Culture and Economic Explanation: Economics in the US and Japan

Culture and Economic Explanation

There is a common view among many economists that one model is capable of explaining a specific type of behavior in all cultural environments. It is only necessary to make appropriate adjustments to bring the model in line with prevailing cultural conditions. This book argues that such an approach can lead to error, in particular to incorrect explanation and understanding of the phenomenon in question, and therefore may result in inappropriate policy recommendations. Katzner’s fascinating book compares the two cultures of Japan and the US and provides insights into the economic workings and differences between the two nations. He shows that an understanding of the culture of a country is essential to the development of appropriate models of economic behavior of economic agents in that country, and that the failure to understand cultural differences weakens the predictive (and prescriptive) power of economic models. The argument is made in a collection of essays supporting the following: (a) thought processes are heavily dependent on cultural environments and (b) because cultures vary widely from society to society, to explain economic behavior in one society may require a model with a completely different structure from that in another. The book applies this argument to elucidate certain features of economic theorizing and to explain the so-called Japanese economic miracle. Challenging the assumption that social values and mores are universal, this book is relevant to all Political and International Economists and those with a particular interest in the US and Japan. Donald W. Katzner is Professor of Economics at the University of Massachusetts, Amherst, USA.

Routledge Frontiers of Political Economy

1

2

Equilibrium Versus Understanding Towards the rehumanization of economics within social theory Mark Addleson Evolution, Order and Complexity Edited by Elias L. Khalil and Kenneth E. Boulding

3

Interactions in Political Economy Malvern after ten years Edited by Steven Pressman

4

The End of Economics Michael Perelman

5

Probability in Economics Omar F. Hamouda and Robin Rowley

6

Capital Controversy, Post Keynesian Economics and the History of Economics Essays in honour of Geoff Harcourt, volume one Edited by Philip Arestis, Gabriel Palma and Malcolm Sawyer

7

Markets, Unemployment and Economic Policy Essays in honour of Geoff Harcourt, volume two Edited by Philip Arestis, Gabriel Palma and Malcolm Sawyer

8

Social Economy The logic of capitalist development Clark Everling

9

New Keynesian Economics/Post Keynesian Alternatives Edited by Roy J. Rotheim

10

The Representative Agent in Macroeconomics James E. Hartley

11

Borderlands of Economics Essays in honour of Daniel R. Fusfeld Edited by Nahid Aslanbeigui and Young Back Choi

12

Value, Distribution and Capital Essays in honour of Pierangelo Garegnani Edited by Gary Mongiovi and Fabio Petri

13

The Economics of Science Methodology and epistemology as if economics really mattered James R. Wible

14

Competitiveness, Localised Learning and Regional Development Specialisation and prosperity in small open economies Peter Maskell, Heikki Eskelinen, Ingjaldur Hannibalsson, Anders Malmberg and Eirik Vatne

15

Labour Market Theory A constructive reassessment Ben J. Fine

16

Women and European Employment Jill Rubery, Mark Smith, Colette Fagan, Damian Grimshaw

17

Explorations in Economic Methodology From Lakatos to empirical philosophy of science Roger Backhouse

18

19

Subjectivity in Political Economy Essays on wanting and choosing David P. Levine The Political Economy of Middle East Peace The impact of competing trade agendas Edited by J.W. Wright, Jnr

26

Political Economy and the New Capitalism Edited by Jan Toporowski

27

Growth Theory A philosophical perspective Patricia Northover

28

The Political Economy of the Small Firm Edited by Charlie Dannreuther

29

Hahn and Economic Methodology Edited by Thomas Boylan and Paschal F. O’Gorman

30

Gender, Growth and Trade The miracle economies of the postwar years David Kucera

20

The Active Consumer Novelty and surprise in consumer choice Edited by Marina Bianchi

31

Normative Political Economy Subjective freedom, the market and the State David Levine

21

Subjectivism and Economic Analysis Essays in memory of Ludwig Lachmann Edited by Roger Koppl and Gary Mongiovi

32

Economist with a Public Purpose Essays in honour of John Kenneth Galbraith Edited by Michael Keaney

33

Involuntary Unemployment The elusive quest for a theory Michel De Vroey

34

The Fundamental Institutions of Capitalism Ernesto Screpanti

35

Transcending Transaction The search for self-generating markets Alan Shipman

36

Power in Business and the State An historical analysis of its concentration Frank Bealey

37

Editing Economics Essays in honour of Mark Perlman Hank Lim, Ungsuh K. Park and Geoff Harcourt

22

Themes in Post-Keynesian Economics Essays in honour of Geoff Harcourt, volume three Edited by Claudio Sardoni and Peter Kriesler

23

The Dynamics of Technological Knowledge Cristiano Antonelli

24

The Political Economy of Diet, Health and Food Policy Ben J. Fine

25

The End of Finance Capital market inflation, financial derivatives and pension fund capitalism Jan Toporowski

38

Money, Macroeconomics and Keynes Essays in honour of Victoria Chick volume 1 Philip Arestis, Meghnad Desai and Sheila Dow

39

Methodology, Microeconomics and Keynes Essays in honour of Victoria Chick, volume 2 Philip Arestis, Meghnad Desai and Sheila Dow

50

Rethinking Capitalist Development Essays on the economics of Josef Steindl Edited by Tracy Mott and Nina Shapiro

51

An Evolutionary Approach to Social Welfare Christian Sartorius

52

Kalecki’s Economics Today Edited by Zdzislaw L. Sadowski and Adam Szeworski

53

Fiscal Policy from Reagan to Blair The left veers right Ravi K. Roy and Arthur T. Denzau

54

The Cognitive Mechanics of Economic Development and Institutional Change Bertin Martens

40

Market Drive & Governance Reexamining the rules for economic & commercial contest Ralf Boscheck

41

The Value of Marx Political economy for contemporary capitalism Alfredo Saad-Filho

42

Issues in Positive Political Economy S. Mansoob Murshed

55

Individualism and the Social Order The social element in liberal thought Charles R. McCann Jnr.

