is communicated and used. This theoretical insight allowed Machlup to ‘operationalize’ his concept of knowledge as being composed of four elements: education, R&D, communication and information. According to Machlup, the largest sector of the knowledge economy is concerned with distribution, and education itself is the largest part of this ‘industry’. To Machlup, education includes not just formal education in school, but also informal education. Eight categories or sources of education were identified: home (mothers educating children), school, training on the job (systematic and formal, excluding learning on the job), church, armed forces, television, self-education, and experience. Machlup

268


concentrated his analysis on the first six, in which knowledge is systematic or transmitted by a teacher, but was able to measure only the first four due to statistical difficulties. The second component of knowledge, the creation of knowledge or R&D, was what Machlup called the narrow sense of knowledge, as opposed to his wider definition, which included its distribution. To Machlup, R&D, commonly defined as the sum of basic research, applied R&D, was inappropriate.6 In lieu of the existing classification as used by the US National Science Foundation, for example, he offered a classification based on a four-stage ‘model’ – which culminated in innovation, a term Machlup explicitly preferred not to use: Basic research S Inventive work S Development S Plant construction Machlup was here taking stock of the new literature on the economics of innovation and its linear model (e.g. Godin, 2006b, 2008). To economists, innovation included more than R&D. Economists defined innovation as different from invention as studied by historians. Innovation was defined as the commercialization of invention by firms. To Machlup, adhering to such an understanding was part of his analytical move away from the primacy of scientific knowledge, or intellectual work. The third component of Machlup’s ‘operationalization’ of knowledge was media (of communication). Since all kinds of knowledge were relevant knowledge to Machlup, not only scientific knowledge but also ordinary knowledge, he considered a large range of vehicles for distribution: printing (books, periodicals, newspapers), photography and phonography, stage and cinema, broadcasting (radio and television), advertising and public relations, telephone, telegraph and postal service, and conventions. The final component of Machlup’s ‘operationalization’ was information, itself composed of two elements: information services and information machines (technologies). Information services, the eligibility for inclusion of which ‘may be questioned’ in a narrow concept of knowledge (Machlup, 1962a, p. 323), were: professional services (legal, engineering, accounting and auditing, and medical), finance, insurance and real estate, wholesale trade, and government. Information machines, of which he says ‘the recent development of the electronic-computer industry provides a story that must not be missed’ (ibid., p. 295), included signalling devices, instruments for measurement, observation and control, office information machines and electronic computers. Where does Machlup’s idea of defining knowledge as both production and distribution come from? Certainly, production and distribution have been key concepts of economics for centuries. However, the idea also


269

derives from the mathematical theory of communication, as developed independently by Claude Shannon and Norbert Wiener in the late 1940s (Shannon, 1948; Wiener, 1948; Shannon and Weaver, 1949). In the following decades, this theory became very popular in several disciplines, such as biology (e.g. Kay, 2000) and the social sciences (e.g. Heims, 1991). Economics (e.g. Mirowsky, 2002), and the economics of information (e.g. Marschak, 1959), would be no exception, and neither would science studies (e.g. Rogers and Shoemaker, 1970). The theory of communication defined information in terms of probability and entropy, and the content of information as resulting from the probability of this message being chosen among a number of alternative communication channels. Schematically, the theory portrayed information as a process involving three elements (see Weaver, 1949): Transmitter S Message S Receiver To Machlup, modern communication theory has given a description of the process between and within two persons or units in a system, one the transmitter, the other the receiver of the message. The transmitter selects the message from his information store, transmits it, usually after encoding it into a signal, through a communication channel to the receiver, who, after decoding the signal, puts the message into his information store. (Machlup, 1962a, p. 31)

What types of communicators are involved in this process? To Machlup, communicators, or knowledge-producers, as he suggested calling them, were of several types, according to the degree to which the messages delivered to a person differ from the messages he has previously received. He identified six types of knowledge-producers: 1. 2. 3.

4. 5.

transporter: delivers exactly what he has received, without changing it in the least (e.g. a messenger carrying a written communication); transformer: changes the form of the message received, but not its content (e.g. a stenographer); processor: changes both form and content of what he has received, by routine procedures like combinations or computations (e.g. an accountant); interpreter: changes form and content, but use imagination to create new form effects (e.g. a translator); analyser: uses so much judgement and intuition that the message that he communicates bears little or no resemblance to the message received;

270

6.


original creator: adds so much of his inventive genius and creative imagination that only weak and indirect connections can be found between what he has received and what he communicates.

To Machlup, knowledge covers the ‘entire spectrum of activities, from the transporter of knowledge up to the original creator’ (ibid., p. 33). His selection of industries for operationalizing knowledge’s activities illustrates this variety. He selected 30 specific groups of industries, or activities, shown in Table 10.1, covering the whole spectrum from creators to transporters. Despite his use of communication theory,7 Machlup did not retain the theory of communication’s key term – information. As he explained later, in a book he edited on information in 1984, information in cybernetics is either a metaphor (as in the case of machines) or has nothing to do with meaning but is a statistical probability of a sign or signal being selected (as in the case of transmission): ‘real information can come only from an informant. Information without an informant – without a person who tells something – is information in an only metaphoric sense’ (Machlup, 1983, p. 657). Machlup preferred to use the term knowledge. In fact, he refused to distinguish information (events or facts) from knowledge. To Machlup, knowledge has a double meaning: ‘both what we know and our state of knowing it’ (Machlup, 1962a, p. 13). The first is knowledge as state, or result, while the second meaning is knowledge as process, or activity. From an economic point of view, the second (transmission of knowledge) is as important as the first: ‘Knowledge – in the sense of what is known – is not really complete until it has been transmitted to some others’ (ibid., p. 14). This was Machlup’s rationale for using the term knowledge rather than information: ‘Information as that which is being communicated becomes identical with knowledge in the sense of that which is known’ (ibid., p. 15). Thus Machlup suggested that it is ‘more desirable to use, whenever possible, the word knowledge’, like the ordinary use of the word, where all information is knowledge (see ibid., p. 8).

MEASURING KNOWLEDGE When Machlup published The Production and Distribution of Knowledge, the economic analysis of science was just beginning (see Hounshell, 2000). A ‘breakthrough’ of the time was R.M. Solow’s paper on using the production function to estimate the role of science and technology in economic growth and productivity (Solow, 1957). The production function


271

Table 10.1 Education In the home On the job In the church In the armed forces Elementary and secondary Colleges and universities Commercial, vocational and residential Federal funds Public libraries R&D Basic research Applied research Printing and publishing Books and pamphlets Periodicals Newspapers Stationery and other office suppliers Commercial printing and lithography Photography and phonography Photography Phonography Stage, podium and screen Theatres and concerts Spectator sports Motion pictures Radio and television Advertising Telecommunication media Telephone Telegraph Postal service Conventions

Information machines Printing trades machines Musical instruments Motion picture apparatus and equipment Telephone and telegraph equipment Signaling devices Measuring and controlling instruments Typewriters Electronic computers Other office machines Office-machine parts Personal services Legal Engineering and architectural Accounting and auditing Medical Financial services Cheque-deposit banking Securities brokers Insurance agents Real estate agents Wholesale agents Miscellaneous business services Government Federal State and local

is an equation, or econometric ‘model’, that links the quantity produced of a good (output) to quantities of input. There are at any given time, or so economists argue, inputs (labour, capital) available to the firm, and a large variety of techniques by which these inputs can be combined to yield the desired (maximum) output. Using the production function, Solow formalized early works on growth accounting (decomposing GDP into

272


capital and labour), and equated the residual in his equation with technical change – although it included everything that was neither capital nor labour – as ‘a shorthand expression for any kind of shift in the production function’ (p. 312). Integrating science and technology was thus not a deliberate initiative, but it soon became a fruitful one. Solow estimated that nearly 90 per cent of growth was due to the residual. In the following years, researchers began adding variables to the equation in order to better isolate science and technology (e.g. Denison, 1962, 1967), or adjusting the input and capital factors to capture quality changes in output (e.g. Jorgenson and Griliches, 1967). According to Machlup, a mathematical exercise such as the production function was ‘only an abstract construction designed to characterize some quantitative relationships which are regarded as empirically relevant’ (Machlup, 1962b, p. 155). What the production function demonstrated was a correlation between input and output, rather than any causality: ‘a most extravagant increase in input might yield no invention whatsoever, and a reduction in inventive effort might by a fluke result in the output that had in vain been sought with great expense’ (ibid., p. 153). To Machlup, there were two schools of thought: According to the acceleration school, the more that is invented the easier it becomes to invent still more – every new invention furnishes a new idea for potential combination . . . According to the retardation school, the more that is invented, the harder it becomes to invent still more – there are limits to the improvement of technology. (Ibid., p. 156)

To Machlup, the first hypothesis was ‘probably more plausible’, but ‘an increase in opportunities to invent need not mean that inventions become easier to make; on the contrary, they become harder. In this case there would be a retardation of invention . . .’ (ibid., p. 162), because ‘it is possible for society to devote such large amounts of productive resources to the production of inventions that additional inputs will lead to less than proportional increases in output’ (ibid., p. 163). For measuring knowledge, Machlup chose another method than econometrics and the production function, namely national accounting. National accounting goes back to the eighteenth century and what was then called political arithmetic (see Deane, 1955, Buck, 1977, 1982; Cookson, 1983; Endres, 1985; Mykkanen, 1994; Hoppit, 1996). But national accounting really developed after World War II with the establishment of a standardized System of National Accounts, which allowed a national bureau of statistics to collect data on the production of economic goods and services in a country in a systematic way (see Studenski, 1958; Ruggles and Ruggles, 1970; Kendrick, 1970; Sauvy, 1970; Carson, 1975; Fourquet, 1980; Vanoli,


273

2002). Unfortunately for Machlup, knowledge was not – and is still not – a category of the National System of Accounts. There are, argued Machlup, ‘insurmountable obstacles in a statistical analysis of the knowledge industry’ (Machlup, 1962a, p. 44). Usually, in economic theory, ‘production implies that valuable input is allocated to the bringing forth of a valuable output’, but with knowledge there is no physical output, and knowledge is most of the time not sold on the market (ibid., p. 36). The need for statistically operational concepts forced Machlup to concentrate on costs, or national income accounting. To estimate costs8 and sales of knowledge products and services, Machlup collected numbers from diverse sources, both private and public. However, measuring costs meant that no data were available on the internal (non-marketed) production and use of knowledge, for example inside a firm: ‘all the people whose work consists of conferring, negotiating, planning, directing, reading, note-taking, writing, drawing, blueprinting, calculating, dictating, telephoning, card-punching, typing, multigraphing, recording, checking, and many others, are engaged in the production of knowledge’ (Machlup, 1962a, p. 41). Machlup thus looked at complementary data to capture the internal market for knowledge. He conducted work on occupational classes of the census, differentiating classes of white-collar workers who were knowledge-producing workers from those that were not, and computing the national income of these occupations (ibid., pp. 383 and 386). Machlup then arrived at his famous estimate: the knowledge economy was worth $136.4 million, or 29 per cent of GNP in 1958, had grown at a rate of 8.8 per cent per year over the period 1947–58, and occupied people representing 26.9 per cent of the national income (see Table 10.2). In conducting his accounting exercise, Machlup benefited from the experience of previous exercises conducted on education (e.g. Wiles, 1956) and human capital (e.g. Walsh, 1935; Mincer, 1958; Schultz, 1959, 1960, 1961a, 1961b, 1962; Becker, 1962; Hansen, 1963), and, above all, on research or R&D. The US National Science Foundation, as the producer Table 10.2

Education R&D Media of communication Information machines Information services

$ (millions)

%

60.194 10.990 38.369 8.922 17.961

44.1 8.1 28.1 6.5 13.2

274


of official statistics on science in the USA, started collecting data on R&D expenditures in the early 1950s (see Godin, 2005). Regular surveys were conducted on four economic sectors: government, universities, firms, and non-profit organizations. Then, in 1956, the Foundation published its ‘first systematic effort to obtain a systematic across-the-board picture’ (NSF, 1956). It consisted of the sum of the results of the sectoral surveys for estimating national funds for R&D. The National Science Foundation calculated that the national budget for R&D amounted to $5.4 billion in 1953. From the start, the data on R&D from the National Science Foundation were inserted into the System of National Accounts’ framework as a model: surveys were conducted according to economic sectors, the classifications used corresponded to available classifications, the matrix of R&D money flows imitated the input–output tables accompanying the System of National Accounts, and a ratio R&D to GNP was constructed. To the National Science Foundation, such an alignment with the System of National Accounts was its way to relate R&D to economic output statistically: describing ‘the manner in which R&D expenditures enter the gross national product in order to assist in establishing a basis for valid measures of the relationships of such expenditures to aggregate economic output’ (NSF, 1961, p. i). Machlup made wide use of the National Science Foundation’s data for his own accounting. As Nelson once stated: ‘the National Science Foundation has been very important in focusing the attention of economists on R&D (organized inventive activity), and the statistical series the NSF has collected and published have given social scientists something to work with’ (Nelson, 1962, p. 4). The organization’s numbers were one of many sources Machlup added together in calculating his estimate of the size of the knowledge economy. In fact, for most of his calculations, Machlup did not use the System of National Accounts, as Porat would for his work on the information economy. Instead he looked liberally at the literature for available numbers, like the National Science Foundation data, and conducted many different calculations (summations, mathematical projections, estimations and computations of opportunity costs). Neither was Machlup addicted to accounting. Although he chose costs for his estimate of the knowledge economy, he discussed and suggested many other statistics. For media of communication, he looked at the number of books and periodicals, their circulation and content; for information, he collected figures on types of technology, and use of technologies in households; on education, he recommended using figures for attendance, years of schooling, achievement tests, class hours, amount of homework, and subject-matter requirements; for R&D, he proposed a list of measures


275

on input and output (see Appendix 1), and relationships or ratios between the two9. Machlup was realistic about his own accounting, qualifying some of his estimates as being speculative (Machlup, 1962a, p. 62), that is, ideas of magnitude and trends based on conjecture rather than exact figures (ibid., p. 103), and he qualified some of his comparisons ‘with several grains of salt’ (ibid., p. 374). To Machlup, it was the message rather than the statistical adequacy that was important. The very last sentence of the book reads as follows: ‘concern about their accuracy [statistical tables] should not crowd out the message it conveys’ (ibid., p. 400).

THE MESSAGE Apart from theoretical borrowings from philosophy, mathematics, economics and national accounting, we can identify policy issues and even professional interests in Machlup’s analysis at several levels. First, Machlup was concerned with the challenges facing the education and research system of which he was part. Second, he was concerned, as analyst, with the new information technology ‘revolution’. For each of the four components operationalizing his definition of knowledge, Machlup identified policy issues, and this partly explains the inclusion of the components in the operationalization. The policy issues Machlup identified were mainly economic. To begin with education, the central question discussed was productivity. Machlup distinguished productivity in education (or performance) from productivity of education (or simply productivity). With regard to productivity in education, Machlup suggested compressing the curriculum to accelerate the production of well-trained brainpower and therefore economic growth. We need an educational system that will significantly raise the intellectual capacity of our people. There is at present a great scarcity of brainpower in our labor force. . .. Unless our labor force changes its composition so as to include a much higher share of well-trained brainpower, the economic growth of the nation will be stunted and even more serious problems of employability will arise. (Machlup, 1962a, p. 135)

Concerning the productivity of education, he suggested considering (and measuring) education as an investment rather than as a cost, and as an investment not only in the individual (earnings) but also in society (culture), in line with studies on social rates of return on research (e.g. Shultz, 1953; Griliches, 1958). As to the second component – R&D – Machlup confessed that ‘this

276


subject was his first interest in the field of knowledge production. The temptation to expand the area of study to cover the entire industry came later, and proved irresistible’ (Machlup, 1962a, p. 48). To Machlup, the policy issues involving R&D were twofold. One was the decline of inventions. From the early 1950s, Machlup had studied monopolies and the role of patents in competition (Machlup, 1952), and particularly the role of the patent system in inducing invention (e.g. Machlup and Penrose, 1950; Machlup, 1958b). Following several authors, among them J. Schmookler (e.g. Schmookler, 1954), he calculated a decline in patenting activity after 1920 (Machlup, 1961). He wondered whether this was due to the patent system itself, or to other factors. In the absence of empirical evidence, he suggested that ‘faith alone, not evidence, supports’ the patent system. To Machlup, it seems ‘not very likely that the patent system makes much difference regarding R&D expenditures of large firms’ (Machlup, 1962a, p. 170). A second policy issue concerning R&D was the productivity of research, and his concern with this issue grew out of previous reflections on the allocation of resources to research activities and the inelasticity in the shortterm supply of scientists and engineers (Machlup, 1958a). To Machlup, research, particularly basic research, is an investment, not a cost. Research leads to an increase in economic output and productivity (goods and services), and society gains from investing in basic research with public funds: the social rate of return is higher than private ones (see Griliches, 1958), and ‘the nation has probably no other field of investment that yields return of this order’ (Machlup, 1962a, p. 189). But there actually was a preference for applied research in America, claimed Machlup: ‘American preference for practical knowledge over theoretical, abstract knowledge is a very old story’ (ibid., pp. 201–2). That there was a ‘disproportionate support of applied work’ (ibid., pp. 203) was a popular thesis of the time among scientists (see Reingold, 1971). To Machlup, there was a social cost to this: echoing V. Bush, according to whom ‘applied research invariably drives out pure’ research (Bush, 1945 [1995], p. xxvi), Machlup argued that industry picks up potential scientists before they have completed their studies, and dries up the supply of research personnel (shortages). Furthermore, if investments in basic research remain too low (8 per cent of total expenditures on R&D), applied research will suffer in the long run, since it depends entirely on basic research. Such was the rhetoric of the scientific community’s members at the time. These were the main policy issues Machlup discussed. Concerning the last two components of his definition – communication and information – Machlup was very brief. In fact, his policy concern was mainly with information technologies and the technological revolution. To Machlup,


277

the important issue here was twofold. The first part was rational decisionmaking: the effects of information machines are ‘improved records, improved decision-making, and improved process controls . . . that permit economies’ (Machlup, 1962a, p. 321). Machlup was here offering what would become the main line of argument for the information economy in the 1980s and after: information technologies as a source of growth and productivity. The second part was the issue of structural change and unemployment (‘replacement of men by machines’). Structural change was a concern for many in the 1940s and 1950s, and the economist Wassily Leontief devoted numerous efforts to measuring it using input–output tables and accounting as a framework (e.g. Leontief, 1936, 1953, 1986; see also Leontief, 1952, 1985; Leontief and Duchin, 1986). ‘There has been’, stated Machlup, ‘a succession of occupations leading [the movement to a knowledge economy], first clerical, then administrative and managerial, and now professional and technical personnel . . . , a continuing movement from manual to mental, and from less to more highly trained labor’ (Machlup, 1962a, pp. 396–7). To Machlup, ‘technological progress has been such as to favor the employment of knowledge-producing workers’ (ibid., p. 396), but there was the danger of increasing unemployment among unskilled manual labour (ibid., p. 397). In the long run, however, ‘the demand for more information may partially offset the labor-replacing effect of the computer-machine’ (ibid., p. 321). With regard to communication, the fourth component of his ‘operationalization’ of knowledge, Machlup discussed no specific policy issue. But there was one in the background, namely the information explosion (see Godin, 2007). In the 1950s, the management of scientific and technical literature emerged as a concern to many scientists and universities, and increasingly to governments. According to several authors, among them science historian Derek J. de Solla Price, scientific and technical information, as measured by counting journals and papers, was growing exponentially. Science was ‘near a crisis’, claimed Price, because of the proliferation and superabundance of literature (Price, 1956). Some radically new technique must be evolved if publication is to continue as a useful contribution (Price, 1961). The issue gave rise to scientific and technical information policies starting from the early 1960s, as a precursor to policies on the information economy and, later, on information technology (see Godin, 2007). In 1962, Machlup did not discuss the issue of information explosion. He even thought that counting the number of books was a ‘very misleading index of knowledge’ (Machlup, 1962a, p. 122). However, in the 1970s, he conducted a study on ‘The production and distribution of scientific and technological information’, published in four volumes as Information

278


through the Printed World (Machlup and Lesson, 1978–80). Produced for the National Science Foundation, the study looked at books, journals, libraries, and their information services from a quantitative point of view, as had been done in The Production and Distribution of Knowledge: the structure of the industries, markets, sales, prices, revenues, costs, collections, circulation, evaluation and use. Machlup wrote on knowledge at a time when science, or scientific knowledge, was increasingly believed to be of central importance to society – and scientists benefited largely from public investments in research. Economists, according to whom ‘if society devotes considerable amounts of its resources to any particular activity, will want to look into this allocation and get an idea of the magnitude of the activity, its major breakdown, and its relation to other activities’ (Machlup, 1962a, p. 7), started measuring the new phenomenon, and were increasingly solicited by governments to demonstrate empirically the contribution of science to society – cost control on research expenditures was not yet in sight. Machlup was part of this ‘movement’, with his own intellectual contribution.

CONCLUSION Machlup’s study on the knowledge economy accomplished three tasks. It defined knowledge, measured it, and identified policy issues. The message was that knowledge was an important component of the economy, but does not completely respond to an economic logic. With The Production and Distribution of Knowledge, Machlup brought the concept of knowledge into science policies and science studies. His conception of knowledge was synthesized from three intellectual trends of the time: ‘disintellectualizing’ and ‘subjectivizing’ knowledge (ordinary knowledge); looking at knowledge as a communication process (production and distribution); and measuring its contribution to the economy (in terms of accounting). In the early 1980s, Machlup began updating his study on the knowledge economy with a projected ten-volume series entitled Knowledge: its Creation, Distribution, and Economic Significance (Machlup, 1980–84). He died after finishing the third volume. By then, he was only one of many measuring the knowledge or information economy. With this new project, Machlup kept to his original method as developed in 1962: national accounting. This was a deliberate choice. In fact, there were two types of accounting measurement in the economic literature of the time. One was growth accounting. It used econometrics, and was the cherished method among quantitative economists. With the aid of equations and


279

statistical correlations, economists tried to measure the role of knowledge in economic growth, following in Solow’s footsteps. Machlup did not believe in this method. The second method was national accounting. This method was not very attractive to economists – although developed by one of them (Simon Kuznets). It relied on descriptive statistics rather than formalization. Its bad reputation, and the reluctance of economists to use national accounting, have a long tradition, going back to the arguments of eighteenth-century classical economists against political arithmetic (see Johannisson, 1990). It was such a reluctance that economist R.R. Nelson expressed while reviewing Machlup’s book in Science in 1963. Nelson expressed his disappointment that Machlup had not studied the role and function of knowledge: ‘Machlup is concerned principally with identifying and quantifying the inputs and outputs of the knowledge-producing parts of the economy and only secondarily with analyzing the function of knowledge and information in the economic system’ (Nelson, 1963, pp. 473–4).

MACHLUP’S SOURCES OF INSIGHT (see Figure 10.1) Today, the measurement of knowledge is often of a third kind. Certainly, knowledge is still, most of the time, defined as Machlup suggested (creation and use) – although the term has also become a buzzword for any writing and discourse on science, technology and education. But in the official literature, knowledge is actually measured using indicators. Such Field

Concept

Philosophy

Subjective knowledge

Machlup

(Ryle, Polanyi) KNOWLEDGE Economics

Information

(Hayek, Arrow)

Mathematics

Communication

(Shannon and Wiener)

Statistics

Accounting

(NSF, human capital)

Figure 10.1

COMMUNICATION INFORMATION

Machlup’s sources of insight

EDUCATION R&D

280


measurements are to be found in publications from the OECD and the European Union, for example. Here, knowledge is measured using a series or list of indicators gathered under the umbrella term ‘knowledge’ (see Godin, 2006a). There is no summation (or composite value), as in accounting, but a collection of available statistics on several dimensions of knowledge, that is, science and technology, among them those on information technologies (see Appendix 2). The methodology of indicators for measuring knowledge, information or simply science comes partly from Machlup. We have seen how Machlup complemented his accounting exercise with discussions on various sorts of statistics, among them statistics on R&D organized into an input–output framework. In 1965, the British economist Christopher Freeman, as consultant to the OECD, would suggest such a collection of indicators to the organization (Freeman and Young, 1965). In the 1970s, the National Science Foundation initiated such a series, entitled Science Indicators, which collected multiple statistics for measuring science and technology. To statistics on input, among them money devoted to R&D, the organization added statistics on output such as papers, citations, patents, high-technology products and so on. The rationale behind the collection of indicators was precisely that identified by Machlup as a policy issue: the ‘productiveness’, or efficiency of the research system (National Science Board, 1973, p. iii). Since then, the literature and measurement on knowledge has growth exponentially, and knowledge is now a central concept of many conceptual frameworks like the triple helix. While Machlup has been influential on many aspects of the analysis of knowledge, among them definition and measurement, current measurements of knowledge are still restricted to scientific knowledge and information technology. Certainly, many aspects of knowledge remain non-accountable, as they were in the 1960s, but the economic orientation of policies and official statistics (economic growth and productivity) probably explains much of this orientation, to which Machlup has contributed.

NOTES 1. The author thanks Michel Menou for comments on a preliminary draft of this chapter. 2. He taught at the University of Buffalo (1935–47), then Johns Hopkins (1947–60), then Princeton (1960–71). After retiring in 1971, he joined New York University until his death. 3. At about the same time, B. Russell distinguished between what he called social and individual knowledge, the first concerned with learned knowledge, the other with experience. See Russell (1948); see also Schutz (1962) and Schutz and Luickmann (1973).


281

4. Knowledge of others’ behaviour (strategic), knowledge of institutions and rules, bounded rationality. 5. Basic versus applied (difficulties in separating the two), scientific versus historical (focusing largely on school learning), enduring versus transitory (the latter nevertheless has great economic value), instrumental versus intellectual versus spiritual (no place for knowledge of transitory value). 6. Other distinctions he discussed were: discovery versus invention (W.C. Kneale), major versus minor inventions (S.C. Gilfillan and W.F. Ogburn). 7. Machlup quoted Weaver at two points in his book. 8. Machlup preferred the concept of investments in the case of education and R&D. 9. For an in-depth discussion of Machlup on this topic, see Machlup (1960).

REFERENCES Amadae, S.M. (2003), Rationalizing Capitalist Democracy: The Cold War Origins of Rational Choice Liberalism, Chicago, IL: University of Chicago Press. Arrow, K.J. (1962a), ‘The economic implication of learning-by-doing’, Review of Economic Studies, 29, 155–73. Arrow, K.J. (1962b), ‘Economic welfare and the allocation of resources for invention’, in National Bureau of Economic Research, The Rate and Direction of Inventive Activity, Princeton, NJ: Princeton University Press, pp. 609–25. Arrow, K.J. (1973), ‘Information and economic behavior’, lecture given at the 1972 Nobel Prize Celebration, Stockholm: Federation of Swedish Industries. Arrow, K.J. (1974), ‘Limited knowledge and economic analysis’, American Economic Review, 64, 1–10. Arrow, K.J. (1979), ‘The economics of information’, in M.L. Deltouzos and J. Moses (eds), The Computer Age: A Twenty-Year View, Cambridge, MA: MIT Press, pp. 306–17. Arrow, K.J. (1984), The Economics of Information, Cambridge, MA: Harvard University Press. Ayer, A.J. (1956), The Problem of Knowledge, Harmondsworth: Penguin Books. Becker, G.S. (1962), ‘Investment in human capital: a theoretical analysis’, Journal of Political Economy, 70 (5), 9–49. Bernstein, R.J. (1971), Praxis and Action, Philadelphia, PA: University of Pennsylvania Press. Boulding, K.E. (1966), ‘The economics of knowledge and the knowledge of economics’, American Economic Review, 56 (1–2), 1–13. Buck, P. (1977), ‘Seventeenth-century political arithmetic: civil strife and vital statistics’, ISIS, 68 (241), 67–84. Buck, P. (1982), ‘People who counted: political arithmetic in the 18th century’, ISIS, 73 (266), 28–45. Bush, V. (1945) [1995], Science: The Endless Frontier, North Stratford, NH: Ayer Company Publishers. Carson, C.S. (1975), ‘The history of the United States National Income and Product Accounts: the development of an analytical tool’, Review of Income and Wealth, 21 (2), 153–81. Cookson, J.E. (1983), ‘Political arithmetic and war in Britain, 1793–1815’, War and Society, 1, 37–60. Deane, P. (1955), ‘The implications of early national income estimates for the

282


measurement of long-term economic growth in the United Kingdom’, Economic Development and Cultural Change, 4 (1), Part I, 3–38. Denison, E.F. (1962), The Sources of Economic Growth in the United States and the Alternatives Before Us, Committee for Economic Development, New York. Denison, E.F. (1967), Why Growth Rates Differ, Washington, DC: Brookings Institution. Endres, A.M. (1985), ‘The functions of numerical data in the writings of Graunt, Petty, and Davenant’, History of Political Economy, 17 (2), 245–64. Foray, D. (2004), Economics of Knowledge, Cambridge, MA: MIT Press. Fourquet, F. (1980), Les comptes de la puissance, Paris: Encres. Freeman, C. and A. Young (1965), The Research and Development Effort in Western Europe, North America and the Soviet Union: An Experimental International Comparison of Research Expenditures and Manpower in 1962, Paris: OECD. Godin, B. (2005), Measurement and Statistics on Science and Technology: 1920 to the Present, London: Routledge. Godin, B. (2006a), ‘The knowledge-based economy: conceptual framework or buzzword?’, Journal of Technology Transfer, 31, 17–30. Godin, B. (2006b), ‘The linear model of innovation: the historical construction of an analytical framework, science’, Technology and Human Values, 31 (6), 1–29. Godin, B. (2007), ‘The information economy: the history of a concept through its measurement, 1949–2005’, History and Technology, 24 (3), 255–87. Godin, B. (2008), In the Shadow of Schumpeter: W. Rupert Maclaurin and the Study of Technological Innovation, Minerva, 46 (3), 343–60. Griliches, Z. (1958), ‘Research costs and social returns: hybrid corn and related innovation’, Journal of Political Economy, 46, 419–31. Hansen, W.L. (1963), ‘Total and private rates of return to investment in schooling’, Journal of Political Economy, 71, 128–40. Hayek, F.A. (1937), ‘Economics and knowledge’, Economica, 4, 33–54. Hayek, F.A. (1945), ‘The use of knowledge in society’, American Economic Review, 35 (4), 519–30. Hayek, F.A. (1968) [1978], ‘Competition as a discovery procedure’, in New Studies in Philosophy, Economics and the History of Ideas, London: Routledge, pp. 179–90. Heims, S.J. (1991), Constructing a Social Science for Postwar America: The Cybernetics Group, 1946–1953, Cambridge, MA: MIT Press. Hoppit, J. (1996), ‘Political arithmetic in 18th century England’, Economic History Review, 49 (3), 516–40. Hounshell, D.A. (1997), ‘The Cold War, RAND, and the generation of knowledge, 1946–1962’, Historical Studies in the Physical and Biological Sciences, 27 (2), 237–67. Hounshell, D.A. (2000), ‘The medium is the message, or how context matters: The RAND Corporation builds an economics of innovation, 1946–1962’, in A.C. Hughes and H.P. Hughes (eds), Systems, Experts, and Computers, Cambridge, MA: MIT Press, pp. 255–310. Johannisson, K. (1990), ‘Society in numbers: the debate over quantification in 18th century political economy’, in T. Frangsmyr et al. (eds), The Quantifying Spirit in the Eighteenth Century, Berkeley, CA: University of California Press, pp. 343–61. Jorgenson, D.W. and Z. Griliches (1967), ‘The explanation of productivity change’, Review of Economic Studies, 34 (3), 249–83.


283

Kay, L.E. (2000), ‘How a genetic code became an information system’, in A.C. Hughes and H.P. Hughes (eds), Systems, Experts, and Computers, Cambridge, MA: MIT Press, pp. 463–91. Kendrick, J.W. (1970), ‘The historical development of national-income accounts’, History of Political Economy, 2 (1), 284–315. Knudsen, C. (2004), ‘Alfred Schutz, Austrian economists and the knowledge problem’, Rationality and Society, 16 (1), 45–89. Lamberton, D. (1971), Economics of Information and Knowledge, Harmondsworth: Penguin. Langlois, R.N. (1985), ‘From the knowledge of economics to the economics of knowledge: Fritz Machlup on methodology and on the knowledge society’, Research in the History of Economic Thought and Methodology, 3, 225–35. Leontief, W. (1936), ‘Quantitative input and output relations in the economic systems of the United States’, Review of Economic Statistics, 18 (3), 105–25. Leontief, W. (1952), ‘Machines and man’, Scientific American, 187, 150–60. Leontief, W. (ed.) (1953), Studies in the Structure of the American Economy, New York: Oxford University Press. Leontief, W. (1985), ‘The choice of technology’, Scientific American, 252, 37–45. Leontief, W. (1986), Input–Output Economics, Oxford: Oxford University Press. Leontief, W. and F. Duchin (1986), The Future Impact of Automation on Workers, Oxford: Oxford University Press. Machlup, F. (1952), The Political Economy of Monopoly: Business, Labor and Government Policies, Baltimore, MD: Johns Hopkins Press. Machlup, F. (1958a), ‘Can there be too much research?’, Science, 128 (3335), 1320–25. Machlup, F. (1958b), An Economic Review of the Patent System, Study no. 15, Committee on the Judiciary, 85th Congress, Second Session, Washington, DC. Machlup, F. (1960), ‘The supply of inventors and inventions’, Weltwirtschaftliches Archiv, 85, 210–54. Machlup, F. (1961), ‘Patents and inventive efforts’, Science, 133 (3463), 1463–6. Machlup, F. (1962a), The Production and Distribution of Knowledge in the United States, Princeton, NJ: Princeton University Press. Machlup, F. (1962b), ‘The supply of inventors and inventions’, in National Bureau of Economic Research, The Rate and Direction of Inventive Activity, Princeton, NJ: Princeton University Press, pp. 143–69. Machlup, F. (1980–84), Knowledge: Its Creation, Distribution, and Economic Significance, Princeton, NJ: Princeton University Press. Machlup, F. (1983), ‘Semantic quirks in studies of information’, in F. Machlup and U. Mansfield (eds), The Study of Information: Interdisciplinary Messages, New York: John Wiley, p. 657. Machlup, F. and K. Lesson (1978–80), Information through the Printed World: The Dissemination of Scholarly, Scientific, and Intellectual Knowledge, New York: Praeger. Machlup, F. and E. Penrose (1950), ‘The patent controversy in the nineteenth century’, Journal of Economic History, 1 (1), 1–29. Marschak, J. (1959), ‘Remarks on the economics of information’, in Contributions to Scientific Research in Management, Western Data Processing Center, Los Angeles: University of California, pp. 79–98. Marschak, J. (1968), ‘Economics of inquiring, communicating, deciding’, American Economic Review, 58 (2), 1–18.

284


Marschak, J. (1974), Economic Information, Decision and Prediction, Dordrecht: Reidel. Mincer, J. (1958), ‘Investment in human capital and personal income distribution’, Journal of Political Economy, 66 (4), 281-302. Mirowsky, P. (2002), Machine Dreams: Economics Becomes a Cyborg Science, Cambridge: Cambridge University Press. Mykkanen, J. (1994), ‘To methodize and regulate them: William Petty’s governmental science of statistics’, History of the Human Sciences, 7 (3), 65–88. OECD (2001), STI Scoreboard: Towards a Knowledge-Based Economy, Paris: OECD. National Science Board (1973), Science Indicators 1972, Washington, DC: NSF. Nelson, R.R. (1962), ‘Introduction’, in National Bureau of Economic Research, The Rate and Direction of Inventive Activity, Princeton, NJ: Princeton University Press, p. 4. Nelson, R.R. (1963), ‘Role of knowledge in economic growth’, Science, 140 (3566), 473–4. NSF (1956), ‘Expenditures for R&D in the United States: 1953’, Reviews of Data on R&D, 1, NSF 56-28, Washington, DC. NSF (1961), ‘R&D and the GNP’, Reviews of Data on R&D, 26, NSF 61-9, Washington, DC, p. 1. Polanyi, M. (1958), Personal Knowledge: Towards a Post-Critical Philosophy, Chicago, IL: University of Chicago Press. Porat, M.U. (1977), The Information Economy, 9 vols, Office of Telecommunication, US Department of Commerce, Washington, DC. Price, D.D.S. (1956), ‘The exponential curse of science’, Discovery, 17, 240–43. Price, D.D.S. (1961), Science since Babylon, New Haven, CT: Yale University Press. Reingold, N. [1971], ‘American indifference to basic research: a reappraisal’, in N. Reingold, Science: American Style, New Brunswick, NJ and London: Rutgers University Press, 1991, pp. 54–75. Rogers, E.M. and F.F. Shoemaker (1970), Communication of Innovation: a CrossCultural Approach, New York: Free Press. Ruggles, N. and R. Ruggles (1970), The Design of Economic Accounts, National Bureau of Economic Research, New York: Columbia University Press. Russell, B. (1948), Human Knowledge: Its Scope and Limits, New York: Simon and Schuster. Ryle, G. (1949), The Concept of Mind, Chicago, IL: University of Chicago Press. Sauvy, A. (1970), ‘Histoire de la comptabilité nationale’, Économie et Statistique, 14, 19–32. Schmookler, J. (1954), ‘The level of inventive activity’, Review of Economics and Statistics, 36 (2), 183–90. Schultz, T.W. (1953), The Economic Organization of Agriculture, New York: McGraw-Hill. Schultz, T.W. (1959), ‘Investment in man: an economist’s view’, Social Service Review, 33 (2), 109–17. Schultz, T.W. (1960), ‘Capital formation by education’, Journal of Political Economy, 68 (6), 571–83. Schultz, T.W. (1961a), ‘Investment in human capital’, American Economic Review, 51 (1), 1–17. Schultz, T.W. (1961b), ‘Education and economic growth’, in N.B. Henry (ed.),


285

Social Forces Influencing American Education, Chicago, IL: University of Chicago Press, pp. 46–88. Schultz, T.W. (1962), ‘Reflections on investment in man’, Journal of Political Economy, 70 (5), 1–8. Schutz, A. (1962), Collected Papers I: the Structure of Social Reality, Dordrecht: Kluwer. Schutz, A. and T. Luickmann (1973), The Structures of the Life-World, Evanston, IL: Northwestern University Press. Shannon, C.E. (1948), ‘A mathematical theory of communication’, in N.J.A. Sloane and A.D. Wyner (eds), C. E. Shannon, Collected Papers, New York: IEEE Press, pp. 5–83. Shannon, C.E. and W. Weaver (1949), Mathematical Theory of Communication, Urbana, IL: University of Illinois Press. Solow, R.M. (1957), ‘Technical change and the aggregate production function’, Review of Economics and Statistics, 39, 312–20. Stigler, G.J. (1961), ‘The economics of information’, Journal of Political Economy, LXIX (3), 213–25. Stigler, J.E. (1985), ‘Information and economic analysis: a perspective’, Economic Journal, 95, 21–41. Studenski, P. (1958), The Income of Nations: Theory, Measurement, and Analysis, Past and Present, New York: New York University Press. Vanoli, A. (2002), Une histoire de la comptabilité nationale, Paris: La Découverte. Walsh, J.R. (1935), ‘Capital concept applied to man’, Quarterly Journal of Economics, 49 (2), 255–85. Weaver, W. (1949), ‘The mathematics of communication’, Scientific American, 181 (1), 11–15. Wiener, N. (1948), Cybernetics: Or Control and Communication in the Animal and the Machine, Cambridge, MA: MIT Press. Wiles, J.D. (1956), ‘The nation’s intellectual investment’, Bulletin of the Oxford University Institute of Statistics, 18 (3), 279–90. Winter, S.G. (1987), ‘Knowledge and competence as strategic assets’, in D.J. Teece (ed.), The Competitive Challenge, Cambridge, MA: Ballinger, pp. 159–84.

286

II ‘Inventive Work’ (including minor improvements but excluding further development of inventions) Laboratories Materials, fuel, power

2. Scientific problems and hunches (old stock and output

∂

Outlays, current and deflated Outlay per man

and deflated

∂ Payrolls, current

Men, man-hours

Scientists Non-scientist inventors Engineers Technical aides Clerical aides

A. Raw inventions, technological recipes a. Patented inventions b. Patentable inventions, not patented but published

C. New practical problems and ideas

1. Scientific knowledge (old stock and output from I-A)

∂

B. New scientific problems and hunches


Laboratories Materials, fuel, power

2. Scientific problems and hunches (old stock and output from I-B, II-B and III-B)

a. Patent applications and patents b. Technological papers and memoranda

—

—

Research papers and memoranda: formulas

A. New scientific knowledge: hypotheses and theories

Men, man-hours Payrolls, current and deflated

Scientists Technical aides Clerical aides

1. Scientific knowledge (old stock and output from I-A) ∂

Measurable

Intangible

Measurable

Tangible

Output

Intangible

Input

The flow of ideas through the stages of research, invention and development to application

I ‘Basic research’ [Intended output ‘formulas’]

Stage

Table 10A.1

APPENDIX 1

287

III ‘Development work’ [Intended output: ‘Blueprints and specifications’]

[Intended output: ‘Sketches’]

Laboratories Materials, fuel, power

2. Technology (old stock and output from III-A) Pilot plants

Scientists Engineers Technical aides Clerical aides

1. Scientific knowledge (old stock and output from I-A)

3. Practical problems and ideas (old stock and output from I-C, II-C, III-C and IV-A)

from II-A and III-A)

∂

∂

Investment


Payrolls, current and deflated

Men, man-hours


A. Developed inventions: blueprints, specification, samples



c. Patentable inventions, neither patented nor published d. Non-patentable inventions, published e. Non-patentable inventions, not published f. Minor improvements

—

Blueprints and specifications

—

—

f. —

e. —

d. Papers and memoranda

c. —

288

Source:

Machlup (1962), pp. 180–81.

4. Enterprise (venturing)

3. Financial resources

2. Business acumen and market forecasts

1. Developed inventions (output from III-A)

4. Raw inventions and improvements (old stock and output from II-A)

3. Practical problems and ideas (old stock and output from I-C, II-C, III-C and IV-A)

Intangible

(continued)

IV ‘New-type plant construction’ [Intended output: ‘New-type plant’]

Stage

Table 10A.1

Building materials Machines and tools

Entrepreneurs Managers Financiers and bankers Builders and contractors Engineers

Tangible

Input

$ investments in new-type plant

Measurable

New-type plant producing a. novel products b. better products c. cheaper products

—


A. New practical problems and ideas

Measurable

Intangible

Output


APPENDIX 2

289

INDICATORS ON THE KNOWLEDGEBASED ECONOMY

A. Creation and Diffusion of Knowledge Investments in knowledge Domestic R&D expenditure R&D financing and performance Business R&D R&D in selected ICT industries and ICT patents Business R&D by size classes of firms Collaborative efforts between business and the public sector R&D performed by the higher education and government sectors Public funding of biotechnology R&D and biotechnology patents Environmental R&D in the government budget Health-related R&D Basic research Defence R&D in government budgets Tax treatment of R&D Venture capital Human resources Human resources in science and technology Researchers International mobility of human capital International mobility of students Innovation expenditure and output Patent applications Patent families Scientific publications B. Information Economy Investment in information and communication technologies (ICT) Information and communication technology (ICT) expenditures Occupations and skills in the information economy Infrastructure for the information economy Internet infrastructure Internet use and hours spent on-line Access to and use of the Internet by households and individuals Internet access by enterprise size and industry Internet and electronic commerce transactions Price of Internet access and use

290


Size and growth of the ICT sector Contribution of the ICT sector to employment growth Contribution of the ICT sector to international trade Cross-border mergers, acquisitions and alliances in the ICT sector C. Global Integration of Economic Activity International trade Exposure to international trade competition by industry Foreign direct investment flows Cross-border mergers and acquisitions Activity of foreign affiliates in manufacturing Activity of foreign affiliates in services Internationalization of industrial R&D International strategic alliances between firms Cross-border ownership of inventions International co-operation in science and technology Technology balance of payments D. Economic Structure and Productivity Differences in income and productivity Income and productivity levels Recent changes in productivity growth Labour productivity by industry Technology and knowledge-intensive industries Structure of OECD economies International trade by technology intensity International trade in high and medium-high technology industries Comparative advantage by technology intensity Source:

OECD (2001).