43

The Enigma of Globalisation A journey to a new stage of capitalism Robert Went

56

Affirmative Action in the United States and India A comparative perspective Thomas E. Weisskopf

44

The Market Equilibrium, stability, mythology S. N. Afriat

57

45

The Political Economy of Rule Evasion and Policy Reform Jim Leitzel

Global Political Economy and the Wealth of Nations Performance, institutions, problems and policies Edited by Phillip Anthony O’Hara

58

46

Unpaid Work and the Economy Edited by Antonella Picchio

Structural Economics Thijs ten Raa

59

47

Distributional Justice Theory and measurement Hilde Bojer

48

Cognitive Developments in Economics Edited by Salvatore Rizzello

Macroeconomic Theory and Economic Policy Essays in honour of Jean-Paul Fitoussi Edited by K. Vela Velupillai

60

The Struggle Over Work The “End of Work” and employment alternatives in post-industrial societies Shaun Wilson

49

Social Foundations of Markets, Money and Credit Costas Lapavitsas

74

Growth and Development in the Global Political Economy Social structures of accumulation and modes of regulation Phillip Anthony O’Hara

75

Uncertainty in Economic Theory Essays in honor of David Schmeidler’s 65th birthday Edited by Itzhak Gilboa

The New Economy and Macroeconomic Stability A neo-modern perspective drawing on the complexity approach and Keynesian economics Teodoro Dario Togati

76

The New Institutional Economics of Corruption Edited by Johann Graf Lambsdorff, Markus Taube and Matthias Schramm

The Future of Social Security Policy Women, work and a citizens basic income Ailsa McKay

77

Clinton and Blair The political economy of the third way Flavio Romano

78

Marxian Reproduction Schema Money and aggregate demand in a capitalist economy A.B. Trigg

61

The Political Economy of Global Sporting Organisations John Forster and Nigel Pope

62

The Flawed Foundations of General Equilibrium Theory Critical essays on economic theory Frank Ackerman and Alejandro Nadal

63

64

65

The Price Index and its Extension A chapter in economic measurement S.N. Afriat

66

Reduction, Rationality & Game Theory in Marxian Economics Bruce Philp

67

Culture and Politics in Economic Development Volker Bornschier

79

The Core Theory in Economics Problems and solutions Lester G. Telser

68

Modern Applications of Austrian Thought Edited by Jürgen G. Backhaus

80

Economics, Ethics and the Market Introduction and applications Johan J. Graaf land

69

Ordinary Choices Individuals, incommensurability, and democracy Robert Urquhart

81

70

Labour Theory of Value Peter C. Dooley

Social Costs and Public Action in Modern Capitalism Essays inspired by Karl William Kapp’s theory of social costs Edited by Wolfram Elsner, Pietro Frigato and Paolo Ramazzotti

71

Capitalism Victor D. Lippit

82

72

Macroeconomic Foundations of Macroeconomics Alvaro Cencini

Globalization and the Myths of Free Trade History, theory and empirical evidence Edited by Anwar Shaikh

83

73

Marx for the 21st Century Edited by Hiroshi Uchida

Equilibrium in Economics: Scope and Limits Edited by Valeria Mosini

84

Globalization State of the art and perspectives Edited by Stefan A. Schirm

85

Neoliberalism National and regional experiments with global ideas Edited by Ravi K. Roy, Arthur T. Denzau, Thomas D. Willett

86

87 88

Post-Keynesian Macroeconomics Economics Essays in honour of Ingrid Rima Edited by Mathew Forstater, Gary Mongiovi and Steven Pressman Consumer Capitalism Anastasios S. Korkotsides Remapping Gender in the New Global Order Edited by Marjorie Griffin Cohen and Janine Brodie

89

Hayek and Natural Law Eric Angner

90

Race and Economic Opportunity in the Twenty-First Century Edited by Marlene Kim

91

Renaissance in Behavioural Economics Harvey Leibenstein’s impact on contemporary economic analysis Edited by Roger Frantz

92

Human Ecology Economics A new framework for global sustainability Edited by Roy E. Allen

93

Imagining Economics Otherwise Encounters with identity/difference Nitasha Kaul

94

Reigniting the Labor Movement Restoring means to ends in a democratic labor movement Gerald Friedman

95

The Spatial Model of Politics Norman Schofield

96

The Economics of American Judaism Carmel Ullman Chiswick

97

Critical Political Economy Christian Arnsperger

98

Culture and Economic Explanation Economics in the US and Japan Donald W. Katzner

Culture and Economic Explanation Economics in the US and Japan

Donald W. Katzner

First published 2008 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN Simultaneously published in the USA and Canada by Routledge 270 Madison Avenue, New York, NY 10016 This edition published in the Taylor & Francis e-Library, 2008. “To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk.” Routledge is an imprint of the Taylor & Francis Group, an informa business

© 2008 Donald W. Katzner All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging in Publication Data Katzner, Donald W., 1938– Culture and economic explanation : economics in the US and Japan / Donald W. Katzner. p. cm. Includes bibliographical references and index. ISBN 978-0-415-77479-6 (hbk.) -- ISBN 978-0-203-93105-9 (e-book) 1. Economics--Japan. 2. Economics--United States. 3. Culture--Economic aspects. I. Title. HB126.J2K377 2008 330.952'05--dc22 2007033806 ISBN 0-203-93105-X Master e-book ISBN

ISBN 10: 0-415-77479-9 (hbk) ISBN 10: 0-203-93105-X (ebk) ISBN 13: 978-0-415-77479-6 (hbk) ISBN 13: 978-0-203-93105-9 (ebk)

Contents

List of illustrations Preface 1 Introduction: culture, economics, and economic behavior 2 Western economics and the economy of Japan

x xi 1 17

Reply to Thiruvadanthai

33

Profit maximization and the Japanese firm: a reply to Coffey and Tomlinson

37

3 “What are the questions?”