11.

Measuring the knowledge base of an economy in terms of triple-helix relations Loet Leydesdorff, Wilfred Dolfsma and Gerben Van der Panne

Ever since evolutionary economists introduced the concept of a ‘knowledgebased economy’ (Foray and Lundvall, 1996), the question of the measurement of this new type of economic coordination has come to the fore (Carter, 1996; OECD, 1996). Godin (2006) argued that the concept of a knowledge-based economy has remained a rhetorical device because the development of specific indicators has failed. However, the ‘knowledgebased economy’ has been attractive to policy-makers. For example, the European Summit of March 2000 in Lisbon was specifically held in order ‘to strengthen employment, economic reform, and social cohesion in the transition to a knowledge-based economy’ (European Commission, 2000; see also European Commission, 2005). When originally proposing their program of studies of the knowledgebased economy at the OECD, David and Foray (1995, p. 14) argued that the focus on national systems of innovation (Lundvall, 1988, 1992; Nelson, 1993) had placed too much emphasis on the organization of institutions and economic growth, and not enough on the distribution of knowledge itself. The hypothesis of a transition to a ‘knowledge-based economy’ implies a systems transformation at the structural level across nations. Following this lead, the focus of the efforts at the OECD and Eurostat has been to develop indicators of the relative knowledge intensity of industrial sectors (OECD, 2001; 2003) and regions (Laafia, 1999, 2002a, 2002b). Alternative frameworks for ‘systems of innovation’ like technologies (Carlsson and Stankiewicz, 1991) or regions (Braczyk et al., 1998) were also considered (Carlsson, 2006). However, the analysis of the knowledge base of innovation systems (e.g. Cowan et al., 2000) was not made sufficiently relevant for the measurement efforts (David and Foray, 2002). In the economic analysis, knowledge was hitherto not considered as a coordination mechanism of society, but mainly as a public or private good. 291

292


Knowledge as a coordination mechanism was initially defined in terms of the qualifications of the labor force. Machlup (1962) argued that in a ‘knowledge economy’ knowledge workers would play an increasingly important role in industrial production processes. Employment data have been central to the study of this older concept. For example, employment statistics can be cross-tabled with distinctions among sectors in terms of high- and medium-tech (Cooke, 2002; Schwartz, 2006). However, the concept of a ‘knowledge-based economy’ refers to a change in the structure of an economy beyond the labor market (Foray and Lundvall, 1996; Cooke and Leydesdorff, 2006). How does the development of science and technology transform economic exchange processes (Schumpeter, 1939)? The social organization of knowledge production and control was first considered as a systemic development by Whitley (1984). Because of the reputational control mechanisms involved, the dynamics of knowledge production and diffusion are different in important respects from economic market or institutional control mechanisms (Dasgupta and David, 1994; Mirowski and Sent, 2001; Whitley, 2001). When a third coordination mechanism is added as a sub-dynamic to the interactions and potential co-evolution between (1) economic exchange relations and (2) institutional control (Freeman and Perez, 1988), non-linear effects can be expected (Leydesdorff, 1994, 2006). The possible synergies may lead to the envisaged transition to a knowledge-based economy, but this can be expected to happen to a variable extent: developments in some geographically defined economies will be more knowledge-based than others. The geographical setting, the technologies as deployed in different sectors, and the organizational structures of the industries constitute three relatively independent sources of variance. One would expect significant differences in the quality of innovation systems among regions and industrial sectors in terms of technological capacities (Fritsch, 2004). The three sources of variance may reinforce one another in a configuration so that a knowledge-based order of the economy can be shaped. Our research question is whether one can operationalize this configurational order and then also measure it. For the operationalization we need elements from our two research programs: economic geography and scientometrics. We use Storper’s (1997, pp. 26 ff.) notion of a ‘holy trinity’ among technology, organization, and territory from regional economics, and we elaborate Leydesdorff’s (1995, 2003) use of information theory in scientometrics into a model for measuring the dynamics of a triple helix.

Measuring the knowledge base of an economy

293

A COMBINATION OF THEORETICAL PERSPECTIVES Storper (1997, p. 28) defined a territorial economy as ‘stocks of relational assets’. The relations determine the dynamics of the system: Territorial economies are not only created, in a globalizing world economy, by proximity in input–output relations, but more so by proximity in the untraded or relational dimensions of organizations and technologies. Their principal assets –because scarce and slow to create and imitate – are no longer material, but relational.

The ‘holy trinity’ is to be understood not only in terms of elements in a network, but as the result of the dynamics of these networks shaping new worlds. These worlds emerge as densities of relations that can be developed into a competitive advantage when and where they materialize by being coupled to the ground in regions. For example, one would expect the clustering of high-tech services in certain (e.g. metropolitan) areas. The location of such a niche can be considered as a consequence of the selforganization of the interactions (Bathelt, 2003; Cooke and Leydesdorff, 2006). Storper argued that this extension of the ‘heterodox paradigm’ in economics implies a reflexive turn. In a similar vein, authors using the model of a triple helix of university– industry–government relations have argued for considering the possibility of an overlay of relations among universities, industries and governments to emerge from these interactions (Etzkowitz and Leydesdorff, 2000). Under certain conditions the feedback from the reflexive overlay can reshape the network relations from which it emerged. Because of this reflexive turn, the parties involved may become increasingly aware of their own and each other’s expectations, limitations and positions. These expectations and interactions can be further informed by relevant knowledge. Thus the knowledgebased overlay may increasingly contribute to the operation of the system. The knowledge-based overlay and the institutional layer of triple-helix relations operate upon one another in terms of frictions that provide opportunities for innovation both vertically within each of the helices and horizontally among them. The quality of the knowledge base in the economy depends on the locally specific functioning of the interactions in the knowledge infrastructure and on the interface between this infrastructure and the self-organizing dynamics at the systems level. A knowledge base would operate by reducing the uncertainty that prevails at the network level, that is, as a structural property of the system. The correspondence between these two perspectives can be extended to the operationalization. Storper (1997, p. 49), for example, uses a depiction of ‘the economy as a set of intertwined, partially overlapping domains

294


an emerging overlay of relations

Figure 11.1

The emerging overlay in the triple-helix model

of action’ in terms of recursively overlapping Venn diagrams denoting territories, technologies and organizations. Using the triple-helix model, Leydesdorff (1997, p. 112) noted that these Venn diagrams do not need to overlap in a common zone. In a networked arrangement, an overlay of interrelations among the bilateral relations at interfaces can replace the function of central integration by providing an emerging structure (Figure 11.1). The gap in the overlap between the three circles in Figure 11.1 can be measured as a negative entropy, that is, a reduction of the uncertainty in the system (Ulanowicz, 1986). Unlike the mutual information in two dimensions (Shannon, 1948; Theil, 1972), information among three dimensions can become negative (McGill, 1954; Abramson, 1963). This reduction of the uncertainty is a consequence of the networked configuration. The ‘configurational information’ is not present in any of the subsets (Jakulin and Bratko, 2004).1 In other words, the overlay constitutes an additional source or sink of information. A configurational reduction of the uncertainty locally counteracts the prevailing tendency at the systems level towards increasing entropy and equilibrium (Khalil, 2004).

METHODS AND DATA Data The data consist of 1 131 668 records containing information based on the registration of enterprises by the Chambers of Commerce of the Netherlands. These data are collected by Marktselect plc on a quarterly


295

basis. Our data specifically correspond to the CD-Rom for the second quarter of 2001 (Van der Panne and Dolfsma, 2003). Because registration with the Chamber of Commerce is obligatory for corporations, the dataset covers the entire population. We brought these data under the control of a relational database manager in order to enable us to focus on the relations more than on the attributes. Dedicated programs were developed for further processing and computation where necessary. The data contain three variables that can be used as proxies for the dimensions of technology, organization and geography at the systems level. Technology will be indicated by the sector classification (Pavitt, 1984; Vonortas, 2000), organization by the company size in terms of numbers of employees (Pugh and Hickson, 1969; Pugh et al., 1969; Blau and Schoenherr, 1971), and the geographical position by the postal codes in the addresses. Sector classifications are based on the European NACE codes.2 In addition to major activities, most companies also provide information about second and third classification terms. However, we use the main code at the two-digit level. Postal codes are a fine-grained indicator of geographical location. We used the two-digit level, which provides us with 90 districts. Using this information, the data can be aggregated into provinces (NUTS-2) and NUTS-3 regions.3 The Netherlands is thus organized in 12 provinces and 40 regions, respectively. The distribution by company size contains a class of 223 231 companies without employees. We decided to include this category because it contains, among others, spin-off companies that are already on the market, but whose owners are employed by mother companies or universities. Given our research question, these establishments can be considered as relevant economic activities. Knowledge Intensity and High Tech The OECD (1986) first defined knowledge intensity in manufacturing sectors on the basis of R&D intensity. R&D intensity was defined for a given sector as the ratio of R&D expenditure to value added. Later this method was expanded to take account of the technology embodied in purchases of intermediate and capital goods (Hatzichronoglou, 1997). This new measure could also be applied to service sectors, which tend to be technology users rather than technology producers. The discussion continues about how best to delineate knowledge-intensive services (Laafia, 1999, 2002a, 2002b; OECD, 2001, 2003, p. 140). The classification introduced in STI Scoreboard will be used here (OECD, 2001, pp. 137 ff.). The relevant NACE categories are shown in Table 11.1.

296


Table 11.1

Classification of high-tech and knowledge-intensive sectors according to Eurostat

High-tech manufacturing

Knowledge-intensive sectors (KIS)

30 Manufacturing of office machinery and computers 32 Manufacturing of radio, television and communication equipment and apparatus 33 Manufacturing of medical precision and optical instruments, watches and clocks

61 62 64 65

Medium-high-tech manufacturing

70 71

24 Manufacture of chemicals and chemical products 29 Manufacture of machinery and equipment n.e.c. 31 Manufacture of electrical machinery and apparatus n.e.c. 34 Manufacture of motor vehicles, trailers and semi-trailers 35 Manufacturing of other transport equipment

66 67

72 73 74 80 85 92

Water transport Air transport Post and telecommunications Financial intermediation, except insurance and pension funding Insurance and pension funding, except compulsory social security Activities auxiliary to financial intermediation Real-estate activities Renting of machinery and equipment without operator and of personal and household goods Computer and related activities Research and development Other business activities Education Health and social work Recreational, cultural and sporting activities

Of these sectors, 64, 72 and 73 are considered high-tech services. Source:

Laafia (2002a), p. 7.

These classifications are based on normalizations across the member states of the European Union and the OECD, respectively. However, the percentages of R&D and therefore the knowledge intensity at the sectoral level may differ in the Netherlands from the average for the OECD or the EU. Statistics Netherlands (CBS, 2003) provides figures for R&D intensity as percentages of value added in 2001. Unfortunately, these data were aggregated at a level higher than the categories provided by Eurostat and the OECD. For this reason and, furthermore, because the Dutch economy is heavily internationalized so that knowledge can easily spill over from neighboring countries, we decided to use the Eurostat categories provided in Table 11.1 to distinguish levels of knowledge intensity among sectors.


297

Regional Differences The geographical make-up of the Netherlands is different from its image. The share of employment in high- and medium-tech manufacturing in the Netherlands rates only above Luxembourg, Greece and Portugal in the EU-15 (OECD, 2003, pp. 140 ff.). The economically leading provinces of the country, like North- and South-Holland and Utrecht, rank among the lowest on this indicator in the EU-15 (Laafia, 1999, 2002a). The southeast part of the country is integrated in terms of high- and medium-tech manufacturing with neighboring parts of Belgium and Germany. More than 50 percent of private R&D in the Netherlands is located in the regions of Southeast North-Brabant and North-Limburg (Wintjes and Cobbenhagen, 2000). The core of the Dutch economy has traditionally been concentrated on services. These sectors are not necessarily knowledge-intensive, but the situation is somewhat brighter with respect to knowledge-intensive services than in terms of knowledge-based manufacturing. Utrecht and the relatively recently reclaimed province of Flevoland score high on this employment indicator,4 while both North- and South-Holland are in the middle range. South-Holland is classified as a leading EU region in knowledge-intensive services (in absolute numbers), but the high-tech end of these services remain underdeveloped. The country is not homogeneous on any of these indicators. Methodology Unlike a covariation between two variables, a dynamic interaction among three dimensions can generate a complex pattern (Schumpeter, 1939, pp. 174 ff.; Li and Yorke, 1975): one can expect both integrating and differentiating tendencies. In general, two interacting systems reduce the uncertainty on either side with the mutual information or the transmission. Using Shannon’s formulae, this mutual information is defined as the difference between the sum of the uncertainty in two systems without the interaction (Hx + Hy) minus the uncertainty contained in the two systems when they are combined (Hxy). This can be formalized as follows:5 Txy = Hx + Hy – Hxy

(11.1)

Abramson (1963, p. 129) derived from the Shannon formulae that the mutual information in three dimensions is: Txyz = Hx + Hy + Hz – Hxy – Hxz – Hyz + Hxyz

(11.2)

298


While the bilateral relations between the variables reduce the uncertainty, the trilateral term adds to the uncertainty. The layers thus alternate in terms of the sign. The sign of Txyz depends on the magnitude of the trilateral component (Hxyz) relative to the mutual information in the bilateral relations. For example, the trilateral coordination can be associated with a new coordination mechanism that is added to the system. In the network mode (Figure 11.1) a system without central integration reduces uncertainty by providing a differentiated configuration. The puzzles of integration are then to be solved in a non-hierarchical, that is, reflexive or knowledgebased mode (Leydesdorff, 2010).

RESULTS Descriptive Statistics Table 11.2 shows the probabilistic entropy values in the three dimensions (G = geography, T = technology/sector and O = organization) for the Netherlands as a whole and the decomposition at the NUTS-2 level of the provinces. The provinces are very different in terms of the numbers of firms and Table 11.2

Expected information contents (in bits) of the distributions in the three dimensions and their combinations HG

NL % Hmax Drenthe Flevoland Friesland Gelderland Groningen Limburg N.-Brabant N.-Holland Overijssel Utrecht S.-Holland Zeeland

HT

6.205 4.055 95.6 69.2 2.465 1.781 3.144 3.935 2.215 2.838 3.673 3.154 2.747 2.685 3.651 1.802

4.134 4.107 4.202 4.091 4.192 4.166 4.048 3.899 4.086 3.956 3.994 4.178

HO

HGT

2.198 10.189 61.3 82.5 2.225 2.077 2.295 2.227 2.220 2.232 2.193 2.116 2.259 2.193 2.203 2.106

6.569 5.820 7.292 7.986 6.342 6.956 7.682 6.988 6.793 6.611 7.582 5.941

HGO

HTO

HGTO

N

8.385 6.013 12.094 1 131 668 83.2 63.7 75.9 4.684 3.852 5.431 6.158 4.427 5.064 5.851 5.240 5.002 4.873 5.847 3.868

6.039 6.020 6.223 6.077 6.059 6.146 6.018 5.730 6.081 5.928 5.974 6.049

8.413 7.697 9.249 9.925 8.157 8.898 9.600 8.772 8.749 8.554 9.528 7.735

26 210 20 955 36 409 131 050 30 324 67 636 175 916 223 690 64 482 89 009 241 648 24 339


299

their geographical distribution over the postal codes. While Flevoland contains only 20 955 units, South-Holland provides the location for 241 648 firms.6 This size effect is reflected in the distribution of postal codes: the uncertainty in the geographical distribution – measured as HG – correlates significantly with the number of firms N (r = 0.76; p < 0.05). The variance in the probabilistic entropies among the provinces is high (> 0.5) in this geographical dimension, but the variance in the probabilistic entropy among sectors and the size categories is relatively small (< 0.1). Thus the provinces are relatively similar in terms of their sector and size distributions of firms,7 and can thus be meaningfully compared. The second row of Table 11.2 informs us that the probabilistic entropy in the postal codes of firms is larger then 95 percent of the maximum entropy of this distribution at the level of the nation. Since the postal codes are more fine-grained in metropolitan than in rural areas, this indicates that the firm density is not a major source of variance in relation to the population density. However, the number of postal-code categories varies among the provinces, and postal codes are nominal variables that cannot be compared across provinces or regions. The corresponding percentages for the technology (sector) and the organization (or size) distributions are 69.2 and 61.3 percent, respectively. The combined uncertainty of technology and organization (HTO) does not add substantially to the redundancy. In other words, organization and technology have a relatively independent influence on the distribution different from that of postal codes. In the provincial decomposition, however, the highly developed and densily populated provinces (Northand South-Holland, and Utrecht) show a more specialized pattern of sectoral composition (HT) than provinces further from the center of the country. Flevoland shows the highest redundancy in the size distribution (HO), perhaps because certain traditional formats of middle-sized companies are still underrepresented in this new province. The Mutual Information Table 11.3 provides the values for the transmissions (T) among the various dimensions. These values can be calculated straightforwardly from the values of the probabilistic entropies provided in Table 11.2 using Equations 11.1 and 11.2 provided above. The first line for the Netherlands as a whole shows that there is more mutual information between the geographical distribution of firms and their technological specialization (TGT = 72 mbits) than between the geographical distribution and their size (TGO = 19 mbits). However, the mutual information

300

Table 11.3


The mutual information in millibits among two and three dimensions disaggregated at the NUTS-2 level (provinces) TGT

TGO

TTO

TGTO

NL

72

19

240

−34

Drenthe Flevoland Friesland Gelderland Groningen Limburg N.-Brabant N.-Holland Overijssel Utrecht S.-Holland Zeeland

30 68 54 40 65 47 39 65 40 31 62 38

5 6 8 4 7 6 16 30 4 5 6 39

320 164 274 242 353 251 223 285 263 221 223 234

−56 −30 −56 −43 −45 −33 −36 −17 −35 −24 −27 −39

between technology and organization (TTO = 240 mbits) is larger than TGO by an order of magnitude. The provinces exhibit a comparable pattern. While the values for TGT and TGO can be considered as indicators of the geographical clustering of economic activities (in terms of technologies and organizational formats, respectively), the TTO provides an indicator for the correlation between the maturity of the industry (Anderson and Tushman, 1991) and the specific size of the firms involved (Suárez and Utterback, 1995; Utterback and Suárez, 1993; see Nelson, 1994). The relatively low value of this indicator for Flevoland indicates that the techno-economic structure of this province is less mature than in other provinces. The high values of this indicator for Groningen and Drenthe indicate that the techno-economic structure in these provinces is perhaps relatively over-mature. This indicator can thus be considered as representing a strategic vector (Abernathy and Clark, 1985; Watts and Porter, 2003). All values for the mutual information in three dimensions (TTGO) are negative. When decomposed at the NUTS-3 level of 40 regions, these values are also negative, with the exception of two regions that contain only a single postal code at the two-digit level. (In these two cases the uncertainty is by definition zero.)8 However, these values cannot be compared among geographical units without a further normalization.


301

THE REGIONAL CONTRIBUTIONS TO THE KNOWLEDGE BASE OF THE DUTCH ECONOMY One of the advantages of statistical decomposition analysis is the possibility of specifying the within-group variances and the between-group variances in great detail (Theil, 1972; Leydesdorff, 1995). However, a full decomposition at the lower level is possible only if the categories for the measurement are similar among the groups. As noted, the postal codes are nominal variables and cannot therefore be directly compared. Had we used a different indicator for the regional dimension – for example, percentage ‘rural’ versus percentage ‘metropolitan’ – we would have been able to compare and therefore to decompose along this axis, but the unique postal codes cannot be compared among regions in a way similar to the size or the sectoral distribution of the firms (Leydesdorff and Fritsch, 2006). In Leydesdorff et al. (2006, p. 190) we elaborated Theil’s (1972) decomposition algorithm for transmissions. The algorithm enables us to define the between-group transmission T0 as the difference between the sum of the parts and the transmission at the level of the composed system of the country (Table 11.4). Table 11.4

The mutual information in three dimensions statistically decomposed at the NUTS-2 level (provinces) in millibits of information ΔTGTO (= ni * Ti /N) in millibits of information

Drenthe Flevoland Friesland Gelderland Groningen Limburg N.-Brabant N.-Holland Overijssel Utrecht S.-Holland Zeeland Sum (∑i Pi Ti) T0 NL

ni

−1.29 −0.55 −1.79 −4.96 −1.20 −1.96 −5.56 −3.28 −1.98 −1.86 −5.84 −0.83

26 210 20 955 36 409 131 050 30 324 67 636 175 916 223 690 64 482 89 009 241 648 24 339

−31.10 −2.46 −33.55

1 131 668 N = 1 131 668

302


ΔT > –0.50 > –1.00 ≤ –1.00

Figure 11.2

Contribution to the knowledge base of the Dutch economy at the regional (NUTS-3) level

The knowledge base of the country is concentrated in South-Holland (ΔT = −5.84 mbits), North-Brabant (−5.56), and Gelderland (−4.96). North-Holland follows with a contribution of −3.28 mbits of information. The other provinces contribute to the knowledge base less than the in-between provinces interaction effect at the national level (T0 = −2.46 mbit). The further disaggregation (Figure 11.2) shows that the contribution of South-Holland is concentrated in the Rotterdam area, the one in NorthBrabant in the Eindhoven region, and North-Holland exclusively in the agglomeration of Amsterdam. Utrecht, the Veluwe (Gelderland) and the northern part of Overijssel also have above-average contributions on this indicator. These findings correspond with common knowledge about the industrial structure of the Netherlands (e.g. Van der Panne and Dolfsma, 2003), although the contribution of northern Overijssel to the knowledge base of the Dutch economy is a surprise. Perhaps this region profits from a spillover effect of knowledge-based activities in neighboring regions.