42

4 An analytical vision of the workings of the United States economy

61

5 The workings of the Japanese economy

82

6 Explaining the Japanese economic miracle

101

7 The role of optimization in economics

119

8 Economics and the principle of uniformity

131

9 Cultural variation in the theory of the firm

147

10 Culture and the explanation of choice behavior Index

161 179

Illustrations

Tables 6.1 Comparative labor productivity levels by sector: Japan as a percentage of the US 6.2 Total factor productivity in Japan relative to the US 6.3 Returns to capital in Japan and the US 1967–1988

114 115 116

Figures 1.1 Rational behavior by an American; irrational behavior by a Japanese 1.2 Rationalizing the Japanese behavior 6.1 Profitability at various wages and employments in the Japanese firm

9 10 110

Preface

Economists tend to date the beginnings of modern Economics with the publication of Adam Smith’s Wealth of Nations in 1776. In that volume, Smith wrote the oftquoted words, “It is not from the benevolence of the butcher, the brewer, or the baker, that we expect our dinner, but from their regard to their own interest.”1 Smith also spoke of an invisible hand, by which he meant the freely functioning market system, guiding individuals pursuing their own self-interest in a direction beneficial to the common good. And the ideas of self-interest and a guiding invisible hand, liberally interpreted, have remained at the core of Economics ever since. During the next 125 years, self-interest was frequently translated into the language of rationality or maximization,2 and the invisible hand was expressed with respect to the simultaneous behavior of rational economic agents (i.e. consumers and firms) operating under competitive conditions. Perhaps due, in part, to the nature of these conceptualizations and their complexity, economic analysis turned in the direction of mathematics. This movement was abetted by several mathematically trained engineers, such as Augustin Cournot and Leon Walras, who left their original disciplines to explain the workings of the economic system in mechanical and mathematical terms. However, it was only during the second half of the twentieth century that economics became overwhelmingly mathematical. Before that time, of course, significant challenges to the previously existing analytical paradigm had been proposed. For example, Chamberlin3 and Robinson4 questioned the assumption of perfectly competitive product markets, and Keynes,5 building on a significant body of work that preceded the Wealth of Nations, argued that the market system would not inevitably equilibrate at full-employment levels of economic activity.

1 Randon House edition: New York, 1937. 2 Although there are differences, for purposes of simplicity I will here treat self-interest, rationality, and maximization as alternative manifestations of the same thing. 3 E.H. Chamberlin, The Theory of Monopolistic Competition, 8th ed. (Cambridge: Harvard University Press, 1962). 4 J. Robinson, The Economics of Imperfect Competition, 2nd ed. (London: Macmillan, 1939). 5 J.M. Keynes, The General Theory of Employment, Interest and Money (New York: Harcourt Brace, 1936).

xii Preface But while that is so, today’s widely held vision of the functioning of the full microeconomy is essentially a dressed-up, mathematical version of that of Adam Smith – with a few wrinkles, such as the idea of a market game, thrown in. And in all of this, the notions of self-interest and the invisible hand, now referred to as, respectively, rationality and competition, reign supreme. Such ideas seem to be accepted outright by economists as a general description of the way the economy works. Few question whether the vision is applicable from one context or culture to the next. In 1992, Gary Becker won the Nobel Prize in Economics for showing that economic analysis, in particular self-interest or rationality, extends well beyond the traditional confines of the subject-matter of economics, into the territory of Social and Behavioral Science in general.6 Becker, himself, asserts explicitly in the Introduction to a collection of his essays entitled The Economic Approach to Human Behavior (Chicago: University of Chicago Press, 1976; p. 14) that all human behavior, that is, everything that you and I do, is explainable in the economic terms of rationality. In his view, every potential action is carefully weighed in advance, and the decision to undertake it depends on an orderly assessment of its costs relative to its benefits, implying thereby an allocation of scarce resources that meets whatever objectives are thought to be most desirable. There is no room for thoughtless, whimsical, or impulsive conduct, or acts of passion. Behavior derived from Keynes’ “animal spirits” 7 and what Dostoyevsky called “our own sweet foolish will” 8 is also ruled out. This also appears to be a basic tenet of the so-called Chicago School of Economics, and its acceptance by the economics profession at large was signaled by the awarding of the Nobel Prize to Becker. Of course, one of the implications of Becker’s view, and from my reading of the economics literature this seems to be believed by many economists today, is that rationality or self-interest is the primary behavior-motivating force in all cultures and under all circumstances. The calculation of benefits and costs drives both economic and noneconomic behavior in the bazaars of Damascus, in the crime-infested streets of Detroit and Washington, in the Keiretsu and other large enterprises of Tokyo and Osaka, and in the living rooms and bedrooms of suburbia everywhere. It is this view that I am challenging with the publication of these essays. More generally, I am trying to convince the reader that economists as a group, if they want their discipline to have meaning and relevance everywhere, have to do much more than they have in the past to integrate culture into their considerations.

6 This, of course, may be regarded as an expansion of L. Robbins’ famous definition of economics as “the science which studies human behaviour as a relationship between ends and scarce means which have alternative uses.” (An Essay on the Nature and Significance of Economic Science (London: Macmillan, 1948), p. 16.) 7 op. cit., p. 161. 8 F. Dostoyevsky, Notes from the Underground, C. Garnett, trans. (New York: Heritage Press, 1967), p. 38.

Preface xiii The collecting of the ensuing essays in a single volume presents, I believe, a much more cogent and more complete argument than is possible to convey by leaving each essay published in its original, isolated source. The links between essays become more obvious; the points made in separate essays reinforce each other; and the overall result is considerably more powerful and compelling. Publishing the essays in one place also makes them more readily accessible. I hope these advantages far outweigh any annoyance caused by the fact that there is, inevitably, some overlap across essays. In preparing this anthology, I have corrected typographical and other errors appearing in the original publications. Essay 3 has been substantially expanded and minor changes in wording have been made throughout in order to enhance clarity and the unity of the vision projected by the volume. I am most grateful for the help given to me, as acknowledged in footnotes, on both the original publications and the new, previously unpublished Essays 1 and 4. I would also like especially to thank Douglas Vickers, not only for his help (as cited) on most of the essays, but also for his encouragement and support in compiling this book.

1

Introduction Culture, economics, and economic behavior1

Culture, according to Edward Tylor, the anthropologist who is credited with providing its first explicit definition, “… is that complex whole which includes knowledge, belief, art, morals, law, custom, and any other capabilities and habits acquired by man as a member of society” [13, p. 1]. To this list might be added such things as behavior, language, ideas, taboos, codes, institutions, tools, techniques, rituals, and ceremonies. Culture may therefore be seen, in part, as a complex of ideas, thought-forms, behavior patterns, and judgmental criteria that determine a community’s way of life. Culture shapes the manner in which each person satisfies his needs by providing rules that determine what those needs are, and when and how they are to be satisfied. Thus it not only specifies, for example, what needssatisfying tools are to be employed in what circumstances, but it also indicates when and how those tools are to be used and, to some extent, it influences the formation of the tools themselves. Culture furnishes a system of meaning for understanding reality along with standards for interpretation, judgment, and action. It helps to control the behavior of individuals (i) by supplying norms or guidelines directing that behavior and (ii) through institutions such as the family, the instruments of government, the organization of production, and economic exchange. Culture is a learned phenomenon; it is not passed on via genetic inheritance.2 In particular, children acquire knowledge of culture by observation, by what they are told, and by what they are told to do or prevented from doing. And it is communicated through the use of symbols, mostly, though hardly exclusively, by language. The individual elements or properties of culture are called cultural traits and collections of related traits are referred to as cultural patterns. It is clear from this description that every academic discipline is swayed by, and contributes to the formation of, important cultural traits identified with the part of the society’s culture in which that discipline is located, and these traits constitute, at least implicitly, a significant and inseparable component of the discipline’s content. That is so because each discipline is associated with a portion of many of

1 Thanks are due to Douglas Vickers for his help in preparing this essay. 2 Of course, the structure of culture that emerges in a society may well be influenced by behavior that results from genetic inheritance and the extent to which such behavior acquires social acceptability.

2 Introduction those cultural elements listed above including, among other things, the society’s knowledge, ideas, language, and institutions. In particular, culture influences the kinds of situations to which the discipline responds. With respect to economics, it is clear that economic behavior occurs in a social context and is culturally qualified. Thus different cultures may raise different demands for economic resources, and different priorities for the use of those resources. The use of different technologies across cultures (which may be a consequence of differing cultural imperatives) may not only contribute to the determination of differing resource availabilities, but may also give rise to different demands and priorities as well. The latter demands and priorities may also reflect different taboos and customs. These facts imply that, in each case, whatever behavior is observed may require a different explanation to properly understand it and, from the perspective of the functioning of the economy as a whole, whatever resources are available may have to be adapted differently. Thus the questions economists ask, the answers they provide, and possibly even the methods employed for constructing answers, are all sensitive to the cultural background in which both the investigation as well as the objects of study are embedded. Because each academic discipline may be thought of as having its own specific body of knowledge, ideas, language, and institutions, it may be viewed as having a distinct culture by itself. From this perspective, the discipline’s culture is a “subculture” of society’s culture, that subculture being understood as a complex consisting of a subcollection of all of society’s cultural traits. Thus, in reference to American society, for example, economics, as practiced by the American economics profession, is a subculture of American culture. And, as has been suggested above, what develops as the culture of economics in America is itself influenced by the wider cultural mores of society at large. Moreover, the cultures of disciplines may themselves have subcultures. This is certainly the case in economics. With respect to the latter, at least two subcultures can be distinguished. These are related in part to alternative styles of argument that have been employed by economists, both styles having emerged from the culture of economics at different times.3 And, for the most part, each subculture has narrowly focused on purely economic matters alone without recognition of the broader cultural influences that impinge on those matters. One approach involves the construction of informal analytic argument in response to perceived actual economic problems and states of affairs. An economic event happens, say, or an economic state of affairs emerges, and, without necessarily developing a highly formal structure within which the event can be understood as part of a broad, interrelated whole, the economist provides a relatively simple, direct, and analytically competent explanation of how and why it occurred. To illustrate, when Adam Smith set out to explain why individuals in a capitalist economy produce goods to sell to others he wrote, “It is not from the benevolence of the butcher,

3 Clearly, the culture of any discipline continually changes over time, as does, also, the wider cultural context within which that discipline exists.

Introduction 3 the brewer, or the baker that we expect our dinner, but from their regard to their own interest” [10, p. 14]. In Smith’s analysis there is no utility function defined over a multidimensional commodity space to be maximized given appropriate market prices and an endowment constraint, and no model of market exchange. There is only a straightforward statement of what Smith claims is going on.4 This mode of argument has been common in economics from its earliest beginnings into the twentieth century (the classical economists, Marshall, and Keynes all used it) and remains in vogue today, in varying degrees and sometimes with substantial modifications, among the Austrian economists and many of the socalled institutionalists and post-Keynesians, as well as among applied practitioners in the mainstream of economics. It has produced and continues to produce a certain kind of knowledge, and is associated with certain beliefs and ideas. It has its own language, its own way of understanding reality, its own standards for interpretation, and its own standards for judging good argument. In these regards, it defines a subculture of economics as set out above. The alternative approach to argument is the development of formal analytical structures or models that are generated deductively from a set (usually a small set) of basic axioms or postulates. All issues are addressed as features arising within a web of interacting simultaneous static or dynamic relations. This analytical path was introduced into economics at least in embryonic form by what may be referred to as the methodological revolution of Ricardo. In spite of the fact that it was used by such economists as Cournot and Walras, it remained at the margins of economic discourse until the second half of the twentieth century when it was popularized by Samuelson, Debreu, and others. Today, it is the most widely employed form of economic expression in the discipline. Like the informal mode of argument described above, it, too, is accompanied by its own knowledge, beliefs, ideas, language, way of understanding reality, standards for interpretation, and standards for judging good argument. Therefore, it also identifies its own corresponding subculture within economics. Observe that the two subcultures of economics described here are quite distinct. Each has a different vision of what constitutes economics and, in a general way, the arguments relating to each tend to speak to rather different issues and questions. In other words, the style of economic argument has been, as it should be, associated with the objective of that argument and the character of the questions that are posed. Thus the analysis-in-response-to-problems or “expositional” approach has drawn the questions it addresses directly from the perceived characteristics of the problem being faced (those perceptions are based on the analysis-in-response-toproblems subculture), and the assumptions of its argument have been determined