303

SECTORAL DECOMPOSITION While the geographical comparison is compounded with traditional industrial structure such as firm density, all effects of the decomposition in terms of the sectoral classification of high- and medium-tech sectors and knowledge-intensive services can be expressed as a relative effect, that is, as a percentage increase or decrease of the negative value of the mutual information in three dimensions when a specific selection is compared with the complete population. We shall consistently use the categories provided by the OECD and Eurostat (Table 11.1) and compare the results with those of the full set as a baseline. Table 11.5 provides the results of comparing the subset of enterprises indicated as high-tech manufacturing and knowledge-intensive services with the full set. Column 4 indicates the influence of high-tech sectors on the knowledge base of the economy. The results confirm our hypothesis that the mutual information or entropy that emerges from the interaction among the three dimensions is more negative for high-tech sectors than for the economy as a whole. The dynamics created by these sectors deepen and tighten the knowledge base. Table 11.5 also provides figures and normalizations for high- and medium-tech manufacturing combined and knowledge-intensive services (KIS), respectively. These results indicate a major effect on the indicator for the sectors of high- and medium-tech manufacturing. The effect is by far the largest in North-Holland, with 943 percent increase relative to the benchmark of all sectors combined. Zeeland has the lowest value on this indicator (365 percent), but the number of establishments in these categories is also the lowest. North-Brabant has the largest number of establishments in these categories, but it profits much less from a synergy effect in the network configuration. The number of establishments in knowledge-intensive services is more than half (51.3 percent) of the total number of companies in the country. These companies are concentrated in North- and South-Holland, with North-Brabant in third position. With the exception of North-Holland, the effect of knowledge-intensive services on this indicator of the knowledge base always leads to a decrease of configurational information. We indicate this with an opposite sign for the change. In the case of NorthHolland, the change is marginally positive (+1.0 percent), but this is not due to the Amsterdam region.9 North-Brabant is second on this rank order with a decrease of −16.6 percent. These findings do not accord with a theoretical expectation about the contributions to the economy of services in general and KIS in particular. Windrum and Tomlinson (1999) argued that to assess the role of KIS, the

304

−34

−56 −30 −56 −43 −45 −33 −36 −17 −35 −24 −27 −39


All sectors

2

−93 −36 −136 −94 −66 −68 −58 −34 −79 −39 −44 −67

−60

Hightech

3

67.6 20.6 144.9 120.1 48.1 105.9 61.2 103.4 127.6 65.9 61.7 73.3

80.2

% change (2/3)

4

786 1307 983 4 885 1 204 2 191 6 375 9 346 2 262 4 843 10 392 554

45 128

N

5

7

−349 −206 −182 −272 −258 −245 −190 −173 −207 −227 −201 −180

−219 526 594 227 536 479 647 430 943 496 859 635 365

553

High- and % change medium-tech (2/6) manufacturing

6

406 401 951 2 096 537 1 031 2 820 2 299 1 167 1 020 2 768 342

15 838

N

8

−34 −18 −37 −25 −29 −18 −30 −17 −20 −13 −15 −28

−24

Knowledgeintensive services

9

Mutual information among three dimensions for different sectors of the economy

NL

Txyz in millibits

1

Table 11.5

−39.1 −37.9 −32.6 −40.8 −34.0 −45.1 −16.6 1.0 −42.8 −45.0 −45.5 −27.8

−27.3

% change (2/9)

10

11 312 10 730 14 947 65 112 14 127 30 040 86 262 126 516 30 104 52 818 128 725 10 503

581 196

N

11


305

degree of integration is more important than the percentage of representation in the economy, and expect KIS especially to contribute favorably to the economy (Bilderbeek et al., 1998; Miles et al., 1995; European Commission, 2000, 2005; OECD, 2000). Unlike output indicators, the measure adopted here focuses precisely on the degree of integration in the configuration. However, our results indicate that KIS unfavorably affects the synergy between technology, organization and territory in the techno-economic system of the Netherlands. Knowledge-intensive services seem to be largely uncoupled from the knowledge flow within a regional or local economy. They contribute unfavorably to the knowledge-based configuration because of their inherent capacity to deliver these services outside the region. Thus a locality can be chosen on the basis of considerations other than those relevant for the generation of a knowledge-based economy in the region. Table 11.6 shows the relative deepening of the mutual information in three dimensions when the subset of sectors indicated as ‘high-tech services’ is compared with KIS in general. More than KIS in general, high-tech services produce and transfer technology-related knowledge (Bilderbeek et al., 1998). These effects of strengthening the knowledge base seem highest in regions that do not have a strong knowledge base in high- and medium-tech manufacturing to begin with, such as Friesland Table 11.6

The subset of high-tech services improves on the knowledge base in the service sector

Txyz in millibits

Knowledge intensive services

High- tech services

% change

N

NL

−24

−34

37.3

41 002


−34 −18 −37 −25 −29 −18 −30 −17 −20 −13 −15 −28

−49 −18 −87 −46 −44 −39 −35 −20 −46 −20 −25 −45

45.2 −4.6 131.5 82.3 49.5 118.7 16.1 17.0 133.1 49.8 69.8 59.7

678 1 216 850 4 380 1 070 1 895 5 641 8 676 1 999 4 464 9 650 483

306


and Overijssel. The effects of this selection for North-Brabant and NorthHolland, for example, are among the lowest. However, this negative relation between high- and medium-tech manufacturing on the one hand, and high-tech services on the other, is not significant (r = −0.352; p = 0.262). At the NUTS-3 level, the corresponding relation is also not significant. Thus the effects of high- and medium-tech manufacturing and high-tech services on the knowledge base of the economy are not related to each other.

CONCLUSIONS AND DISCUSSION Our results suggest that: 1. 2.

the knowledge base of a (regional) economy is carried by high-tech, but more importantly by medium-tech manufacturing; the knowledge-intensive services tend to uncouple the knowledge base from its geographical dimension and thus have a relatively unfavorable effect on the territorial knowledge base of an economy. However, high-tech services contribute to the knowledge-based structuring of an economy.

These conclusions have been confirmed for other countries (Leydesdorff and Fritsch, 2006; Lengyel and Leydesdorff, 2010). In terms of policy implications, our results suggest that regions that are less developed may wish to strengthen their knowledge infrastructure by trying to attract medium-tech manufacturing and high-tech services in particular. The efforts of firms in medium-tech sectors can be considered as focused on maintaining absorptive capacity (Cohen and Levinthal, 1989), so that knowledge and technologies developed elsewhere can more easily be understood and adapted to particular circumstances. High-tech manufacturing may be more focused on the (internal) production and global markets than on the local diffusion parameters. High-tech services, however, mediate technological knowledge more than knowledge-intensive services, which are medium-tech. KIS seems to have an unfavorable effect on territorially defined knowledge-based economies. Unlike the focus on comparative statics in employment statistics and the STI Scoreboards of the OECD (OECD, 2001, 2003; Godin, 2006), the indicator developed here measures the knowledge base of an economy as an emergent property (Jakulin and Bratko, 2004; see Ulanowicz, 1986, pp. 142 ff.). This, then, is the most important contribution of this chapter: the knowledge base of an economy (the explanandum of this study) can


307

be measured as an overlay of communications among differently codified communications using the triple-helix model for the explanation (i.e. as the explanans) (Luhmann, 1986, 1995; Leydesdorff, 2006). When market expectations, research perspectives and envisaged retention mechanisms can be interfaced, surplus value can be generated by reducing uncertainty prevailing at the systems level without diminishing room for future explorations using a variety of perspectives (Ashby, 1958).

ACKNOWLEDGMENT A different version of this chapter appeared in Research Policy, Vol. 35, No. 2, 2006, pp. 181–99.

NOTES 1. The so-called interaction or configurational information is defined by these authors as the mutual information in three dimensions, but with the opposite sign (McGill, 1954; Han, 1980). 2. NACE stands for Nomenclature générale des Activités économiques dans les Communautés Européennes. The NACE code can be translated into the International Standard Industrial Classificiation (ISIC) and in the Dutch national SBI (Standaard Bedrijfsindeling) developed by Statistics Netherlands. 3. These classifications are used for statistical purposes by the OECD and Eurostat. NUTS stands for Nomenclature des Unités Territoriales Statistiques (Nomenclature of Territorial Units for Statistics). 4. Flevoland is the only Dutch province eligible for EU support through the structural funds. 5. Hx is the uncertainty in the distribution of the variable x (i.e. Hx = − ∑x px 2log px), and analogously, Hxy is the uncertainty in the two-dimensional probability distribution (matrix) of x and y (i.e. Hxy = − ∑x ∑y pxy 2log pxy). The mutual information will be indicated with the T of transmission. If the basis two is used for the logarithm, all values are expressed in bits of information. 6. The standard deviation of this distribution is 80 027.04 with a means of 94 305.7. 7. The value of H for the country corresponds to the mean of the values for the provinces ¯T = 4.088 ± 0.097 and H ¯O = 2.196 ± 0.065. in these dimensions: H 8. These are the regions Delfzijl and Zeeuwsch-Vlaanderen (COROP / NUTS-3 regions 2 and 31). 9. Only in NUTS-3 region 18 (North-Holland North) is the value of the mutual information in three dimensions more negative when zooming in on the knowledge-intensive services. However, this region is predominantly rural.

REFERENCES Abernathy, W. and Clark, K.B. (1985), ‘Mapping the winds of creative destruction’, Research Policy, 14, 3–22.

308


Abramson, N. (1963), Information Theory and Coding, New York: McGrawHill. Anderson, P. and M.L. Tushman (1991), ‘Managing through cycles of technological change’, Research-Technology Management, 34 (3), 26–31. Ashby, W.R. (1958). ‘Requisite variety and its implications for the control of complex systems’, Cybernetica, 1 (2), 1–17. Bathelt, H. (2003), ‘Growth regimes in spatial perspective 1: innovation, institutions and social systems’, Progress in Human Geography, 27 (6), 789–804. Bilderbeek, R., P. Den Hertog, G. Marklund and I. Miles (1998), Services in Innovation: Knowledge Intensive Business Services (KIBS) as C-producers of innovation, STEP report No. S14S. Blau, P.M. and R. Schoenherr (1971), The Structure of Organizations, New York: Basic Books. Braczyk, H.-J., P. Cooke and M. Heidenreich (eds) (1998), Regional Innovation Systems, London and Bristol, PA: University College London Press. Carlsson, B. (2006), ‘Internationalization of innovation systems: a survey of the literature’, Research Policy, 35 (1), 56–67. Carlsson, B. and R. Stankiewicz (1991), ‘On the nature, function, and composition of technological systems’, Journal of Evolutionary Economics, 1 (2), 93–118. Carter, A.P. (1996), ‘Measuring the performance of a knowledge-based economy’, in D. Foray and B.-Å. Lundvall (eds), Employment and Growth in the KnowledgeBased Economy, Paris: OECD, pp. 203–11. CBS (2003), Kennis en Economie 2003, Voorburg and Heerlen: Centraal Bureau voor de Statistiek. Cohen, W.M. and D.A. Levinthal (1989), ‘Innovation and learning: the two faces of R&D’, The Economic Journal, 99, 569–96. Cooke, P. (2002), Knowledge Economies, London: Routledge. Cooke, P. and L. Leydesdorff (2006), ‘Regional development in the knowledgebased economy’, Journal of Technology Transfer, 31 (1), 5–15. Cowan, R., P. David and D. Foray (2000), ‘The explicit economics of knowledge codification and tacitness’, Industrial and Corporate Change, 9 (2), 211–53. Dasgupta, P. and P. David (1994), ‘Towards a new economics of science’, Research Policy, 23 (5), 487–522. David, P. and D. Foray (1995), ‘Assessing and expanding the science and technology knowledge base’, STI Review, 16, 13–68. David, P.A. and D. Foray (2002), ‘An introduction to the economy of the knowledge society’, International Social Science Journal, 54 (171), 9–23. Etzkowitz, H. and L. Leydesdorff (2000), ‘The dynamics of innovation: from national systems and “Mode 2” to a triple-helix of university–industry– government relations’, Research Policy, 29 (2), 109–23. European Commission (2000), Towards a European Research Area, Brussels, 18 January, at http://europa.eu.int/comm/research/era/pdf/com2000-6-en.pdf. European Commission (2005), Working Together for Growth and Jobs. A New Start for the Lisbon Strategy, at http://europa.eu.int/growthandjobs/pdf/ COM2005_024_en.pdf. Foray, D. and B.-Å. Lundvall (1996), ‘The knowledge-based economy: from the economics of knowledge to the learning economy’, in Employment and Growth in the Knowledge-Based Economy, Paris: OECD, pp. 11–32. Freeman, C. and C. Perez (1988), ‘Structural crises of adjustment, business cycles and investment behaviour’, in G. Dosi, C. Freeman, R.N.G. Silverberg and


309

L. Soete (eds), Technical Change and Economic Theory, London: Pinter, pp. 38–66. Fritsch, M. (2004), ‘Cooperation and the efficiency of regional R&D activities’. Cambridge Journal of Economics, 28 (6), 829–46. Godin, B. (2006), ‘The knowledge-based economy: conceptual framework or buzzword’, Journal of Technology Transfer, 31 (1), 17–30. Han, T.S. (1980), ‘Multiple mutual information and multiple interactions in frequency data’, Information and Control, 46 (1), 26–45. Hatzichronoglou, T. (1997), Revision of the High-Technology Sector and Product Classification, Paris: OECD, http://www.olis.oecd.org/olis/1997doc.nsf/LinkTo/ OCDE-GD(97)216. Jakulin, A. and I. Bratko (2004), Quantifying and Visualizing Attribute Interactions: An Approach Based on Entropy, http://arxiv.org/abs/cs.AI/0308002. Khalil, E.L. (2004), ‘The three laws of thermodynamics and the theory of production’, Journal of Economic Issues, 38 (1), 201–26. Laafia, I. (1999), Regional Employment in High Technology, Eurostat, http://epp. eurostat.ec.europa.eu/cache/ITY_OFFPUB/CA-NS-99-001/EN/CA-NS-99001-EN.PDF. Laafia, I. (2002a), ‘Employment in high tech and knowledge intensive sectors in the EU continued to grow in 2001’, Statistics in Focus: Science and Technology, Theme, 9 (4), at http://epp.eurostat.ec.europa.eu/cache/ITY_OFFPUB/KS-NS02-004/EN/KS-NS-02-004-EN.PDF. Laafia, I. (2002b), ‘National and regional employment in high tech and knowledge intensive sectors in the EU – 1995–2000’, Statistics in Focus: Science and Technology, Theme 9 (3), http://epp.eurostat.ec.europa.eu/cache/ITY_ OFFPUB/KS-NS-02-003/EN/KS-NS-02-003-EN.PDF. Lengyel, B. and L. Leydesdorff (2010), ‘Regional innovation systems in Hungary: the failing synergy at the national level’, Regional Studies (in press). Leydesdorff, L. (1994), ‘Epilogue’, in L. Leydesdorff and P. Van den Besselaar (eds), Evolutionary Economics and Chaos Theory, London and New York: Pinter, pp. 180–92. Leydesdorff, L. (1995), The Challenge of Scientometrics: The Development, Measurement, and Self-Organization of Scientific Communications, Leiden: DSWO Press. Leydesdorff, L. (1997), ‘The new communication regime of university–industry–government relations’, in H. Etzkowitz and L. Leydesdorff (eds), Universities and the Global Knowledge Economy, London and Washington, DC: Pinter, pp. 106–17. Leydesdorff, L. (2003), ‘The mutual information of university–industry– government relations’, Scientometrics, 58 (2), 445–67. Leydesdorff, L. (2006), The Knowledge-Based Economy: Modeled, Measured, Simulated, Boca Rota, FL: Universal Publishers. Leydesdorff, L. (2010), ‘Redundancy in systems which entertain a model themselves: interaction information and self-organization of anticipation’, Entropy, 12 (1), 63–79. Leydesdorff, L. and M. Fritsch (2006), ‘Measuring the knowledge base of regional innovation systems in Germany in terms of a triple helix dynamics’, Research Policy, 35 (10), 1538–53. Leydesdorff, L., W. Dolfsma and G. Van der Panne (2006), ‘Measuring the knowledge base of an economy in terms of triple-helix relations among technology, organization, and territory’, Research Policy, 35 (2), 181–99.

310


Li, T.-Y. and J.A. Yorke (1975), ‘Period three implies chaos’, American Mathematical Monthly, 82 (10), 985–92. Luhmann, N. (1986), Love as Passion: The Codification of Intimacy, Stanford, CA: Stanford University Press. Luhmann, N. (1995), Social Systems, Stanford, CA: Stanford University Press. Lundvall, B.-Å. (1988), ‘Innovation as an interactive process: from user–producer interaction to the national system of innovation’, in G. Dosi, C. Freeman, R. Nelson, G. Silverberg and L. Soete (eds), Technical Change and Economic Theory, London: Pinter, pp. 349–69. Lundvall, B.-Å. (ed.) (1992), National Systems of Innovation, London: Pinter. Machlup, F. (1962), The Production and Distribution of Knowledge in the United States, Princeton, NJ: Princeton University Press. McGill, W.J. (1954), ‘Multivariate information transmission’, Psychometrika, 19 (2), 97–116. Miles, I., N. Kastrinos, K. Flanagan, R. Bilderbeek, P. Den Hertog, W. Hultink and M. Bouman (1995), Knowledge-Intensive Business Services: Users, Carriers and Sources of Innovation, European Innovation Monitoring Service, No. 15, Luxembourg. Mirowski, P. and E.-M. Sent (2001), Science Bought and Sold, Chicago, IL: University of Chicago Press. Nelson, R.R. (ed.) (1993), National Innovation Systems: A Comparative Analysis, New York: Oxford University Press. Nelson, R.R. (1994), ‘Economic growth via the coevolution of technology and institutions’, in L. Leydesdorff and P. Van den Besselaar (eds), Evolutionary Economic and Chaos Theory, London and New York: Pinter, pp. 21–32. OECD (1986), OECD Science and Technology Indicators: R&D, Invention and Competitiveness, Paris: OECD. OECD (1996), New Indicators for the Knowledge-Based Economy: Proposals for Future Work, DSTI/STP/NESTI?GSS/TIP (96) 6. OECD (2000), Promoting Innovation and Growth in Services, Paris: OECD. OECD (2001), Science, Technology and Industry Scoreboard: Towards a Knowledgebased Economy, Paris: OECD. OECD (2003), Science, Technology and Industry Scoreboard, Paris: OECD. OECD/Eurostat (1997), Proposed Guidelines for Collecting and Interpreting Innovation Data, ‘Oslo Manual’, Paris: OECD. Pavitt, K. (1984), ‘Sectoral patterns of technical change’, Research Policy, 13 (6), 343–73. Pugh, D.S. and D.J. Hickson (1969), ‘The context of organization structures’, Administrative Science Quarterly, 14 (1), 91–114. Pugh, D.S., D.J. Hickson and C.R. Hinings (1969), ‘An empirical taxonomy of structures of work organizations’, Administrative Science Quarterly, 14 (1), 115–26. Schumpeter, J. [1939] (1964), Business Cycles: A Theoretical, Historical and Statistical Analysis of Capitalist Process, New York: McGraw-Hill. Schwartz, D. (2006), ‘The regional location of knowledge based economy activities in Israel’, The Journal of Technology Transfer, 31 (1), 31–44. Shannon, C.E. (1948), ‘A mathematical theory of communication’, Bell System Technical Journal, 27 (July and October), 379–423 and 623–56. Storper, M. (1997), The Regional World, New York: Guilford Press. Suárez, F.F. and J.M. Utterback (1995), ‘Dominant design and the survival of firms’, Strategic Management Journal, 16 (6), 415–30.


311

Theil, H. (1972), Statistical Decomposition Analysis, Amsterdam and London: North-Holland. Ulanowicz, R.E. (1986), Growth and Development: Ecosystems Phenomenology, San Jose, CA: Excel Press. Utterback, J.M. and F.F. Suárez (1993), ‘Innovation, competition, and industry structure’, Research Policy, 22 (1), 1–21. Van der Panne, G. and W. Dolfsma (2003), ‘The odd role of proximity in knowledge relations’, Journal of Economic and Social Geography, 94 (4), 451–60. Vonortas, N.S. (2000), ‘Multimarket contact and inter-firm cooperation in R&D’, Journal of Evolutionary Economics, 10 (1–2), 243–71. Watts, R.J. and A.L. Porter (2003), ‘R&D cluster quality measures and technology maturity’, Technological Forecasting & Social Change, 70 (8), 735–58. Whitley, R.D. (1984), The Intellectual and Social Organization of the Sciences, Oxford: Oxford University Press. Whitley, R.D. (2001), ‘National innovation systems’, in N.J. Smelser and P.B. Baltes (eds), International Encyclopedia of the Social and Behavioral Sciences, Oxford: Elsevier, pp. 10303–9. Windrum, P. and M. Tomlinson (1999), ‘Knowledge-intensive services and international competitiveness’, Technology Analysis and Strategic Management, 11 (3), 391–408. Wintjes, R. and J. Cobbenhagen (2000), ‘Knowledge intensive industrial clustering around Océ’, MERIT Research Memorandum No. 00-06. MERIT, University of Maastricht.

12.

Knowledge networks: integration mechanisms and performance assessment Matilde Luna and José Luis Velasco

Triple-helix relations involve communication among systems that have distinctive codes and languages. Therefore they should be conceptualized as complex systems, or as systems of network-structured social relations. Because of their complexity, triple-helix relations are hard to analyse and evaluate. Strictly speaking, they are not economic, political or scientific organizations, and therefore conventional criteria and standards for assessing performance cannot be taken for granted. This chapter addresses a set of theoretical and methodological issues related to the functioning and performance of organizations involved in triple-helix relations. More specifically, we look at four integration mechanisms and their impact on several criteria of performance. We consider two types of network performance: functional and organizational, respectively related to the practical results of collaboration and the conditions for the production of such results. We shall focus on academy–industry relations or knowledge networks conceived as complex problem-solving structures devoted to the generation and diffusion of knowledge through the establishment of collaborative links between academy and industry. We begin by drawing some theoretical insights from the literature on network complexity. We then identify four mechanisms for integrating actors with different and sometimes diverging codes, interests, needs, preferences, resources and abilities: mutual trust, translation, negotiation and deliberation. Next, we focus on trust and translation, respectively related to problems of internal social cohesion and communication. Next we turn to negotiation and deliberation, both related to decision-making. Finally, we try to establish to what extent, in what sense, and under what conditions knowledge networks can or cannot be considered effective, efficacious and efficient, and propose some criteria for assessing functional and organizational performance. The empirical data that illustrate some of our analytical claims come 312

Integration mechanisms and performance assessment

313

from 38 structured interviews with people from academic and economic organizations.1 The interviewees are from different economic sectors, technological fields and regions in Mexico. All of them have participated in joint research projects aimed at the generation and diffusion of knowledge. Most of the firms involved were large and had R&D departments. While not statistically representative, these data allow us to put forward some hypotheses and proposals regarding the operation of integration mechanisms and their impact on network performance. Given the heterogeneity of the cases and the diversity of the actors interviewed, we can reasonably suppose that high frequencies or low levels of data dispersion may indicate highly significant social regularities.