4 It is true that in an economic argument of this kind the analyst would understand the thing or condition to be explained as determined in large degree by the wider cultural influences referred to earlier. In particular, Smith’s explanation is understood to be what it is because of the cultural acceptability, or even priority, in the society to which he implicitly refers, of the pressure for generalized economic gain.

4 Introduction by the subcultural view taken of the nature, scope, and causes of the problem at hand. By contrast the axiomatic-deductive or “model-building” approach has tended to raise issues in terms of the perceived properties (with respect to its own subculture) of general classes of problems, and derive its assumption content from the perceived common attributes across those classes. Moreover, each of these subcultures may well give rise to the development of explanatory systems that involve the rejection of certain basic ideas or assumptions of alternative schemes of thought – even within a given subculture. In that sense further subcultures of economics may emerge. A prominent example of this emergence is the Keynesian rejection of the classical economic postulate that the macroeconomy would automatically operate in such a way as to guarantee the full employment of its labor resources. It is clear, then, that culture plays an important role in shaping the nature and content of economics. On the one hand, what has been seen as the wider determinants of a society’s culture gives rise to the character of the problems and conditions that economics takes as its province of investigation. On the other, the styles of economic argument, or what has been referred to as the subcultures of the discipline, are themselves influenced by the professional and scientific acceptability of analytical procedures, as that is determined, in turn, by the cultural phenomenon of the progress of knowledge. The striking, indeed surprising, thing is that while culture has in these ways pervaded economics to its very core, the arguments that economists actually produce tend to give insufficient recognition to the prominence of cultural determinants in the minds of the subjects whose behavior those arguments are meant to explain. Indeed, there is no little irony in the fact that while culture profoundly influences what economists themselves do professionally, including the behavioral decisions they make concerning the nature of the questions they raise and the arguments they invoke, they do not seem willing to admit that the agents they study are also influenced by cultural backgrounds, at least as profoundly, in the making of the decisions that generate the economic behavior that economists investigate. A possible reply to this criticism is that such cultural influences can be introduced into standard economic constructions by, say, letting preferences vary from culture to culture in the case of the consumer (e.g. Schultz [9, pp. 20–22]),5 or focusing on differences in information availability across cultures with respect to decision-making in firms (e.g. Stiglitz [12, p. 23]). (The former, of course, ignores the admonition of Stigler and Becker [11, p. 76] that “… tastes … [do not] differ importantly between people.”) But such inclusions in standard economics are simply not enough. For to account properly for cultural variation in a particular economic analysis may require the rephrasing of analytical objectives, the restructuring of entire arguments and, if present, the major modification of formal structures. The latter modifications might include, among other things,

5 A detailed description of the literature relating to the impact of culture on economic outcomes through contact with individual preferences is given by Guiso et al. [3].

Introduction 5 the introduction of unquantifiable elements.6 The purpose of this book is to demonstrate that the failure to make such changes when warranted is a serious error with potentially far-reaching consequences. Although the interaction between culture on the one hand and economics and economic behavior on the other clearly flows in both directions, in what follows only the impact of culture on economics and economic action is considered. The basic argument is effected in four stages that establish the respects in which cultural considerations are relevant to economic discourse and economic behavior, and may be briefly summarized as follows below. First, a number of fundamental propositions relating to explanation in economics are traced: the main objective of economic analysis is taken to be the correct explanation of economic happenings and to show, on the basis of that explanation, what may conceivably occur in the future if certain imaginable conditions exist. Generally speaking, to explain something is to make it understandable, to show how it fits into the world of which it is a part, and to furnish the genesis, source, and potential significance of it. Explanations of economic phenomena, then, require the elucidation of the forces that cause those phenomena to appear, and explanations of economic behavior consist of providing reasons why the given behavior occurred. Explanation is correct if it gives an accurate picture of what is actually going on; it is erroneous if it does not. (Prediction is not sufficient for correct explanation because it may be derived from the wrong causes.) Due to the inadequacies of the economist’s “laboratory,” especially the inability to hold the necessary control variables fixed, it is not possible in many circumstances to test statistically whether an explanation is correct – even when considerable data are available. Thus the viability of economic argument, that is its ability to provide a “true” explanation of a state of affairs, often rests on the judgment of the investigator as to its “reasonableness.” But in the end, such reasonableness has, in turn, to be based on the relevance of the explanation to the cultural environment in the context of which the observed behavior is generated. When the latter cultural environment is the same as that of the investigating economist, the judgment of an explanation’s validity becomes personal – introspective in character and derived from private life experience. That is, the judgment is dependent to a considerable degree on the cultural background of the investigator. In that case, for him to be ready to accept an explanation, the investigating economist has to feel that its assumptions accurately reflect his own perception of the reality that is being explained. Second, in light of the preceding it becomes necessary to enter upon an analysis of the respects in which the cultural backgrounds relating to the phenomena under investigation, notably those of the individuals whose behavior is being studied, are central to any inquiry. This is because behavior is derived from thought processes, more or less ordered and articulated, that result in the making of decisions. Thought processes, in turn, are mental acts that rely heavily on the symbols and their

6 Analytical methods for the introduction into, and use of such elements in formal structures have been extensively studied by Katzner [5], [6].

6 Introduction interpretations imported into the mind from the individuals’ cultural backgrounds. The motivations that guide those processes are also part of the cultural baggage included with the imports. Moreover, to make the decisions that generate economic behavior requires that the individual, in his decision-making thought processes, manipulates the imported symbols in culturally learned ways. It follows, then, that individual and institutional behavior (institutions are run by individuals) is significantly dependent on cultural backgrounds through the thought processes and motivations leading to the decisions that generate that behavior. It is worth digressing for a moment to note that there is some laboratory evidence intimating that patterns and processes of thought vary across cultures. Much of it has been described by Nisbett [7]. A brief sampling of the conclusions Nisbett draws for experiments performed with American students and students from several Asian countries is as follows [7, pp. 191–192]: (1) when presented with a series of pictures containing background scenery and objects in the foreground, Japanese students more easily recognized changes in backgrounds while American students more easily recognized changes in foreground objects. (2) When asked which among two apparently contradictory statements were believable, American students tended to choose one or the other. But Chinese students tended to accept both. (3) When shown a thing (say, a clay pot), Japanese students were more likely to think of it as a substance (clay), whereas American students were more likely to regard it as an object (a pot). Although narrow in scope, these results also suggest the symbiotic relation between thought and culture, and hence the dependency of behavior on the latter. Third, account is taken of the fact that significant cultural differences can arise across, and even within, societies. These differences often manifest themselves in ways that cause certain cultural traits to become dominant in some societies and not in others. For present purposes, a dominant trait is defined as one that appears to have more influence than most in generating individual behavior. Cultural traits become dominant when specific events occur that institutionalize those traits, making them a general motivating force for behavior in society at large. Of course, there may be many cultural traits that are common to two societies having different dominant traits. Moreover, cultural evolution has occurred in all societies over the ages in response to the events and histories that each experiences. And differences in histories are responsible, in part, for differences in dominant traits among them. Fourth, and finally, it is recognized that in constructing explanations of economic phenomena and behavior, economists abstract from reality to focus on the factors thought to be the most significant with respect to the issue at hand. It follows from what has been said above, then, that if explanation is to be correct, the elements on which attention centers must include the dominant cultural traits of the society under investigation. Indeed, explanations by economists of phenomena and behavior in Western economies implicitly incorporate the dominant cultural traits of those societies in virtually all of their arguments. One of the most common cultural traits relevant for explanations of the economic behavior of Western individuals is the pursuit of self-interest. That trait is often represented in Western economics as the aspect of rationality known as maximization.

Introduction 7 (The term “maximization” in its present usage is intended to include such things as minimization where appropriate and satisficing in circumstances where full information is not available.7 ) This is not to say that all individual behavior in Western economies is explainable with respect to maximization. Indeed, there is considerable laboratory evidence suggesting the contrary.8 But as a precept for general explanations, maximization is acceptable and useful because of its close link to the underlying dominant cultural forces in those economics. However, this may not be so in other economies. Since different cultures are known to produce different symbols, different symbolic interpretations and manipulations, and different dominant traits and motivations, or, in short, since the individuals of different cultures have different thought processes and motivations at least in relation to decision-making, it is often necessary to provide different explanations of their behavior. In particular, the pursuit of self-interest, rationality, or maximization is neither relevant nor appropriate to correct explanations of economic behavior everywhere. (Laboratory evidence suggesting this divergence is presented in a different context below.) Therefore cultural backgrounds are an important component of correct explanation in economics. Any economist who ignores in his analysis the fullness of the potential impact of culture on behavior is exposing himself to the possibility of significant error. The four-stage argument just described is made in much greater detail in the next two essays. Essay 2, “Western Economics and the Economy of Japan” and the replies to criticism that follow it, largely cover the first, third, and fourth of the above-listed stages (that is, respectively, culture impacts judgments of the correctness of explanation, different dominant cultural traits emerge in different societies, and dominant cultural traits are fundamental to any analysis). Essay 3, “‘What are the Questions?’” deals, among other things, with the second of the above listed stages (culture significantly influences individual behavior). In these essays, as well as in those of the rest of this book, all specific comparisons, cultural and otherwise, are made between the US and Japan. This has been done because, since they are so clear and striking, the comparisons cited are easy to understand and appreciate. The first two essays deal with other points as well. But before proceeding to summarize their remaining contents and the contents of the other essays, it is necessary to consider the relationship of the preceding basic argument to what is called universalism and localism in explanation (more extensive discussion also appears in subsequent essays), and to phrase that argument in terms of a concrete illustration. With regard to the former relationship, note that the analysis-in-response-to-problems and the axiomatic-deductive approaches to

7 In subsequent language employed here, the meanings of self-interest, rationality, and maximization are conflated as follows: self-interest is a motivation. And rationality in the form of maximization is the representation of self-interest in a model. Thus with respect to the consumer, for example, utility maximization is the manifestation of both self-interest and rationality in standard (Western) models of consumer behavior. Of course, these definitions are highly specialized and need not apply generally. 8 See, for example, Fehr and Gächter [1].