NETWORKS AS COMPLEX SYSTEMS Since the early 1990s, the network metaphor has inspired a variety of theoretical, methodological and technical developments. Despite their differences, three seminal approaches tend to converge on the notion of networks as complex systems: social network analysis, the study of networks as social coordination mechanisms, and actor-network theory. These approaches provide some theoretical foundations for the study of integration mechanisms and network performance. Social network analysis developed from the notion that networks are systems of bonds between nodes and that bonds are structures of interpersonal communication. Nodes can be individuals, collective entities (e.g. organizations or countries), or positions in the network. This approach has centred on the morphological dimension of networks, asking how actors are distributed in informal structures of relations and what are the boundaries or limits of a network. Its main concerns include the operationalization, measurement, formalization and representation of ties. Only recently has this approach developed mathematical models and tools to analyse network dynamics. The dominant image in this literature is one of a dense, egocentric network, comprising homogeneous actors linked by strong (intense) ties. One of the most interesting problems analysed through this approach is ‘the strength of weak ties’ (Granovetter, 1973). A tie with a low level of interpersonal intensity, it is argued, may have a high level of informative strength if it is a ‘bridge’, that is to say, the only link between two or more groups, each of them formed by individuals connected by strong bonds. Seeing networks as complex organizations, Burt (1992) put forward the notion of ‘structural holes’: sparse regions located between two or more dense regions of relations and representing opportunities for the

314


circulation of original information or new ideas (Walker et al., 1997; see also Valente, 1995). Theories of networks as social coordination mechanisms have distinguished several patterns of social coordination, each associated with a different type of institution. Hierarchical control has been associated with the state, exchange with the market, solidarity with the community, and concerting with the associative model (Eising and Kohler-Koch, 1999; Martinelli, 2002; Streeck and Schmitter, 1992). More recently, networks have also been seen as a specific mode of social coordination. One of the most interesting developments incorporates elements of evolutionary theories. From this perspective, what distinguishes networks from other forms of coordination is their high level of complexity, which results from the differentiation, specialization and interdependence of social systems. Seen as a form of governance, a network is represented as a polycentric configuration and as a system of relations having weak but strong ties and a membership that is both elastic and heterogeneous. Within these configurations, actors follow different codes or coordination principles – for example, exchange or profit, legitimate authority or the law, and solidarity – which may be mutually inconsistently. Arising from the sociology of science, technology and knowledge, actornetwork theory emphasizes the autonomy and indeterminacy of networks and their actors. Bruno Latour (1993, p. 91), its main proponent, holds that the separation of nature, society and language is artificial. Reality, he claims, is ‘simultaneously real like nature, narrated like discourse, and collective like society’. Actor-network theory seeks to describe the association of humans and non-humans (objects, abilities, machines), or ‘actants’. This theory opened social sciences to the non-human part of the social world (Callon and Law, 1997). Thus a network is a configuration of animated and inanimate elements. An actor-network is simultaneously an actor that brings together heterogeneous and differentiated elements, and a network that transforms its own elements. In this view, translation – defined as the combination of negotiations, intrigues, calculations, acts of persuasion, and violence by means of which an actor or force acquires the authority to speak or act on behalf of other actors or forces – provides the link that connects actors or ‘actants’ into a network and defines the network itself. Translation both allows actors to communicate and defines the evolution of the network. In fact, from this perspective, a network can be seen as a system of translation (Leydesdorff, 2001). From these approaches, we can conclude that triple-helix relations should be considered complex systems of relations, resulting from


315

simultaneous processes of differentiation and interdependence among individuals, groups, institutions or subsystems. The collaborative projects analysed here comprise moderate or highly complex interactions that can be usefully conceptualized as networks. Members act according to different codes. The networks that they create have their own rules of interaction and are highly autonomous from the original organizations. Authority within the network is dynamically dispersed and decision-making involves various actors and complex processes of negotiation and deliberation. Relations among members also involve the collective construction of objectives through a variety of coordination mechanisms (such as formal agreements, personal relationships, committees of all sorts, working teams) and diverse communication techniques. Some specific data show other dimensions of this complexity, such as the weakness of the ties among members and institutions, the elasticity of the network, and the uncertainty of their results. On the weakness of ties, 82 per cent of the interviewees considered that there were important obstacles to the flow of knowledge, and nearly 50 per cent recognized that there had been ‘important differences in opinion among the members’. Elasticity is shown by the fact that the original aims of collaboration were changed and new members were included as projects developed. On the uncertainty of results, only half of the people interviewed thought that ‘the problem addressed was solved’ and a relatively high 24 per cent ‘do not know whether the problem addressed was solved’. However, most of them (82 per cent) claimed that other problems, not initially addressed, were solved. Being complex entities, knowledge networks are autonomous in a double sense. First, each of their components (institutions, organizations and individuals) is autonomous and remains so even as interaction and collaboration intensify. Second, the whole network is autonomous, in the sense that it is not subject to a superior entity that regulates its actions. This individual and organizational autonomy entails that there are no settled rules unequivocally determining the rights and obligations of members (self-regulation) and that participants are reasonably free to express their opinions and to choose their options (self-selection). At the level of decision-making, the double autonomy of networks means that no member has absolute authority, each has certain autonomy (Hage and Alter, 1997) and authority is dynamically distributed within the network. In the words of one participant: ‘leadership shifts from one place to another, from one [strategic] member to another’, depending on the nature of the issues addressed. To coordinate heterogeneous actors, structure and process conflicts, collectively make decisions, reach agreements and solve problems, triple-helix

316


relations combine four mechanisms of integration: complex trust, translation, negotiation and deliberation. Each mechanism operates in other social structures as well, for example, markets, politics and communities. What is distinctive of networks is that they require the combined operation of the four mechanisms and assign a central role to three of them: complex trust, translation and deliberation.

TRUST AND TRANSLATION Two of the four integrative mechanisms analysed here have to do with how networks structure their internal relations, creating a common ground where people from universities and business firms may fruitfully interact, cooperate and thereby generate knowledge that is at the same time scientifically valid, technically useful and economically profitable. Trust and Internal Cohesion Interpersonal or mutual trust is a set of positive expectations regarding the actions of other people. Such expectations become important when someone has to choose a course of action knowing that its success depends, to some extent, on the actions of others; and yet such a decision must be made before it is possible to evaluate the actions of the others (Dasgupta, 1988). Trust, hence, has three basic features: interdependence, uncertainty and a positive expectation. There is a trust relation when the success of a person’s action depends on the cooperation of someone else; therefore it entails at least partial ignorance about the behaviour of the others and the assumption that they will not take undue advantages (Lane, 1998, p. 3; Sable, 1993). For knowledge networks, three kinds of trust are particularly significant: calculated or strategic, technical or prestige-based, and normative. Prestige-based or technical trust depends mainly on perceptions about the capabilities and competences of participants. Calculated or strategic trust arises from calculations of costs and benefits and is based on the expectation of mutual benefits from the relationship. Personal or normative trust depends on shared norms, beliefs and values; it is based on social solidarity, rather than on the expected benefits of the interaction. Interviews with participants showed that the three types of trust are highly important. They were asked to rate an indicator of each kind of trust on a scale ranging from 0 to 10, where 10 meant ‘very important for the development of the collaborative project’. Ratings for each dimension of trust averaged between 8.8 and 9.2.


317

We also found that there is a strong internal consistency among the three dimensions of trust; each of them can be either a cause or an effect of the other two; any of them can give rise to transitive chains (A trusts B, B trusts C, therefore A trusts C); they may be mutually supportive, overlapping or conflicting, and the way they combine has important consequences for the origin, development, and dissolution of triple-helix relations. The existence of one type of trust may increase the opportunities for the development of the other types; conversely, the predominance of one type may undermine the others. For instance, a relation initially based on normative trust may give rise to technical trust; or a collaborative project motivated solely by expectations of mutual gain may impede the development of normative or personal trust, which is the firmest basis of interpersonal communication. In the former case, trust may help stabilize or integrate the network, or even operate as a multiplying factor creating new relationships out of an original system of ties. In the latter case, trust creates problems of coordination, efficacy, or efficiency.2 It seems clear, therefore, that an important amount of trust in each of its three types is indispensable in complex networks. That is why we should not refer to normative, strategic and technical ‘trusts’ as separate kinds of trust but as dimensions of it. Therefore, in complex networks, the total amount of trust is a combination (the ‘algebraic sum’) of the three dimensions. If one neglects this fact, it becomes difficult to explain how trust can be generalized among people with different ideas, from diverse disciplinary and organizational cultures and even from different countries, as is the case in knowledge networks. More generally, it could be said that as a property of complex systems – and as a mechanism of integration among actors with different codes and languages – mutual trust takes on a complex character. It involves calculations, solidarity, and a perception of the technical prestige of participants. Thus, as opposed to simple social interactions based on a single type of trust, triple-helix relations involve an unstable balance among these three dimensions of trust. Therefore the evaluation of network performance has to consider not only the total level of mutual trust produced or wasted during the interaction but also the way in which its different dimensions are combined. This leads us to question the assumption, shared by many theories of trust, that there is always a positive relationship between trust and performance. Translation and Communication The problem of translation is located at the core of the weak-tie paradox, according to which the flow of new ideas and original information is

318


associated with the mismatch of language and cognitive orientation (Steward and Conway, 1996; Granovetter, 1973; Valente, 1995; Burt, 2000; Lundvall, 2000). More strongly, translation systems may be seen as the answer to an evolutionary paradox, with integration being a form of de-differentiation (Leydesdorff, 1997). Moreover, from diverse theoretical approaches, the literature on innovation and academy–industry relations has identified boundary personnel under different names, such as gatekeepers, brokers, boundary-spanners, gold-collar workers or translators, and has assigned them a critical role in collaboration (e.g. Leydesdorff, 1997; Steward and Conway, 1996). Drawing on this literature and on data from the projects described above, we hold that translation performs five critical functions in knowledge networks, creating a common ground between different and sometimes conflicting cognitive orientations, coding schemes, ‘local languages’, normative orientations and interests.3 At the level of cognitive orientation, translation connects organizations and people with diverging notions of knowledge and knowledge creation. As Leydesdorff (2001) has noted, while scientific propositions are usually evaluated in terms of ‘truth’, market-oriented communications are normally judged by their utility or profit potential. But there are also important differences in the notion of innovation. For some participants, innovation means new ideas or the rupture of paradigms; for others, changes in the market (products, technology, organization patterns, or trade techniques). Moreover, actors from academic institutions and firms often perceive the mismatch of cognitive orientation as a difference in language, sometimes considered the major impediment to interaction. More generally, interviewees referred to differences in culture, approaches and time spans. The task of reconciling these differences becomes all the more difficult because they often overlap and combine in complex ways. Second, at the inter-organizational level, translation reconciles different organizational structures and procedural mechanisms. Mutual understanding within the network is made difficult by the existence of diverging standards about confidentiality, information flows, intellectual property, patents, evaluation criteria and administration. The need for translation at this level increases because network members remain autonomous and control their own resources, which entails that virtually any decision has to be jointly made. Third, operating at the interdisciplinary level, translation contributes to the development of a problem-solving approach, which commonly involves the collaboration of multiple disciplines. At this level, translation also deals with tensions between basic and applied research. Fourth, at the level of codification, translation combines ‘local’ and


319

‘universal’ knowledge. As one interviewee put it, a ‘translator’ must have the ‘sensitivity’ to simultaneously ‘see an industrial problem’ and ‘the basic science behind it’. Finally, at the level of interests and negotiations, translation deals with power struggles, many of them related to asymmetries and differences in the kinds of ‘goods’ exchanged within the network. Networks are complex entities that have to coordinate particular, divergent and common interests (Messner, 1999). At this level, translation is critical to the efficacy of the network, which is often defined as the capacity to manage conflicts. Translation may be approached at both the structural and individual levels. By bringing together people from academic institutions and firms, knowledge networks function as translation structures. They connect entities from two social subsystems, each with its own cognitive orientations, coding schemes, ‘local languages’ and normative orientations. But, inside the network, some individuals specialize in translation, facilitating communication and understanding among members. In this respect, it is noticeable that most of the interviewees identified a person who facilitated communication between participants. Translators are usually individuals who have worked in both the business and academic sectors. They have interdisciplinary knowledge and skills; understand the cultures and procedures of different organizations; have many varied and informal links with the other members of the network; and are distinguished by personal traits that allow them to act as interpersonal communication facilitators. They posses high amounts of tacit knowledge and frequently command all types of knowledge: knowwho, know-what, know-how and know-why. Although translators may occupy marginal positions in the network, they are not just carriers of messages from one sector to another. They transform scientific knowledge into economically useful information, knowledge, products and processes. Conversely, they transform the practical knowledge requirements of firms into scientifically relevant questions. Trust or Translation? Within certain limits, there seems to be an inversely proportional relation between trust and translation: when members of the network strongly trust each other, communication is easier and translation less necessary. Only those respondents who thought that communication among members was ‘very difficult’, ‘difficult’, or ‘easy’ believed that there was one person facilitating communication. People who thought that communication was ‘very easy’ denied that such a facilitator existed. Another interesting fact is that trustworthiness is an essential

320


characteristic of individuals identified as translators. Personal descriptions of translators are full of phrases or words such as ‘very participative’, ‘intelligent person’, ‘willingness’, ‘self-motivated individual’, ‘empathy’, ‘trust’, ‘reliable person’ and so on. In sum, translators must be trustworthy people – trustworthy in a complex way, with features corresponding to each of the three dimensions of trust.4

NEGOTIATION AND DELIBERATION Given the complexity of triple-helix relations, their collective decisions have to be reached by bargaining, by giving reasons, and by solving conflicts or problems. This is especially true of decisions concerning the aims of the project, the nature of problems to be addressed, and the best ways to solve them – all of which decisively shape the origins, dynamics and evolution of knowledge networks.5 Interviewees showed a generalized perception that a clear definition of aims was decisive for the consolidation of networks. It was seen as even more important than the amount of financial support, the existence of previous relations among participants, the specific capacities of institutions and actors for solving the problem in question, and the incentives provided through public policy instruments. Conflicts and differences about the nature of the problem addressed can be exemplified with the following statement made by a participant from an academic institution: People from industry always claim to know what the problem is from a nontechnical standpoint. But if you get into it, you realize that what lies behind is very different . . . Sometimes, they think they have a problem of processes, and perhaps it is in fact a problem of materials; or they see it as a problem of materials and indeed is a problem of characterisation . . . To break down [the problem] also allows us to identify other problems, of which they were probably unaware even though the problems were about to explode.

According to Elster (1999), there are three basic forms of collective decision-making: voting, bargaining or negotiation, and deliberation. Although not mutually exclusive, they have fundamental differences in their theoretical presuppositions, possibilities and constraints. Voting has been considered a very effective and efficient method, generally producing clear and quick decisions (Jachtenfuchs, 2006). However, it is hardly suitable to knowledge networks, characterized by the search for consensus6 through a mixture of deliberation and negotiation. Therefore we shall concentrate on the features, possibilities and constraints of these two ways of making decisions.


321

In contrast to theories of social coordination (e.g. Messner, 1999) and actor-network theory, which claim that networks are predominantly governed by the logic of negotiation, we argue that deliberation is what distinguishes them from other forms of coordinating actions and making collective decisions. Networks make decisions mainly through the rational exchange of arguments; this sets them apart from political, economic or social structures that rely primarily on power, money, or solidarity.7 This is particularly evident in academy–industry relations, where information, ideas, knowledge and expertise – resources that are essential for argumentation and collectively making judgements – play an important role in authority formation. Negotiation The central distinction between negotiation and deliberation can be approached from the standpoint of interests. Whereas in negotiation interests are fixed and defined beforehand, deliberation involves the collective definition of preferences. Deliberation presupposes that interests are not external to the political process: the debate and exchange of arguments transform preferences, making them more compatible.8 In an interdependent situation, negotiation seeks either ‘to create something new that neither party could do on his or her own, or . . . to resolve a problem or dispute between the parties’. It may be an informal haggling (bargaining) or a ‘formal, civilized process’ through which parties try to ‘find a mutually acceptable solution to a complex conflict’ (Lewicki et al., 2004, p. 3). A negotiation situation has several distinctive characteristics: there are two or more parties, who are interdependent (they need each other) but have conflicting interests; the parties prefer to seek an agreement instead of fighting openly or opting out of the relationship; there is no set of rules that automatically resolves the conflict of interests (Lewicki et al., 2004). The main concern of participants in this competitive cooperation is how to distribute the costs and benefits of the interaction. As long as participants remain diverse and autonomous from each other, each of them independently controlling important resources and sharing in the distribution of power, negotiation will always be necessary. This makes it an essential, permanent feature of network decisionmaking. However, other structural features of knowledge networks place important limits on the use of negotiation. First, interests – and therefore objectives, goals, strategies, gains and losses – are defined and redefined within the network itself. They are internal to the network itself, transformed and even generated by it. The

322


problems that networks usually address are necessarily complex. If participants were able to correctly define them in a way that is both scientifically correct and economically useful, then there would scarcely be any need to create networks like these. This interactive redefinition of interests and problems fundamentally transforms network exchange, making it different from market bargaining or political negotiation. Before being negotiated, those interests have to be defined through other communication and decision-making mechanisms. But, in the second place, networks are more than exchange mechanisms. They are autonomous organizations: collective actors in their own right, with their own interests, goals, strategies, gains, losses and problems to solve. As seen above, trust and translation play critical roles in creating a common ground and solidifying the collective structure of the network, even if individual participants remain mutually independent. Negotiation is embedded in this collective structure and therefore occupies a narrower place than it does in market bargaining and conventional political negotiations. Given that many network decisions concern not the best way to accommodate diverse individual interests but the best solution to common problems, they have to be agreed upon rather than negotiated. Deliberation At its core, deliberation refers to the rational exchange of arguments aiming at reasonable decisions and solutions. Its main goal is to identify a common good, which implies a redefinition of private interests. As Elster (1999, pp. 12–13) puts it, ‘When the private and idiosyncratic wants have been shaped and purged in public discussion about the public good, uniquely determined rational agreement would emerge. Not optimal compromise but unanimous agreement’ would be the result of such a process. As suggested above, this is especially true of knowledge networks, whose main explicit purpose is to solve knowledge-related problems that are not precisely defined beforehand. Even when it fails to identify a common good, deliberation may at least result in a ‘collective evaluation of divergences’ (Oléron, 1983, p. 108). This evaluation may facilitate mutual understanding among members of the network. Moreover, it almost inevitably leads participants to redefine their interests, objectives and goals. Deliberation compels participants to present their arguments in terms of the common interests of the organization. Individual interests are legitimate and publicly defensible only in so far as they may be presented as compatible with, or at least not contrary to, the common interests of participants. This, again, is especially important in knowledge networks, because the problem that must be solved has


323

to be collectively defined or at least significantly redefined by the network itself. In contrast to compromise solutions (as well as to coercion, manipulation, acquiescence, unthinking obedience, or market decisions), deliberation implies justification (Warren, 1996). Through deliberation, participants make collective judgements of discretional character – judgements that are freely and prudently made through debates, driven by reason and good sense, which are not partial or inappropriate. In the cases analysed, most decisions had important deliberative elements. A common way to describe the process through which participants made joint decisions is the following: proposals are elaborated and analysed; arguments are presented and discussed; then, technical support is provided, some tests are conducted and results are compared. Asked how they solve their differences, network participants emphasized the importance of deliberative mechanisms: Differences are solved by seeking more information and studying it more carefully. That is, everybody has to learn how to understand the others . . . We must understand how to measure and evaluate things . . . But this is achieved through studying, and afterward, in some meeting, talking more deeply about mistakes, definitions, and so on. You have to back up your proposals with indicators . . . You have to show the viability of numbers.