8 Introduction argument are both, by themselves, potential parts of economic explanation. Each could be employed from a universalist or localist perspective. In general, universalism and localism are nonoverlapping, opposing categories of explanation defined with respect to the interaction between culture and economic behavior that they allow (Gudeman [2, Ch. 2]). Consider universalism first. From a universalist vantage point, and in denial of the basic argument of this book as outlined previously, economic behavior, while it is perceived as occurring in its own separate and isolated sphere of human activity, is also thought to be explicable in terms of universally applicable principles and methods of analysis. That sphere, that is the economic sphere, is generally understood to be independent of action that derives from cultural imperatives. On this perspective, expositions and models that explain economic behavior are relevant in all cultural settings and throughout all ages. It is only necessary to make relatively minor adjustments, in each application, to the particulars of place and time. Thus, for example, all consumers everywhere are assumed to be utility maximizers, although individuals with alternative cultural backgrounds may behave differently due to variations in preferences. That is, people’s behavior differs only because individuals like and dislike different things. It has been suggested earlier that economists are severely restricted by the limitations of the laboratory they have to work with. Those restrictions imply that universalism cannot be refuted by checking it or its implications against real human activity. For to continue with the case of the consumer, purchases of goods and services observed to be inconsistent with utility maximization can be attributed to changes in preferences over which the economist exercises no control. Thus that explanatory model cannot be falsified and, by changing the characteristics of the preferences assumed, all buying behavior in all cultures can be “explained” in terms of utility maximization. However, the problem with the universalist perspective is not only that the explanation it provides may be incorrect in that the individual’s thought processes may have nothing to do with utility maximization, but also that, because it treats all agents identically regardless of cultural backgrounds, it tends to limit the depth of economic explanation to the shallow waters of a few rather bland economic theories. If all consumers are seen as utility maximizers with only variations in tastes to tell them apart, then after pointing out preferential differences, economic explanations of conflicting behaviors cannot go further. The rest is psychology. Tastes are what they are. There is nothing more to say. The richness of human economic experience, derived from the existence of cross-cultural variations in thought processes and motivations, is confined to a straitjacket of rather limited analytical horizons that is unlikely to have meaning in all cultural environments to which it is applied. And vast landscapes of possible economic interpretations and analyses are excluded from consideration. Thus it is more interesting, more correct (in that it provides more accurate explanation), and therefore potentially more significant, to explore the cultural subtleties of an absence of self-interest or (Western) rationality (see note 7 above) in the behavior of a Japanese individual, say, who knowingly does not purchase a commodity at the lowest available price, than to merely assert that this person is actually behaving rationally but with a different preference ordering.

Introduction 9 That there are strong forces in Japanese culture leading to the predominance of such motives as loyalty over self-interest or rationality is argued in Essay 2. The general point is worth rephrasing and emphasizing in terms of the latter example. Imagine a two-good world in which both goods are identical except that they are purchased from different sellers. There are no special services or warranties that might distinguish them in the mind of the buyer. Nor are there any hidden costs associated with their purchase. In Figure 1.1, quantities of the good sold by seller 1 (also called good 1) are denoted by x1 and those sold by seller 2 (referred to as good 2) are written as x2 . These commodities would normally be treated by an American consumer, say, as perfect substitutes, and such an individual’s preferences among baskets containing them would usually be pictured (Figure 1.1) by an indifference map where each indifference “curve” is a straight line negatively sloped at a 45◦ angle. In this case, it is assumed that more of each good is preferred to less. Let the prices charged by the sellers be P1 and P2 , respectively. Were P1 > P2 , and were the budget constraint drawn as the dashed line AB in Figure 1.1, then the American, assuming utility maximization subject to the budget constraint, would choose basket A containing only quantities of the cheaper good sold by seller 2. But a Japanese, whose family has always purchased this commodity exclusively from seller 1, and who, for cultural reasons, feels obligated to remain loyal to that seller, will purchase basket B. The latter’s behavior clearly appears to be irrational when judged by the standard of this particular utility-maximization model. Of course, a universalist would immediately (and rightly) argue that properly accounting for loyalty in the Japanese person’s utility function would alter his choice in such a manner that, for him, point B would become rational. The general x2 A

B

45° x1

Figure 1.1 Rational behavior by an American; irrational behavior by a Japanese.

10 Introduction x2 A

B x1

Figure 1.2 Rationalizing the Japanese behavior.

argument is made in the context of choice theory in Essay 10. One way to do this in the present case would be to include the quality of loyalty to seller 1 as a characteristic of good 1. With such a modification, the two goods would no longer be identical, and the Japanese person’s preferences among baskets containing quantities of them could be taken to be the lexicographic ordering depicted in Figure 1.2. In that diagram, more of good 1 is always better than less, regardless of how much good 2 he might have; and given any value of x1 , adding quantities of good 2 does not increase his utility or make him better off. With the same dashed budget line of Figure 1.1 reproduced in Figure 1.2, self-interest would lead the individual to select basket B which has now become the rational choice. But motivation by self-interest or rationality cannot be a part of correct explanation because it does not represent the thought processes of the Japanese person. Rather, it imposes a cultural imperative on him, namely self-interest, which, as indicated in Essay 2 and elsewhere, does not arise in Japan in the same way as it does in the US. How, then, can the Japanese person’s behavior be said to have been explained when the categories and standards of that explanation make little sense in the cultural context in which his behavior emerged? To explain an individual’s behavior in terms of self-interest when, in fact, that individual himself believes his behavior is motivated by loyalty might be thought to be progress. But it does substitute a secondary, if not spurious motive for a primary one, and it therefore leaves something to be desired in so far as clarification is concerned. It follows that deeper, more accurate insight might be gained by discarding the idea of constrained utility or preference maximization and devising an alternative and culturally more relevant explanation that permits the explicit examination of the general role of