Certain norms are indispensable for deliberation. Among them are ‘openness, respect, reciprocity, and equality’ (Dryzek, 2000, pp. 134–5). These norms may be enshrined in explicit rules, but even if they are not, they must be respected in practice. Openness means that several decisions are possible in principle. A measure of respect for partners is indispensable for serious discussion. As Gutmann and Thompson (1996, pp. 52–3) point out, reciprocity refers to ‘the capacity to seek fair terms of cooperation’, And ‘equality of opportunity to participate in . . . decision making’ has been considered ‘the most fundamental condition’ of deliberation (Bohman and Rehg, 1999, p. xxiii). Several structural features of networks facilitate deliberation: the autonomy and interdependence of their members, their decentralized power structures and so on. But even within this favourable structure, deliberation needs specific institutional arrangements. Among the facilitating institutions and practices are: frequent meetings where free, horizontal discussion is not only allowed but actively encouraged; the use of appropriate channels of communication that do not require physical presence (e-mail, virtual meetings); the existence of recognized mechanisms for solving disputes and defining the rules of debate. Mechanisms like these

324


allow networks to embed deliberation in their daily life and to take full advantage of its potential for producing legitimate decisions. Deliberation has several advantages over other forms of collective decision-making. In contrast to voting, where the acceptance of majority decision by the minority is a structural problem, negotiation and deliberation do not create absolute losers. But whereas the main goal of negotiation is to reach a compromise among conflicting interests, the main goal of deliberation is to convince the other partners. Thus collective agreements reached through deliberation are self-enforcing and therefore less vulnerable to unilateral action, which is a weakness of negotiated agreements. Deliberation may reinforce the efficacy of networks. According to Weale (2000, p. 170), under certain conditions, transparent processes based on deliberative rationality must lead to solutions that are functionally efficacious in most cases. This will happen if the solution to a given problem complies with the following conditions: it must arguably belong to the set of those decisions that may be reasonably chosen, even if there were other options that could have been reasonably chosen as well; it must be open to scrutiny by those affected or benefited by it. If this is the case, then negotiation and the pressure for unanimity are irrelevant to the extent that their potential outcomes belong to the set of decisions that may be made through deliberation. However, deliberation has important drawbacks. Agreements often exact a heavy price, as they are usually achieved through long and complicated processes of discussion. Besides, there is always the risk that deliberation may lead to non-decisions (Jachtenfuchs, 2006). Deliberation is not only a time-consuming activity; it also requires energy, attention, information and knowledge, which have been considered scarce deliberative resources (Warren, 1996). Moreover, by stimulating public discussion, deliberation may intensify disagreement and increase ‘the risk that things could go drastically wrong’ (Bell, 1999, p. 73). It may even create disagreement where there was none. It can impede or at least complicate the adoption of rules guiding collective discussion and decision-making, which form the basis for subsequent deliberations. Although it has been considered that collective agreements reached through deliberation are ideal for heterogeneous actors, several interviewees think that some differences among participants were never solved. But this apparent deficit of efficiency is perhaps a structural problem of knowledge networks. Members frequently complain that to reach agreements they have to participate in multiple committees of all sorts and to spend much time discussing in formal and informal settings. This is not surprising, as knowledge networks are characterized by a permanent


325

search for equilibrium between academic and profit-making criteria. This equilibrium can be achieved only through regular and frequent interaction among participants. This interaction is highly demanding: it requires the recognition of others’ perceptions and preferences, a relatively generalized and well-balanced trust, and a good disposition to engage in the deliberative process itself.

NETWORK PERFORMANCE The complexity of networks entails that there are no straightforward criteria to define and evaluate their performance. From their respective standpoints, participants may draw varying and even contradictory conclusions about the achievements and failures of collaborative projects. The performance of knowledge networks has to be assessed at two equally important levels: practical (or functional) and organizational. The former refers to actions and decisions aimed at solving technical problems, transforming or generating knowledge, or creating economically useful products by scientifically valid methods. This level of performance has normative, technical and exchange dimensions. Organizational performance conventionally refers to the survival of the network, but it also concerns its stabilization and the preservation, waste, or increase of the bases and opportunities for future collaborative exchanges between academy and industry. Network organizational performance will be higher to the extent that it facilitates the continued operation of the four integration mechanisms analysed above and provides participants with the knowledge and skills necessary for this complex form of collaboration. Dimensions of Functional Performance The decisions made and actions taken by knowledge networks may be assessed along three dimensions, each with its distinctive question: 1. 2.

3.

Normative dimension. Whether decisions and actions are right: the extent to which they comply with the normative standards of participants. Technical dimension. Whether they are true or accurate: how successful they are in solving the problems that the network is meant to address and in finding correct answers to relevant questions. Exchange dimension. Whether they are profitable: how much they satisfy the interests of all participants and how well they deal with their concerns.

326


Actions and decisions must perform reasonably well in each of the three dimensions simultaneously. A good decision or action should be right, true or accurate, and profitable9. An action that is judged normatively sound but fails to bring about accurate solutions to the problems or profitable results to participants would be practically useless. One that has technically accurate and profitable consequences but violates norms and rules that are fundamental to any of the participant entities may undermine the collaborative project. Moreover, given that these networks bring together people from different institutional settings, each of the three dimensions listed above necessarily comprises a variety of standards. The norms and values held by academic and business organizations are obviously different, and all of them must be taken into account when determining whether a decision or action was right or wrong. Similarly, to determine whether a given decision was correct, it is necessary to consider the definitions of truth and accuracy that prevail in all the participant entities. The same holds for finding out whether the results of an action were profitable, since universities and firms have distinct interests and goals, which result in different views of what a ‘profit’ should be. Equally, or even more importantly, those dimensions also comprise the standards created by the network itself. The relevant norms, the nature of the technical problems to be solved, and the interests and goals of participants are defined, shaped and transformed by means of the interaction itself. To the extent that the interaction crystallizes into a genuinely new entity and becomes autonomous from its sponsors, the network acquires its own performance standards. Functional Performance: Effectiveness, Efficacy, Efficiency The complexity of knowledge networks further implies that effectiveness, efficacy and efficiency, the traditional ways to evaluate performance, also become multifaceted. An agent is usually considered effective to the extent that it is capable of having the intended or expected effects.10 It is clear that, according to their own members, the networks analysed here did have the capacity to produce their intended effect: to transmit, transform and even create knowledge. However, effectiveness, while important, is too vague. It does not evaluate that effect by comparing it to a given standard. Efficacy is more precise. It normally means not only the ‘power or capacity to produce effects’ in general, but, more specifically, also the ‘power to effect the object intended’.11 This is associated with the capacity for achieving precise, previously stated, goals. But according to the interviews, problem-solving in knowledge networks involves a complex


327

process of definition and redefinition of the issues addressed. Therefore network efficacy has to be evaluated dynamically, considering the resolution of problems arisen during the interaction itself, rather than solely the achievement of previously set goals. By itself, efficacy is about achieving goals and says nothing about the costs of doing so. Efficiency is the notion that attempts to fill this void. An agent or action is efficient to the extent that it achieves its goals at the lowest costs (time, technical resources, money, physical effort and so on). It implies a balance between means and ends, between costs and benefits. This notion is critical for evaluating the performance of knowledge networks, since it is clear that assessments will vary greatly according to which combination of criteria and standards is taken into account. Therefore it requires a more detailed attention than the others. Efficiency is an empirical, not a normative, matter. But a potentially efficient decision that violates fundamental normative or legal standards may be unacceptable, unsustainable or counterproductive. Hence a previous step is to determine whether an action or decision complies with the relevant norms (values, laws, rules and so on). After this preliminary probe, its efficiency should be evaluated according to the two remaining dimensions: their truth or accuracy, and their profitability. In the technical sense, networks are efficient in so far as they find accurate solutions to the problems they are meant to address at the lowest possible cost, measured in cognitive, economic, technological and related terms. For knowledge networks, this usually means whether the expected technological product was created or whether scientific knowledge was transformed into economically useful goods. But it also means whether important problems, not originally considered, where found and defined in a creative and potentially useful way. Although sophisticated tools may be used to gather information in this respect, the assessment made by participants themselves might be the most important input for analysing this kind of efficiency. But, as said above, networks are not only problem-solving devices. They are also exchange structures. When organizations and individuals contribute their own resources to the collective effort, they seek actual returns. Therefore network efficiency must also be assessed in distributive terms. Does the collaborative effort yield significant benefits to each participant? Are the products of the interaction distributed in a way that satisfies even the least favoured partner? Are the costs of the interaction fairly distributed among participants? Is there no participant that would be better off without this collaboration? In formal econometric terms, is the network Pareto-efficient – that is to say, is it not possible to organize the exchange in a way that would produce greater benefits to at least

328

Table 12.1


Criteria and dimensions for assessing practical performance

Criteria/ dimension

Normative

Technical

Exchange

Effectiveness

Ability to produce normatively sound results

Ability to produce profitable results

Efficacy

Capacity to achieve normatively sound concrete goals

Ability to produce technically correct results Capacity to solve specific problems

Efficiency

Capacity to solve technical or scientific problems at the lowest possible cost

Capacity to provide practical benefits to participants Capacity to provide tangible benefits to each and all participants

some of the participants without damaging any of them? Again, although sophisticated tools may be used to analyse this kind of efficiency, perhaps the best data come from evaluations made by participants themselves. The important question in this respect is whether participants perceive an acceptable proportion between, on the one hand, the efforts made and the resources invested and, on the other, the results obtained. Table 12.1 summarizes the criteria and dimensions that can be combined to assess the practical or functional performance of knowledge networks. It must be added that such assessment should be comparative. For instance, the capacity of a network for efficiently solving technical problems must be compared to the real or potential capacity of other organizational structures to address the same problems. Would a university laboratory, by itself, have been able to solve those problems at a lower cost? Would the R&D department of the firm have provided the quickest and least costly solutions? The same types of questions should be asked for each of the other criteria and dimensions. Organizational Performance Decisions and instrumental actions are not the only results of networks that merit attention when evaluating their performance. It is equally important to observe whether, in making or undertaking them, the network preserved, or undermined, the opportunities for future collaborative exchanges.


Table 12.2

Organizational performance: standards and conditions for the operation of integration mechanisms

Mechanism

Criteria

Trust

● ●

Translation

● ● ● ●

Negotiation

● ● ●

Deliberation

329

● ● ●

Production of normative, strategic and technical trust Balance among the three kinds of trust Creation of common languages Institutionalization of translation Training of individual translators Diminishing interpretative flexibility Reciprocity: respect for the legitimate particular interests of participants Production of rules for future negotiations Creation of mechanisms and sites for conflict negotiation Equal opportunity to participate in decision-making Definition of collective interests, objectives and problems Creation of institutions for deliberation

As Table 12.2 suggests, this means, in the first place, asking whether the operation of the network satisfied the fundamental standards or ideals of the four integration mechanisms analysed above: trust, translation, negotiation and deliberation. But it also depends on whether it strengthened or weakened the conditions necessary for the operation of such mechanisms. Trust can be reinforced by use – but it may also be destroyed when used improperly. Seemingly efficient solutions may violate some basic principles of fairness, discrediting some of the participants or contravening important values. More subtly, a decision or action may alter the balance among trust dimensions that is necessary for a network to maintain itself and operate efficiently in the medium term. It may, for instance, reinforce strategic trust at the expense of normative or prestige-based trust. Thus the preservation, creation, or destruction of trust is a crucial criterion for evaluating network performance. A further criterion is the degree to which the network created common languages, institutionalized the function of translation, and trained individual translators, thereby facilitating future collaboration. Similarly, the interaction may define collective interests, objectives, problems and solutions in a way that facilitates future deliberation. It can also reinforce norms (equality, respect, reciprocity, openness) and institutions that facilitate deliberation; particularly important is the equal opportunity to participate in decision-making (arguably ‘the most fundamental condition’ of deliberation). Decisions may be made according to the principle of reciprocity, which

330

Table 12.3


Organizational performance: network dynamics

Dimensions

Criteria

Autonomy

● ●

Network development

● ● ●

Learning

●

Self-regulation capacities (organizational autonomy) Self-selection capacities (individual autonomy) Stabilization of the network Organizational learning Creation of new networks Individual acquisition of organizational skills and knowledge

is central to negotiation: respecting the legitimate particular interests of all participants, with a fair distribution of the benefits and costs of collaboration. Similarly, the interaction may reinforce existing structures, spaces, mechanisms and procedures for conflict management and interest aggregation – or it may undermine them. Especially important is the creation of sites (meetings, boards, committees) where conflicts can be negotiated or arbitrated. Practical and organizational results are obviously important. But, in some cases, the best result of collaboration may be the creation and preservation of the network itself as a space for bringing together knowledge from different settings, interests from several organizations, problems and potential solutions useful to a variety of actors. To assess this, one must observe the dynamics of networks (Table 12.3). According to the actor-network theory, such dynamics embraces three main phases: emergence, development and stabilization. New networks emerge out of already existing ones, by either subtle changes or revolutionary breakthroughs. A network can evolve into a more convergent or more divergent structure. When coordination is stronger and different elements are better aligned, the network becomes more stable and predictable. In other words, stabilization, or closure, means that interpretive flexibility diminishes. When its diverse elements are more tightly interrelated, the network becomes more complex and stable, because to disconnect an actor from a network, many other connections have to be undone (Stalder, 1997). Finally, by participating in knowledge networks, individuals and organizations may acquire the skills and knowledge necessary for future collaboration: how to interact with people from other academic or business entities, how to communicate with and learn from them, how to distinguish those that are trustworthy from those that are not, how to encourage them to trust you, how to negotiate and deliberate with them, how to


331

collectively make decisions that are normatively right, technically correct and profitable. But participation may also be a frustrating experience. It may show you how difficult it is to communicate with people from different entities, how unreliable they are, how complicated it is to negotiate with them, how futile are the efforts to rationally exchange arguments with them. As with the previous dimensions and criteria of performance, interviews with participants can be the best means to evaluate this learning dimension of networks.

CONCLUSION The proper functioning of knowledge networks requires the concurrence of four integration mechanisms: trust, translation, negotiation and deliberation. Yet, within certain limits, there is an inverse relationship between two pairs of them: between translation and trust, and between negotiation and deliberation. When there is an optimal combination of normative, strategic and technical trust, translation becomes less necessary; the successful practice of deliberation makes negotiation easier and less salient as a decision-making mechanism. Thus well-functioning knowledge networks should show the following distinctive characteristics: strong and well-balanced trust that facilitates communication among participants, a moderate need for translation, an intense practice of deliberation, and a moderate use of negotiation. A well-functioning network, with a proper combination of the four integration mechanisms, can be expected to have good practical or functional performance – bringing about actions and decisions that are normatively sound, technically true or correct, and economically profitable. They can also be expected to have high organizational performance – preserving and strengthening the conditions for future collaboration. In making effective, efficacious and efficient decisions, well-performing networks should facilitate translation, create or preserve trust, and reinforce the institutional and personal basis for negotiation and deliberation. Similarly, they should become more autonomous and stable and give their members greater opportunities for acquiring the knowledge and skills that this complex form of collaboration requires.

NOTES 1.

The interviews were collectively designed and conducted by participants in a research project on knowledge networks (Luna, 2003).

332 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

The capitalization of knowledge For a further analysis of these findings, see Luna and Velasco (2005). For a broader analysis of this topic, see Luna and Velasco (2003). For further analysis on the relationship between trust and translation, see Luna and Velasco (2006). Half of the interviewees spontaneously said that these decisions were collectively made. The other participants gave diverging replies when asked about the source of such decisions, even when they were referring to the same collaborative project. That is, agreements about which there is no expressed opposition by any participant, or agreements that result from the sum of differences. According to Dryzek (2000, p. 134), ‘the most appropriate available institutional expression of a dispersed capacity to engage in deliberation . . . is the network’. On this topic, see Magnette (2003a, 2003b), Eberlein and Kerwer (2002), and Smismans (2000). According to Weber (2005, p. 51), ‘The efficiency of the solution of material problems depends on the participation of those concerned, on openness to criticism, on horizontal structures of interaction and on democratic procedures for implementation.’ ‘Organizational effectiveness is a prerequisite for the organization to accomplish its goals. Specifically, we define organizational effectiveness as the extent to which an organization is able to fulfill its goals’ (Lusthaus et al., 2002, p. 109). Oxford English Dictionary Online, 2006.

REFERENCES Bell, Daniel. A. (1999), ‘Democratic deliberation: the problem of implementation’, in S. Macedo (ed.), Deliberative Politics: Essays on Democracy and Disagreement, New York and Oxford: Oxford University Press, pp. 70–87. Bohman, J. and W. Rehg (eds) (1999), Deliberative Democracy: Essays on Reason and Politics, Cambridge, MA: MIT Press. Burt, R.S. (1992), Structural Holes, Cambridge, MA: Harvard University Press. Burt, R.S. (2000). ‘The contingent value of social capital’, in E.L. Lesser (ed.), Knowledge and Social Capital. Foundations and Applications, Boston, MA: Butterworth/Heinemann, pp. 255–86. Callon, M. and J. Law (1997), ‘L’Irruption des Non-Humaines dans les sciences humaines: quelques leçons tirées de la sociologie des sciences et des techniques’, in J.-P. Dupy, P. Livet and B. Reynaud (eds), Les limites de la rationalité. Tome 2. Les figures du collectif, Paris: La Découverte, pp. 99–118. Dasgupta, P. (1988), ‘Trust as a commodity’, in D. Gambetta (ed.), Trust: Making and Breaking Cooperative Relations, Oxford: Basil Blackwell. Dryzek, J.S. (2000), Deliberative Democracy and Beyond: Liberals, Critics, Contestations. Oxford: Oxford University Press, pp. 49–72. Eberlein, B. and D. Kerwer (2002), ‘Theorising the new modes of European Union governance’, European Integration on Line Papers (EIoP), 6 (5). Eising, R. and B. Kohler-Koch (1999), ‘Introduction: network goverance in the European Union’, in B. Kohler-Koch and R. Eising (eds), The Transformation of Governance in the European Union, London: Routledge, pp. 3–12. Elster, J. (1999), ‘The market and the forum: three varieties of political theory’, in J. Bohman and W. Rehg (eds), Deliberative Democracy: Essays on Reason and Politics, Cambridge, MA: MIT Press, pp. 3–34. Granovetter, M.S. (1973), ‘The strength of weak ties’, American Journal of Sociology, 78 (6), 1360−80.


333

Gutmann, A. and D.F. Thompson (1996), Democracy and Disagreement, Cambridge, MA: Belknap Press of Harvard University Press. Hage, J. and C. Alter (1997), ‘A typology of interorganisational relations and networks’, in J.R. Hollinsworth and R. Boyer (eds), Contemporary Capitalism. The Embeddedness of Institutions, Cambridge: Cambridge University Press, pp. 94–126. Jachtenfuchs, M. (2006), ‘The EU as a polity (II)’, in K.E. Jorgensen, M. Pollack and B.J. Rosamond (eds), Handbook of European Union Politics, London: Sage, pp. 159–74. Lane, C. (1998), ‘Introduction: theories and issues in the study of trust’, in C. Lane and R. Bachman (eds), Trust within and between Organizations: Conceptual Issues and Empirical Applications, Oxford: Oxford University Press, pp. 1–30. Latour, B. (1993), We Have Never Been Modern, Cambridge, MA: Harvard University Press. Lewicki, R.J. et al. (2004), Essentials of Negotiation, 3rd edn, Boston, MA: McGraw-Hill/Irwin. Leydesdorff, L. (1997), ‘The new communication regime of university–industry– government relations’, in H. Etzkowitz and L. Leydesdorff (eds), Universities and the Global Knowledge Economy, London: Pinter, pp. 106–17. Leydesdorff, L. (2001), A Sociological Theory of Communication: The SelfOrganization of the Knowledge Based Society, Parkland, FL: Universal Publishers. Luna, M. (ed.) (2003), Itinerarios del conocimiento: formas, dinámicas y contenido. Un enfoque de redes, Barcelona: Anthropos/IIS-UNAM. Luna, M. and J.L. Velasco (2003), ‘Bridging the gap between business firms and academic institutions: the role of translators’, Industry and Higher Education, 17 (5), 313−23. Luna, M. and J.L. Velasco (2005), ‘Confianza y desempeño en las redes sociales’, Revista Mexicana de Sociología, 67 (1), 127−62. Luna, M. and J.L. Velasco (2006), ‘Redes de conocimiento: principios de coordinación y mecanismos de integración’, in M. Albornoz and C. Alfaraz (eds), Redes de conocimiento: construcción, dinámica y gestión, Buenos Aires: UNESCO, pp. 13–38. Lundvall, B.-Å. (2000), ‘Understanding the role of education in the learning economy. The contribution of economics’, in OECD (ed.), Knowledge Management in the Learning Society. Education and Skills, Paris: OECD, pp. 11–35. Lusthaus, C. et al. (2002), Organizational Assessment: A Framework for Improving Performance, Washington, DC and Ottawa: Inter-American Development Bank and International Development Research Centre. Magnette, P. (2003a), ‘European governance and civic participation: beyond elitist citizenship?’, Political Studies, 51 (1), 1−17. Magnette, P. (2003b), ‘In the name of simplification. Constitutional rhetoric in the convention on the future of Europe’, Working Paper, Bruselas: Institut d’études européennes, Université Libre de Bruxelles. Martinelli, A. (2002), ‘Markets, governments, communities, and global governance’, paper presented at the XV World Congress of Sociology, Brisbane, Australia, 7−13 July. Messner, D. (1999), ‘Del estado céntrico a la sociedad de redes. Nuevas exigencias

334


a la coordinación social’, in N. Lechner, R. Millán and F. Valdés (eds), Reforma del estado y coordinación social, Mexico City: Plaza y Valdés, pp. 77–121. Oléron, P. (1983), L’argumentation, Paris: PUF. Sable, C. (1993), ‘Studied trust: building new forms of cooperation in a volatile economy’, Human Relations, 46 (9), 1133−70. Smismans, S. (2000), ‘The European Economic and Social Committee: towards deliberative democracy via a functional assembly’, European Integration on Line Papers, 4 (12). Stalder, F. (1997), ‘Latour and actor-network theory’, paper available online at http://amsterdam.nettime.org/Lists-Archives/nettime-l-9709/msg00012.html. Steward, F. and S. Conway (1996), ‘Informal networks in the origination of successful innovations’, in R. Coombs et al. (eds), Technological Collaboration: The Dynamics of Cooperation in Industrial Innovation, Cheltenham, UK and Brookfield, MA, USA: Edward Elgar, pp. 201–21. Streeck, W. and P.C. Schmitter (1992), ‘Comunidad, mercado, Estado ¿y las asociaciones? La contribución esperada del gobierno de intereses al orden social’, in R. Ocampo (ed.), Teoría del neocorporativismo: Ensayos de Philippe Schmitter, Mexico City: UIA/ Universidad de Guadalajara, pp. 297–334. Valente, T. (1995), Network Models of the Diffusion of Innovation, Cresskill, NJ: Hampton Press. Walker, G., B. Kogut and W.-J. Shan (1997), ‘Social capital, structural holes and the formation of an industry network’, Organization Science, 8 (2), 109−25. Warren, M.E. (1996) ‘Deliberative democracy and authority’, American Political Science Review, 90 (1), 46−60. Weale, A. (2000), ‘Government by committee. Three principles of evaluation’, in T. Christiansen and E. Kirchner (eds), Committee Governance in the European Union, Manchester: Manchester University Press, pp. 161–70. Weber, S. (2005), ‘Network evaluation as a complex learning process’, Journal of MultiDisciplinary Evaluation, 2, 39−73.

Index Abernathy, W. 300 Abramson, N. 294, 297 academic and industrial researchers, different approaches of 53–4 academic life see polyvalent knowledge and threats to academic life academic software and databases 181–4 academic–business research collaborations (and) 74–97 organizational ecology of innovation 88–92 productive shocks and problematic tensions 77–81 social interactions and collaborative research 82–7 university patenting for technology transfer 74–7 see also innovation; legislation academy–industry relations 2–3, 21, 25, 46–60 effects of different deontic knowledge on 46 incentives for 32 obstacles to 45 transformation of 204–5 see also background knowledge; norms acronyms proprietary, local, authoritarian, commissioned and expert (PLACE) 47, 52 strategic, hybrid, innovative, public and scepticism (SHIPS) 47, 52 ACT-R (Adaptive Control of Thought–Rational) networks 40 agro-food 144, 151–3 Agrawal, A. 5, 6, 46 ahead-of-the-curve science 146

Ainsworth, S. 244 Alcoa 32 Allen, R.C. 170, 171, 173 allocative efficiency theory 125 Alter, C. 315 Amable, B. 227 Amadae, S.M. 264 analytical ontic knowledge 33–42 cognitive mode of 39–42 descriptive 33–4 explanatory 34–9 see also Appert, Nicholas; Pasteur, Louis Anderson, J.R. 40 Anderson, P. 300 anti-commons 181 effects 80, 122 tragedy of 18, 20, 134–5 anti-cyclic triple helix see capitalizing knowledge and the triple helix; polyvalent knowledge and threats to academic life; publishing and patenting; triple helix in economic cycles; triple helix: laboratory of innovation Anton, J.J. 171 Antonelli, Cristiano 16, 99 Aoki, M. 226–7 Appert, Nicholas 36–9, 41, 43, 62 and food preservation 36–7, 38–9 appropriability, impact of 36 appropriability, opportunities and rates of innovation (and) 121–42 blurred relations between appropriability and innovation rates 130–136 failures of ‘market failure’ arguments 123–8 growth in patenting rates and (mis) uses of patent protection 128–30

335

336


opportunities, capabilities and greed 136–9 Arora, A. 136 Arrow, K.J. 88, 99, 121, 169, 265, 266–7 Art of Computer Programming 168 Arundel, A. 68 Ashby, W.R. 307 AT&T 32 Aventis 151 Avnimelech, G. 113 Ayer, A.J. 264 Azoulay, P. 6, 7 background knowledge (and) 45, 47–60 groups 53–4 shared vs unshared 64–5 ‘teamthink’ theory 53 see also cognitive rules; language; norms; research; studies (on) Bagchi-Sen, S. 143, 152, 155 Balconi, M. 41, 64, 75 Baltimore, David 148 banks and polyarchic decision-making 101–2 Bar Hillel, M. 57 Barnet, G. 206 Barré, R. 223, 233 Barsalou, L.W. 40, 44 Barton, J. 135 Bas, T. 146, 148, 159 Bathelt, H. 293 Baumgartner, F. 244 Bayer 151 Bazdarich, M. 151 Becker, G.S. 273 Beffa, J.-L. 234 Bell, Daniel A. 324 Bell Labs 125 Bernstein, R.J. 264 Bessen, J. 135 Beston, A. 243 Beth, E. 44 big pharma 153–4, 155, 156 ‘big science’ projects 63 Bilderbeek, R. 305 biopharma clusters 19 biopharmaceuticals 18, 143–8 , 151–3, 250

Biopolis (Singapore) R&D hub 160 bioprocessing 160 bioregional clusters 157 biosciences globalized 152–5 and knowledge networks 153, 157–8 new research institutes for 160 vehicles for implantation of 153–5 Biosciences Policy Institute 252 see also Life Sciences Network (LSN) Biosystems Genomics, Centre for 153 biotechnology 41, 147–9, 152–3 and cluster theory 144–6 clustering 149–50 collaborative aspect of 148 and dedicated biotechnological firms (DBFs) 19, 144–6, 153 in Germany 19 see also Germany globalization of 150–152 initiatives (Cambridge University) 2 and nanabiotechnology 160 networked hierarchy in 155–61 sociocultural impact of (in New Zealand) 247–8 see also cluster theory; Schumpeter, J.A. biotechnology firms Cambridge (UK) 148 Perkin-Elmer (USA) 149 Birdzell, L.E. 34, 35, 36, 61 Blau, P.M. 295 Bloch, C. 115 Blumenthal, D. 4, 206, 208, 209 Bohman, J. 323 Boltanski L. 219, 227 Borras, S. 234 Boulding, K.E. 266 boundaries 244–5 mental 245 setting 247 social 245 symbolic 245 boundary organizations: their role in maintaining separation in the triple helix (and) 243–60 boundaries 244–5 boundary work 245–7 Life Sciences Network (LSN) 248–58 research project 247–8

Index boundary work 245–7 activities 246 and demarcation 247 Bowie, J. 143 Bozkaya, A. 114 Braczyk, H.-J. 291 Braine, M.D.S. 44 Branciard, A. 233 Bratko, I. 306 Brenner, Sidney 160 Breschi, S. 8 British Petroleum (BP) 153 Broesterhuizen, E. 47 Buck, P. 272 Burt, R.S. 313, 318 Bush, V. 276 Calderini, M. 8 Callon, M. 226, 227, 314 Cal Tech 208 Cambridge (UK) 146, 148, 150, 151 Cambridge University 2, 159 Cambridge–Boston 159, 162 and global bioscientific network hierarchy 158 Camerer, C. 68 Campart, S. 115 Campbell, E.G. 4 Canada biotechnology in 150–151 bioregional clusters in 157 capability clusters 139 capability-based theory of the firm 138 Capitalism, Socialism and Democracy 102 capitalism, varieties of 218 capitalizing knowledge and the triple helix 14–25 see also triple helix in economic cycles; triple helix: laboratory of innovation Carlsson, B. 143, 291 Carnegie, Andrew 31 Carruthers, P. 39 Carson, C.S. 272 Carter, A.P. 291 Casper, S. 146, 148, 150, 232 causal reasoning 44, 57–8 Celera 182

337

Centennial Exhibition, Philadelphia (1876) 134 Chandlerian model of innovation 116 Chase Econometrics 183 chemicals complex, Grangemouth, Scotland 152–3 Cheng, P.W. 44 Chesbrough, H. 113, 153 Chiappello, E. 227 Childhood and Neglected Diseases, Institute for (ICND) 155 Clark, H.H. 50, 51, 52, 65 Clark, K.B. 300 Cleeremans, A. 45 cluster penetration for open innovation 154 cluster theory 144–6 cluster-building 19, 147 Coase, R. 105, 123 and Coase Theorem 76 Cobbenhagen, J. 297 Coenen, L. 151, 153 cognitive rules (and) 54–60, 64 application of 45 decision-making 58–60 differences in 66 intuitive and analytical thinking 54–5 problem-solving 55–6 reasoning 56–8 cognitive style, shared vs unshared 65–7 Cohen, J. 67 Cohen, W. 18, 132–3, 136, 174, 306 Colbertist model 223 Collins, A.M. 4044 Collins, S. 174, 250, 253, 255–6 Colman, Alan 160 communication 53–4 and information 276–7 linguistic 64–5 non-verbal 51–2 theory 269–70 Complex Adaptive Systems (CAD) 25 Computer Science, National Institute of (INRIA) 233 Conway, S. 318 Cook-Deegan, R. 90 Cooke, P. 18, 143, 145, 159, 292, 293 Cookson, J.E. 272

338


copyleft license see Generalized Public License (GPL) copyright 17, 32 see also patents/patenting cotton research centre (Raleigh– Durham) 152 Cowan, R. 64, 291 Craig Venter 149 creative destruction 101, 146–7 Crespi, G.A. 75, 76 Crystal Palace Exhibition, London (1951) 133–4 da Vinci, Leonardo 31 Dahlman, C.J. 105 Dalle, J.M. 167 Dasgupta, P. 4, 172, 292, 316 Data Resources, Inc 183 Davenport, Sally 23 David, P.A. 4, 5, 126, 170, 172, 182, 191, 291, 292 De la Mothe, J. 143 De Liso, N. 110 de Solla Price, Derek J. 277 Dealers, National Association of 113 Deane, P. 272 Dearborn, D.C. 56 Debreu, Gérard 265 decision-making (and) 58–60, 78, 101, 264, 277, 312, 315, 329, 331 centralization of 223 collective 320–325 bargaining/negotiation 320, 321–2 deliberation 320, 322–5 voting 320 hierarchical 100–101, 103 irrational escalation 59 polyarchic 100–101 rational 277 regret and loss aversion 58–9 risk 58 temporal discounting/‘myopia’ 59–60 dedicated biotechnological firms (DBFs) 19, 144–8, 150, 152–3, 155, 160, 162 exploitation-intensive 156 financing of 156 Defense Research Projects Agency (DARPA) 11

and SUN, Silicon Graphics and Cisco 11 definition of knowledge as production and distribution 268 territorial economy 293 Denison, E.F. 272 DiMinin, A. 7 dogmatism 48, 56–7 Dolfsma, W. 24, 295, 302 Dosi, Giovanni 17, 127, 136, 137, 138 dot.com bubble 147 Dow 151 Dryzek, J.S. 323 dual careers 21 Duchin, F. 277 due diligence 85 DuPont 31 Eastman Kodak 31 Economic Co-operation and Development, Organisation for (OECD) 2, 144, 218, 219, 228, 230, 234, 262, 291, 295, 296, 297, 305, 306 STI Scoreboards of 306 econometric software packages (case study) 184–9 EasyReg 186–7 price discrimination in 189 protection of 185 statistics for 186 education 21, 23, 263, 273–6, 279 categories/sources of 267 economics of 262 higher 225, 230, 239 productivity of 275 Eisenberg, R.S. 122, 135, 180 Eising, R. 314 Elster, J. 59, 320, 322 Endres, A.M. 272 entrepreneurial science 201–2, 207 origins of 206 entrepreneurial scientists (and) 201–17 academic entrepreneurial culture 212–14 academic world 201–2 forming companies 202–5 industrial penumbra of the university 209–11

Index normative impetus to firm foundation 211–12 research funding difficulties 207–12 and transformation of industry– university relations 204–5 Epicyte 144–6 patent-holder: Scripps Research Institute 145 and Plantibodies: patents for DNA antibody sequences 144, 145 Ernst & Young 231 Espiner, G. 248 Etzkowitz, Henry 3, 6, 13, 21, 60, 201, 203, 207, 210, 215, 220, 226 European Commission 78, 183, 234, 291, 305 funding of SESI project 239 European Environmental Agency 247 European Framework Programmes 2 European NACE codes 295 European Patent Office (EPO) 75–6 European Research Area (ERA) 77, 78 European Summit (Lisbon, 2000) 291 European Union (EU) 218, 219, 239 and admin system of European universities 77 EU-15 297 knowledge ecology of 89 legislation (Bayh–Dole-inspired) 79 member states 74, 296 publications on measurement of knowledge 280 universities 91 Eurostat 291, 296, 303 Evans, R. 247 externalities, lack of 123–4 Eymard-Duvernay, F. 219 Fabrizio, K.R. 7 Faulkner, W. 205 feedback positive processes of 90–91, 115 from reflective overlay 24 from reflexive overlay 293 Feldman, M. 146, 147, 148, 150 Feldman, Stuart 15, 79 finance and innovation (and) 99–104 Type 1 and/or Type 2 errors 100– 101, 104 see also innovation

339

Finegold, D. 155, 160 First Industrial Revolution 31, 34, 35, 40, 41, 43 Fischhoff, B. 58 Fisher, D. 250 Fondazione Rosselli 66 Foray, D. 171, 173, 232, 261, 262, 291, 292 Fourquet, F. 272 France/French 228 Grands corps de l’Etat 224 Grandes écoles 224 and Grenoble innovation complex 146, 234 RDI policies (and) 219, 232–4 Grenoble technological district 234 Guillaume Report 232–3 and Sophia Antipolis 2 and state as entrepreneur system 232–4 Francis, J. 146, 147, 148, 150 Franzoni, C. 8 Freeman, C. 146, 280, 292 Friends of the Earth 250, 251 Fritsch, M. 292, 301, 306 Fuchs, G. 143 Gallini, N. 132 Gambardella, Alfonso 20, 21 GE Plastics 153 Geiger, R. 203 Geiryn, T. 246, 247 Gene Technology Information Trust see Genepool; Life Sciences Network (LSN) General Electric 32 Generalized Public License (GPL) 21, 168–9, 183, 186, 188, 197–8 as coordination device 175–7 lesser (LGPL) 190–191 and nature and consequences of GPL coordination 177–8 GenePool 248, 253 Genetic Modification, Royal Commission into (RCGM) 244, 250–252, 258 genetic modification (GM) 248–58 and anti-GM activists 244 role in New Zealand 253–4

340


see also Life Sciences Network (LSN) genetically modified organism (GMO) 23, 251 genomics 62, 154, 160, 233 Centre for Biosystems Genomics 153 functional 155 Genomics Institute of Novartis Research Foundation (GNF) 155 Joint Center for Structural Genomics (JCSG) 155 Gentner, D. 45 Georghiou, K. 229, 230 Germany/German 2, 32, 102, 152, 219–20, 228, 297 BioM 150 BioRegio programme in 150 biotechnology in 19, 149 Federal Ministry of Education and Research 231, 232 ICT sector 232 professional networks/university entrepreneurship 230–232 RDI policies/regime 219, 228, 231–2 revolution in organic chemistry in 32 university model 21 Geuna, A. 68, 174 Gibbons, M. 4, 225 Giere, R.N. 39 Gieryn, T. 215 Gilbert, Walter 148 Glaxo 143 global bioregions: knowledge domains, capabilities and innovation system networks 143–66 ‘Globalization 2’ 161, 162 globalization in biosciences see biosciences Godin, B. 23, 262, 268, 274, 277, 280, 291, 306 Goffman, E. 50 Gompers, P. 111, 113 Grandstrand, O. 126 Granovetter, M.S. 313, 318 Greenpeace 250, 251 Gregersen, B. 148 GREMI group 150 Grice, H.P. 54 Griliches, Z. 272, 275

Grossman, M. 45 growth accounting 23, 263, 271–2, 278–9 Guillaume, H. 232 Guston, D. 243, 245, 258 Gutmann, A. 323 Hage, J. 315 Hall, B. 128, 139, 134 Hall, B.H. 20, 21, 100, 164, 171, 174 Hall, P. 148, 218, 227 Hansen, W.L. 273 Harhoff, D. 171, 173 Harvard Medical School (HMS) 159, 162 Harvard University 1, 148, 202, 206 Haseltine, Walter 148 Hawkes, J. 250, 255, 256 Hayek, F.A. 263, 265–6 Health, National Institutes of (NIH) 211 research funding 156–7 Heims, S.J. 269 Heller, M.A. 122, 135, 180 Hellstrom, T. 243 Henderson, R. 5, 6, 46, 149 Hendry’s model development methodology 186 Heracleous, L. 243 Hernes, T. 243, 244–5 Hertzfeld, H.R. 174 Hewlett, Bill 210 Hewlett-Packard 228 Hiatt and Hine 145 Hicks, John 265 Hickson, D.J. 295 higher education institutions (HEIs) 82, 83–4, 242 Hilaire-Perez, L. 171, 173 Hinsz, V. 53 Hodgson, G.M. 98 Holyoak, K.J. 44 Holyoke, T. 244 Hoppit, J. 272 Hortsmann, I. 130 Hounshell, D.A. 267, 270 human genome decoding/Celera 147–8, 149 human genome mapping 63 Hurd, J. 151

Index Hussinger, K. 111 hybrid organizations 22, 32, 41–2, 60, 64, 66–7, 98, 226, 243 Incyte 182 information and communication technology (ICT) 5, 41, 98, 125–6, 147–9, 151, 229, 232, 233, 261, 264 core technologies of 18 emergence of 125–6 revolution 64, 116 imagery 9 incentives for knowledge production with many producers 169–72 and granting of IPR 170 intellectual curiosity 170 open science regime 170–171 and system for ‘collective invention’ 171 Industry and Trade 88 initial public offerings (IPOs) 12, 112, 113–14 innovation(s) 1–3, 9–20, 22, 31–9, 42–4, 61–2, 67–8, 88–92 bundling finance and competence with 111 and finance 99–104 and imitation, model of 127 and innovative activity 88 national systems of 291 oilseed rape 152 patent 5, 125 policy response 88–9 imitation of patent-protected 131 opportunities for 137–8 protection of 132–3 profiting from 132 rates of 122–3 and RDI policies 218–42 see also open innovation innovative opportunities 125, 127–8, 137 intellectual property (IP) 4 costs of protecting 134–5 lawyers 135 monopolies 81, 126 ownership of 77 protection 79–80, 124–6 and appropriability 124–5

341

and royalty stacking 80 weak regime of 126 intellectual property rights (IPR) 15–18, 22, 32, 80, 121–3 capitalization of knowledge through 22 and corporate strategies on legal claim of 129 enforcement strategies 127 fragmentation of 183 influence on knowledge 130–131 and innovation 18 and joint R&D ventures 86 knowledge-intensive 111–13 -leads-to-profitability model 138 legal protection of 123 markets for technologies/rates of innovation/diffusion 136 and modes of appropriation 121 and new technologies 126–7 as obstacle to research/innovation 18, 80 over-the-county (OTC) offerings of 112 protection and income distribution 122–3 protection and innovation 135–7 regimes changes in 132 lack of effects of different 133 strengthening of 137–8 -related incentives 139 and technology transfer officers (TTOs) 21 university access to 173–4 see also markets interaction and discussion between academics and business 19 investment on research and innovative outcomes 134 Isabelle, M. 170 IT industry 228 Itai, S.244 Italy 75 patenting and publishing study 8–9 and Piedmont regional government 2 Politecnico of Turin 2 Jachtenfuchs, M. 320, 324 Jackendoff, R. 44

342


Jackson, S. 53 Jacob, M. 243 Jaffe, A. 128, 133, 137 Jakulin, A. 306 Janus scientist(s) 14, 22, 25, 32 and proximity 60–68 Jensen, R. 5 job losses, UCLA Anderson Business School report on 151 Jobert, B. 235 Johannisson, K. 279 Johnson, B. 227 Johnson, J. 146 Johnson-Laird, P. 39, 44, 57 Jorgenson, D.W. 272 Kaghan, W. 206 Kahneman, D. 45, 54–5, 57, 58 Kaiser, R. 146, 149, 150 Karamanos, A. 146, 148, 150 Karolinska University/Institute 11, 159 Kauffman, S. 38 Kaufmann, D. 146, 149 Kay, L.E. 269 Kelly, S. 243, 244 Kendrick, J.W. 272 Khalil, E.L. 294 King, R.G. 102 Kista science park (Stockholm) 151 Klevorick, A. 136, 137 knowledge analytical ontic 33–42 base of an economy 24 as coordination mechanism 24, 292 creation of 268 defining 24 defining as production and distribution 268–9 deontic (procedural) 39–40, 42, 46 different deontic 46 as economic good 99 elements of: education, R&D, communication, information 267–8 fountain-heads 146 generation 124 governance 98–120 and identification with information 125 and IPR regimes 130–131

management 155 measurement of 279–80 ontic (declarative) 39–40, 42, 64 operational elements of 23 operationalization of 277 problem-solving (of firms) 138 production and control 292 productive 125 public domain (PD) 173, 174 public-good in private-good patent 124 spillover 155 tacit 64 transfer processes 91 see also analytical ontic knowledge; venture capitalism Knowledge: its Creation, Distribution and Economic Significance 278 knowledge-based economy 291 knowledge capitalization 65 and norms 43 software research 20 knowledge economy 261–90 flow of ideas (Appendix 1) 286–8 indicators on knowledge-based economy (Appendix 2) 289–90 and knowledge measurement 270–275 see also Machlup, Fritz knowledge networks: integration mechanisms and performance assessment 312–34 as complex systems 313–16 trust and translation 316–20 negotiation and deliberation 320–325 network performance 325–31 see also decision-making; networks; trust knowledge production, (Gibbon’s) Mode 2 of 4, 225 knowledge production see incentives for knowledge production with many producers knowledge transfer 14–15, 19, 43, 53, 64–5, 67, 91, 153, 157 epistemological and cognitive constraints in 32–42 and analytical ontic knowledge 33–42 global 171

Index and hybrid organizations 42 technology 79 knowledge-based development 10 knowledge-driven capitalization of knowledge 16, 19–20, 31–73 see also academy–industry relations; background knowledge; knowledge transfer; ‘nudge’ suggestions to triple helix knowledgeable clusters 161, 162 Knudsen, C. 265 Knuth, Donald 168 Kohler-Koch, B. 314 Koput, K. 143 Kortum, S. 128, 129, 130 Kossylin, S.M. 45 Krimsky, S. 206, 213 Kuan, J. 181 Kuhn, T. 54 Kuznets, S. 38, 279 Laafia, I. 291, 295, 297 Lakatos, I. 5 Lam, A. 225, 227, 230 Lambert Review 82, 86, 87 Lamberton, D. 266 Lamont, M. 219, 244, 245, 247 Lanciano-Morandat, C. 22, 227, 234 Landes, D. 121 Lane, C. 316 Lane, D.A. 99 Langer, E. 58 Langlois, R.N. 262–3 language (and) 32–3, 50–60, 314, 317–19, 329 command 168, 184–5 co-ordination difficulties 52 mathematical 33, 44 natural/formal 39–40 non-verbal communication 51–2 technology transfer (TT) agents 52 Larédo, P. 220, 232 Latour, B. 212, 314 Law, J. 314 Lawton Smith, H. 143, 152, 155 learning 110, 139, 178 collective 233 -curve advantages 132 by doing 265 implicit 45

343

interactive 68, 109 and rules 45 Leder and Stewart patent 18, 131 Lederberg, Joshua (Nobel Laureate) 213 Lee, C. 145 Lee, Y.S. 170 Leech, B. 244 legislation and antitrust litigation 125 Bayh–Dole Act (USA) 74, 76, 78, 79 Danish Law on University Patenting (2000) 79 Employment Retirement Income Security Act 113 European Directive on patenting of software in Europe 182–3 Leitch, Shirley 23 Lengyel, B. 306 Leontief, W. 277 Lerner, J. 111, 113, 127, 128, 129, 130, 133, 175, 178, 190 Lesson, K. 278 Levin, R. 18, 132, 136 Levin, S.G. 4 Levine, R. 102 Levinthal, D.A. 178, 306 Lewicki, R.J. 321 Leydesdorff, L. 24, 220, 206, 243, 292, 293, 294, 301, 306, 307, 314, 318 Life Sciences Network (LSN) 23, 244, 248–58 campaign strategy for RCGM 251–2 Constitution of 248 evolution into Biosciences Policy Institute 252–3 and Genepool 248 membership of 250 objectives of 249 Linnaeus’ natural taxnomy 62 Liu, Edison 160 loss aversion 58–9 and irrational escalation 59 Luhmann, N. 307 Luna, Matilde 25 Lundvall, B.-Å. 146, 227, 234, 291, 292, 318

344


McGill, W.J. 294 Machlup, Fritz (and) 23–4, 126, 261–80, 292 Information through the Printed World 278 measuring knowledge/national accounting 270–275 policy issues 275 The Production and Distribution of Knowledge 262, 278 sources of insight 279–80 see also Machlup’s construction Machlup’s construction 263–70 and communication teory 269–70 defining knowledge 264–7 ‘operationalizing’ knowledge 267–70 McKelvey, M. 143 Mackie, J. 57 Maillat, D. 150 Mansfield, E. 5,131 Manz, C.C. 53 March, J.G. 60, 178 markets 104–17 classification of 106–7 as devices for reducing transaction costs 105 as economic problem 104–6 functions of 107–10 as coordination mechanisms 108–9 as risk management mechanisms 109–10 as selection and incentive mechanisms 108 as signaling mechanisms 108 and market failure arguments 123–8 and neoclassical theory of exchange 105–6 public capital – focused on IPOs 113–14 as social institutions 105 for technologies and IPR 136 trading knowledge-intensive property rights in 112–13 Marengo, Luigi 17 Markman, A.B. 45 Markusen, A. 148 Marschak, J. 266, 269 Marshall, Arthur 88 Martinelli, A. 314

Marx, K. 121 Maskin, E. 135 Matkin, G. 204 Maurer, S.M. 167, 182, 184, 188 Maurice, M. 219 Maxfield, R.R. 99 measuring knowledge base of an economy in terms of triple-helix relations (and) 291–311 combination of theoretical perspectives 293–4 methods and data 294–8 data 294–5 knowledge intensity and high tech 295–6 regional differences 297 methodology 297–8 regional contributions to knowledge base of Dutch economy 301–2 results 298–300 descriptive statistics 298–9 mutual information 299–300 sectoral decomposition 303–6 Mendeleev’s periodic table of elements 62 Menard, C. 98, 106 Merges, R. 129, 131 Merton, R. 42, 47, 170, 221 Messner, D. 319, 321 Metcalfe, J.S. 15, 88 Meyer-Krahmer, F. 227, 231, 232 Miles, I. 305 Miller, C. 243, 247 Miller’s magical number 40 Mincer, J. 273 Mirowski, P. 292 Mirowsky, P. 269 MIT (Massachusetts Institute of Technology) 6, 158–9, 171, 202, 206, 210–211 MIT–Stanford model 21 Mitroff, I.I. 47 modus tollens 45 Mokyr, J. 31, 38, 62 Molnar, V. 244, 245, 247 Monsanto 151, 248, 253 Moore, G.E. 42 Moore, K. 243, 246–7, 255, 256 Moser, P. 134 Motion, J. 248

Index Mowery, D. 32, 34, 62, 74, 129 multi-level perspectives see research and development (R&D): national policies Mustar, P. 220, 232 Mykkanen, J. 272 NACE categories 295 NASDAQ 17, 146 regulation (1984) and ‘Alternative 2’ 129 and venture capitalism 110–116 national accounting 24, 261, 262–3, 272–3, 275, 278 –9 and System of National Accounts 262, 272–4 National Science Board 207, 280 National Science Foundation (US) 39, 41, 268, 273–4, 278 grant 211 Science Indicators 280 National Venture Capital Association 1 Neck, C.P. 53 Nelson, R.R. 90, 99, 131, 146, 169, 274, 279, 291, 300 Nesta, L.174 Nestlé 151 Netherlands 153, 294–306 Chambers of Commerce of 294 geographical make-up of 297 regional contributions to knowledge base of economy of 301–2 sectoral decomposition of 303–6 and Statistics Netherlands (CBS) 296 network performance 325–31 dimensions of functional 325–6 functional: effective, efficacy, efficiency 326–8 organizational 328–31 networks 153, 157–8, 312–120 ACT-R 40 and actor-network theory 314 as complex systems 313–6 dynamics of 330 as social coordination mechanisms 314 and social network analysis 313 neuroeconomics 68

345

New York Times 79 New Zealand 247–58 Association of Crown Research Institutes (ACRI) 248, 254, 255 biotechnology in 247–8 and Biotenz 250 Foundation for Research Science & Technology 247–8 King Salmon 248 Life Sciences Network in 23 role of GM in 253–4, 257 and Royal Commission on Genetic Modification 243, 244, 250–252 and Royal Society of New Zealand (RSNZ) 248, 254, 255 Newcastle University (and) 1 professors of practice (PoPs) 12–13 Regional Development Agency 12 researchers of practice (RoPs) 12–13 Nguyen-Xuan, A. 45 Niosi, J. 143, 146 148, 159 Nisbett, R.E. 44, 45, 66 Nohara, H. 227, 234 Nooteboom, B. 46 norms 25, 42–4, 47–8, 60–61, 64, 66, 67, 90, 170–171, 173, 177, 182, 187, 190, 211, 245, 316, 323, 326–7, 329 disclosure 80–81 institutional 60, 77 Mertonian 47 operational 48, 50, 52, 55, 56, 58, 60, 68 and pragmatic schemes 44 priority 22 scientific 6 social 44, 48, 64, 178 of universalism vs localism 55 and social value 21 technical 43–4, 63–4 North, Douglass 60 Novartis 152, 161 Genomics Institute of the Novartis Research Foundation (GNF) 155 Institutes for Biomedical Research (NIBR) 155 Nozick, R. 59

346


Nuvolari, A. 173 ‘nudge’ suggestions to triple helix 60–68 generality vs singularity 61–2 complexity vs simplicity 62–3 explicitness vs tacitness 63–4 shared vs unshared background knowledge 64–5 shared vs unshared cognitive style 65–7 see also Janus scientist(s) ‘nudging’ capitalization of knowledge 14 Obama Administration 2 and programme for green technologies 63 OEU 91 Oléron, P. 322 Olson, Mancur 174, 175 and theory of collective action 21 Open Collaborative Research Program 78–9 Cisco Systems 78 Hewlett-Packard 78 IBM 78 Intel 78 US universities 78 see also research open innovation 154 and ‘Globalization 2’ 161, 162 open science 20, 80–81, 91–2, 126, 162, 169, 170, 171, 178, 188, 221 open source production complementary investments in 179–81 instability of 174–5 Oresund (Copenhagen/southern Sweden) 9 organizational re-engineering 77–8 Orsenigo, L. 67, 143, 144 Owen-Smith, J. 144, 170, 235 Packard, Dave 210 apon, P. 223, 233 Parkes, C. 151 Pasquali, Corrado 17 Pasteur, Louis 36–9, 62 and function of bacteria in food preservation 36–7, 38, 39

patents/patenting 5, 6–9, 18, 32, 41, 64, 125–6, 128–30, 144–5 applications from US corporations 128 fences 132–3 growth in applications for 128–9 imitation costs of 131 increase in rates of 133 and innovation 132 ‘onco-mouse’ (Leder and Stewart) 18, 131 private-good 124 and reasons for not patenting 132 and ‘regulatory capture’ 130 Selden 131 strategic value of 129 transfer 65 uses of 132 wildcat 134 and Wright brothers 131 Paulsen, N. 243 Pavitt, K. 137, 295 Penrose, E. 162 Perez, C. 98, 292 Perkin-Elmer 149 Personal Knowledge 265 Pfister, E. 115 Pfizer 143, 288 pharma, corporate 152 pharmaceuticals 228 Phillips, C. 152 Piaget, J. 44 Polanyi, M. 23, 221, 265 Politzer, G. 45 Polyami (artificial fibres) 153 polyvalent knowledge and threats to academic life 3–5 and IP protection for research 4 Porat, M.U. 262, 274 Porter, A.L. 300 Porter, M. 226 and Porterian clusters 19 Powell, W.W. 143, 144, 170, 201, 235 Pozzali, A. 25, 32, 42, 64 , 68 Principles of Economics 88 Production and Distribution of Knowledge in the United States, The 261, 262, 270, 278

Index proprietary vs public domain licensing: software and research products 167–98 academic software and databases 181–4 econometric software packages (case study) 184–9 Generalized Public License (GPL) 175–8 incentives for knowledge production with many producers 169–72 open source production 179–81 public domain vs proprietary research 172–9 prospect theory 58–9 proximity 8, 14, 19, 24, 32, 60–68, 145, 146, 230, 245, 293 geographical 221 and functional 155–6 public domain vs proprietary research (and) 172–9 configuration of open source equilibrium 172–4 Generalized Public Licence (copyleft) as coordination device 175–7 instability of open source production 174–5 nature and consequences of GPL coordination 177–9 public research organizations (PROs) 89, 90–91, 170 and ‘common-use pools’ of patents 80 publishing bioscientific 157–60 collaborations 154 and patenting 5 complementarity between 6–9 in Europe 7–8 Pugh, D.S. 295 Quéré, M. 99 Quillian, M.R. 40, 44 Quinn, S. 209 R&YD 263 Raagmaa, G. 160 Racal 228–9 Rahm, D. 205

347

Rallet, A. 156 Ranga, L.M. 8 Reber, A.S. 45 regional innovation organizer (RIO) 9 Rehg, W. 323 Reingold, N. 276 Renfro, Charles 184, 185, 187 Republic of Science 22, 81, 167, 170, 190, 220, 221–3, 228, 234, 235 research 3–5, 6, 61–2, 161–2, 171–2 academic 66 collaboration 46, 65, 82–7 as costly but fundamental 57 funding 156–7, 207–9 funding for academic 49–50 globalized 155 industrial 49 as investment 24 IP protection for 4 IP protection and innovation 133–4 by National Institutes of Health 147 Open Collaborative Research Program 78–9 productivity of 276 proprietary (PR) 173, 174 public domain vs proprietary 172–9 publication collaboration 158 and Report of Forum on Universitybased Research (EC, 2005) 78 scientific 41 on sociocultural impact of biotechnology in New Zealand 247–8 strategic, hybrid, innovative, public and scepticism (SHIPS) 47 proprietary, local, authoritarian, commissioned and expert (PLACE) 47 time constraints on 48–9 universities 224 in Western universities 92 see also public domain vs proprietary research; studies (on) research, development and innovation (RDI) policies 218–42 and four patterns of RDI policymaking conventions 221–7 ‘Republic of Science’ 221–3 state as entrepreneur 223–4

348


state as facilitator (of technological products) 225–7 state as regulator 224–5 methodological appendix for 239–42 research and development (R&D) 1–2, 8, 22, 32, 77, 85–6, 89, 91, 92, 124, 127, 130, 132, 134–5, 147, 156, 160, 161, 205, 207 collaboration 67 expenditure 143, 144, 295 investment 80–81, 122, 123 and IPR 86 -intensive corporations/firms 80, 83 national policies 218–42 policy issues 275–6 spending, growth in 133 statistics 280 see also France; Germany; United Kingdom (UK research and development (R&D): national policies 218–42 analytical framework: construction of policy-making conventions 219–21 public regimes of action in the UK, Germany and France 228–34 see also France; Germany; research, development and innovation (RDI) policies; United Kingdom (UK) research institutes Burnham 145 Gottfreid-Wilhelm-Leibniz Association of Research Institutes 231 La Jolla 145 Salk 145, 154, 158, 159 Scripps Research Institute 145, 155, 158, 159 Torrey Mesa Research Institute (San Diego) 152 reasoning 56–8 causal 57–8 deductive 57 probabilistic 56–7 Richardson, G.B. 110 Richter, R. 109 Rip, A. 47

risk 16, 66–7, 85–6, 99–100, 102–3, 324 adversity 15 aversion 85, 108 capital 1 management 109, 111, 249 perception 66 propensity 58–9 of short-sightedness and merchandization in UK 228–30 Rogers, E.M. 269 Rolleston, William 250 Rohm & Haas 153 Rosenberg, N. 32, 34, 35, 36, 61, 62 Ross, B.H. 45 Ruggles, N. 272 Ruggles, P. 272 Rumain, B. 44 Ryan, C. 152 Ryle, G. 23, 264–5, 266, 267 Sable, C. 316 Sah, R.K. 100 Salais, R. 220, 224, 226 Sampat, B.N. 74 Samson, A. 248, 250 Samuelson, Paul 265 San Diego 156 Sassen, S. 160 Sauvy, A. 272 Saxenian, A.L. 226 schemas 44–5 pragmatic 44 Schippers, M.C. 53 Schmitter, P.C. 314 Schmoch, U. 227 Schmookler, J. 114 Schoenherr, R. 295 Schultz, T.W. 273,275 Schumpeter, J.A. 101–2, 121, 146–9, 292, 297 and post-Schumpeterian symbiotics 144 Schumpeterian(s) corporation 102 innovation/entrepreneurship 161 insights 147 model 19 and Neo-Schumpeterian innovation theorists 146, 148 tradition 148

Index Schwartz, D. 292 Science 279 Science, Technology & Human Values 247 science and technology, integration between 4 Science Board, National 207 Science Foundation, National (NSF) Grant 211 science–industry relations 218–20, 229, 231 scientific software and databases 182–3 in Europe 182–3 privatization of 182 Scotchmer, S. 131, 169, 170 Scott, Anthony 254 Scully, J. 152, 155 Searle, J. 45 Selden patent 131 Second Academic Revolution 6 Second Industrial Revolution 35, 41, 43 Siegel, D. 5, 46, 47 Senker, J. 205 Sent, E.-M. 292 Shannon, C.E. 269, 294 formulae of 297 Shimshoni, D. 206 Shinn, T. 225 Shirley, M.M. 98 Shoemaker, F.F. 269 Silicon Valley 9, 11 venture capitalists 2 Simon, H. 56, 60 Sloman, S.A. 54 small and medium-sized enterprises (SMEs) 143, 145, 161, 231–2 Small Business Economics 143 Smith, Adam 109, 121 Smith, E.E. 45 Smith-Doerr, L. 143 social interactions and collaborative research 82–7 social network analysis 313 Solow, R.M. 23, 270, 271–2, 279 Soskice, D. 218, 227 Space Imaging Corporation 182 Sperber, D. 54 Standard Oil 32

349

Stanford University 158, 159, 201, 208, 210–211 and MIT model 21 Stankiewicz, R. 291 Stanovich, K. 59 Stephan, P.E. 4, 6, 169 Sternberg, R.J. 40, 56 Stevens, A. 207 Steward, F. 318 STI Scoreboard 295 Stigler, G.J. 266 Stiglitz, J.E. 100, 101, 125, 266 Stokes, D.E. 4 Storper, M. 219, 220, 224, 292, 293–4 Streeck, W. 314 structural change and unemployment 277 Studenski, P. 272 studies (on) with Alzheimer patients 45 social norms 47–8 Suárez, F.F. 300 Sunstein, C.R. 3, 14, 61, 67 surveys on effects of changes in IPR regimes (Jaffe, 2000) 137 Survey of Doctorate Recipients 6 Sweden 9, 10, 11, 151 Syngenta 151, 152, 154 System of National Accounts 262, 272–4 Tamm, P. 160 Tartu (Estonia) 160–161 technological paradigm 127 technologies 5 biotechnology 5, 41 information and communication (ICT) 5 nanotechnology 5, 41 technology, description of 127–8 Technology Kingdom 228 Technology Investment Program (Advanced Technology Programme) 2 technology transfer 12, 15, 34, 67–8, 74–5, 81, 206, 210, 229, 232 agents (TTAs) 21, 25, 46, 52, 65, 67 offices 12, 85, 202, 203, 204 Teece, D. 122, 125, 138, 39

350


Terman, Frederick 210 Teubal, Morris 16, 113 Thaler, R.H. 3, 14, 50, 61, 67 Theil, H. 294, 301 theories DNA 62 information 62 generality of application of 62 Theory of Economic Development 101 Thevenot, L. 219 Thompson, D.F. 323 Thursby, J. 179 Thursby, M. 5, 179 time constraints on problem-solving 56 research 48–9 Tirole, J. 129, 172, 175, 178, 190 Tomlinson, M. 303 top-down and bottom-up initiatives 13 Torre, A. 156 triple helix in economic cycles 1–3 triple helix: laboratory of innovation 9–14 triple-helix model 60 anticyclic role of 2 triple-helix relations and economy knowledge base 291–311 triple-helix spaces 10–11 trust (and) internal cohesion 316–17 or translation 319–20 translation and communication 317–19 Tushman, M.L. 300 Tversky, A. 45, 57, 58 Ulanowicz, R.E. 294, 306 Unilever 151 United Kingdom (UK) (and) chemicals companies in 153 government reforms in R&D 229–30 Higher Education Innovation Fund 229 Joint Infrastructure Fund 230 Oxbridge 230 Racal 228–9 RDI policies 219, 228–30 risks of short-sightedness/ merchandization 228–30 Science Enterprise Challenge 229

science–industry relations in 229 universities in 229–30 University Challenge 229 virtual centres of excellence (VCE) 229–30 United States of America (US) 150, 207–9 agricultural innovation 13 Bush Administration in 2 Court of appeals for the Federal Circuit (CAFC) 129 Department of Commerce 130 Justice Department 125 National Science Foundation 273–4 patent applications/cases in 128, 129–30 and privatization of Landsat images 182 Reagan Administration in 182 research universities 82–5 universities in 6, 7, 15 University of California and San Diego 2, 145 universalism 47, 55, 60, 66 vs localism 48 universities 1–3 passim; 32, 82–5 passim; 125, 129, 153, 156, 159, 170, 174, 201–7, 210, 215, 221, 243, 250, 274, 193, 295 ‘Centres of Excellence’ and research 143–4 in EU 91 industrial penumbra of 209–11 and research 11–12, 46, 201, 224 and strategic alliances/joint ventures 11 UK and corporate liaison offices/ officers 86–7 virtual 11 see also academy–industry relations university patenting for technology transfer 74–81 Utterback, J.M. 300 Valente, T. 314, 318 Van der Panne, G. 24, 295, 302 Van Knippenberg, D. 53 Van Looy, B. 7 Vanoli, A. 272 Velsasco, José Luis 25

Index venture capitalism 10, 16–17 as mechanism for knowledge governance 98–120 and NASDAQ 110–116 bundling finance and competence with innovation 111 knowledge intensive property rights 111–12 see also finance and innovation; innovation; markets; NASDAQ Verdier, E. 22, 233 Viale, Riccardo 3, 14, 25, 32, 35, 38, 39, 42, 43, 46, 47, 54, 64, 66, 67, 68 Vinck, D. 226 Von Hippel, E. 171, 173 Von Krogh, G. 173 von Mises, Ludwig 263 Von Wright, G.H. 42, 43, 44 Vonortas, N.S. 295 Wakoh, H. 174 Walker, G. 314 Walsh, J.R. 273 Walshok, M. 145 Warren, M.E. 323, 324

Wason, P.C. 57 Watson, James 148 Watts, R.J. 300 Weale, A. 324 Weaver, C.K. 248 Weaver, W. 269 Weiss, W. 100 Wevers, Francis 248, 252, 253, 254 Whitley, R.D. 292 Wiener, N. 269 Wilson, D. 54 Windrum, P. 303 Winter, S. 125, 127, 134, 138, 265 Wintjes, R. 297 Wolter, K. 146 Woolgar, S. 212 Worrall, J. 250 Yao, D.A. 171 Yoshaki Ito 160 Young, A. 280 Zeller, C. 154–5 Ziedonis, R. 129, 132, 134, 171 Ziman, J.M. 47 Zucker, L. 144, 162

351

The Capitalization of Knowledge: A Triple Helix of University-industry-government

The Triple Helix: Gene, Organism, and Environment

The Triple Helix: Gene, Organism, and Environment

Orphans of the Helix

Orphan of the Helix

Orphan of the Helix

The Triple Helix: Gene, Organism, and Environment

Orphans of the Helix

05 - Orphan Of the Helix

Dan Simmons - Orphan of the Helix

Dan Simmons - Orphans of the Helix

Triple

Triple

Triple

Triple

Triple

TRIPLE

Dan Simmons - Orphans of the Helix

Dan Simmons - Orphan of the Helix

The double helix. a personal account of the discovery of the structure of DNA

The Double Helix: A Personal Account of the Discovery of the Structure of DNA

The Sword of Knowledge

The Possibility of Knowledge

The Beginning of Knowledge

The Sword of Knowledge

The Philosophy of Knowledge

The Metaphysics of Knowledge

The Metaphysics of Knowledge

The concept of knowledge

The Possibility of Knowledge

Chattamanavaka Gatha (Stanzas of the Triple Gem)

The Capitalization of Knowledge: A Triple Helix of University-industry-government

The Triple Helix: Gene, Organism, and Environment

The Triple Helix: Gene, Organism, and Environment

Orphans of the Helix

Orphan of the Helix

Orphan of the Helix

The Triple Helix: Gene, Organism, and Environment

Orphans of the Helix

05 - Orphan Of the Helix

Dan Simmons - Orphan of the Helix

Dan Simmons - Orphans of the Helix

Triple

Triple

Triple

Triple

Triple

TRIPLE

Dan Simmons - Orphans of the Helix

Dan Simmons - Orphan of the Helix

The double helix. a personal account of the discovery of the structure of DNA

The Double Helix: A Personal Account of the Discovery of the Structure of DNA

The Sword of Knowledge

The Possibility of Knowledge

The Beginning of Knowledge

The Sword of Knowledge

The Philosophy of Knowledge

The Metaphysics of Knowledge

The Metaphysics of Knowledge

The concept of knowledge

The Possibility of Knowledge

Chattamanavaka Gatha (Stanzas of the Triple Gem)

Recommend Documents