Introduction 11 loyalties in determining economic decisions and their properties. For this reason, the universalist view is hereafter discarded. There are, it turns out, other reasons for rejecting universalism. These are discussed in a small but growing literature that argues that culture matters in determining economic behavior in general, and that, in particular, the use of the assumption of maximization in explanations of individual economic activity is not always appropriate. A number of sources that make such points in various ways are cited in the essays that follow. But one of the more recent studies of this sort deserves special mention here. Henrich, et al. [4], examined ultimatum games played by participants from 15 small-scale societies with distinctive cultures. They found that the assumption of what would correspond to rational behavior (as the term is used here) in those games could not be sustained in any of the societies studied.9 That is, the use of the model of the “self-interested, … payoff-maximizing actor” is inappropriate in explaining individual behavior in those cases. Turn now to the localist category of explanation. In this context the determinants of the economic behavior of individuals are thought of as unique to their cultural heritages. Economic behavior, as previously noted, is social behavior that is culturally conditioned and that can only be understood in relation to the cultural environment within which it takes place. There is no such thing as an economic argument or model that, a priori, can be applied everywhere at all times. Whether or not a particular argument or model is relevant in a particular place or at a particular time depends on the cultural characteristics of the people to whom it might possibly be applied. There are difficulties with the localist position too. Individuals can act only on the basis of their own understandings of the reality that surrounds them, and those understandings are defined by the cultural categories and standards with which they are familiar.10 And, as has already been implied by the discussion to this point, the same categories and standards influence what the individual will accept and understand as explanation. That being the case, how is it possible for a person embedded in one culture to understand accurately and explain the economic behavior of a person embedded in another? The former would tend to perceive and

9 In these games, one individual is given a sum of money equivalent to the typical earnings from one or two days’ work in the society in which he lives. This person must then offer to give some portion of that sum to a second player (living in the same society). If the second player accepts the offer, each person keeps the respective amounts involved (the first person being left with the original sum minus what he offered to the second person); if the second person rejects the offer, neither player keeps anything. Communication between the players other than the amount of the offer and its acceptance or rejection is not permitted. The objective of the game is for the first person to make an offer that will be accepted by the second, thereby avoiding the outcome that both players end up with no money. In this case, rationality implies that the second player will accept any positive offer and the first player, knowing that, will offer as little as seems possible. 10 Even within the same community, individuals generally have different inheritances, personal histories, education and, perhaps, religious beliefs. They therefore may have different cultural patterns.

12 Introduction judge the behavior of the latter in terms of his (the former’s) own categories and standards, and the likely result is misunderstanding and error. To avoid this pitfall, the culture and behavior of the latter has to be translated into terms that make sense to the former and convey the correct meaning to him (Pálsson [8, Ch. 2]). And from the localist vantage point in which each person is trapped in his own culture, this simply cannot be done. Indeed, it follows from the localist position that the whole notion of using formal models as the basis for explanation has been to a considerable extent culturally bound to Western economic thought and may not make sense in the frameworks of some non-Western cultures. Regardless, the basic argument of this book as set out earlier may be seen as taking an approach in between universalism and localism that is predicated on the following two assumptions: (i) enough communication through conversation and other forms of human interaction can take place among peoples of different cultural backgrounds to permit translations of meanings across cultural boundaries, however difficult and imperfect, to be achieved. (ii) In light of assumption (i), both expositional (analysis-in-response-to-problems) and modeling (axiomaticdeductive) approaches to explanation are applicable in a meaningful way in all cultures. Of course, the latter is not intended to imply that general decision-making structures like the utility-maximization model of consumer behavior and the profitmaximization model of firm behavior are relevant a priori for explaining behavior in every culture. Although, upon investigation, that may turn out to be the case with respect to some cultures, the general presumption is that different explanations, that is different expositions or models, are required in different cultural circumstances. Returning to the content of Essay 2 (“Western Economics and the Economy of Japan” together with the two responses to criticism of its argument), consideration is given there, over and above that which has already been described, to the main differences in the dominant cultural traits (dominant, recall, with respect to their influence on individual behavior) now present in Western societies and Japan, and to the manner in which those traits rose to dominance in each case. In this context, the appropriateness of using postulates of rationality in explanations of economic behavior is examined. In addition to its relevance to the third stage of the basic argument put forth above, Essay 3 (“‘What are the Questions?’”) is partly concerned with the way cultural traits are represented in economic structures. Specific examples of the appearance of American cultural traits in models thought to be relevant to the American economy and American economic behavior are given. It is then suggested that, because those cultural traits are absent in Japan, the application of these models to the Japanese situation is inappropriate. Such inappropriate application not only results in the erroneous explanation that improper application generally implies, but it also carries the potential of faulty policy recommendations as well. Policy recommendations, of course, are a legitimate and necessary part of the discipline of economics. The essay further argues that the questions asked about an economy, if they are to be meaningful in the general society to which it is connected, must of necessity emerge in relation to the cultural environment in which that economy is located.

Introduction 13 The remaining essays relate to a number of issues that grow out of the basic argument presented above. Essay 4, “An Analytical Vision of the Workings of the United States Economy,” and Essay 5, “The Workings of the Japanese Economy,” when taken together are intended to show, in a very particular case and in a very detailed way, why models employed to explain how one economy works are not always appropriate and relevant for explaining the workings of another. Essay 4 emphasizes the manner in which culture – especially cultural values – impacts economic decision-making, and hence economic behavior, in the American economy. It identifies both specific American cultural traits influencing decision-making and the appearance of those traits in the standard Walrasian vision often used to explain the operation of that economy. The efficiency of outcomes derived from the vision along with so-called breakdowns or market failures are also considered. Much of what is presented in the essay is well known. Its main purpose is to present a foil against which Japanese decision-making, Japanese economic behavior, and the Japanese economy can be compared. That comparison is the subject matter of the essay that follows. Thus Essay 5 (“The Workings of the Japanese Economy”) begins by demonstrating that the Walrasian vision should not be used as the basis for explaining how the Japanese economy works because the cultural prerequisites for doing so are not present. Japanese cultural values would seem to require the creation and use of an alternative explanatory vision. In addition, certain features of the Japanese industrial structure, the Japanese labor market and distribution system, and the manner in which the Japanese government participates in the Japanese economy all serve to distinguish the Japanese economy more completely from the American economy and also to separate the relevant explanation of it farther from the Walrasian vision of economic behavior. It is not surprising, then, that the Japanese economy cannot and does not live up to the standards of efficiency applied to the American economy and derived from the traditional Walrasian assumptions. Essay 6, “Explaining the Japanese Economic Miracle,” argues that, contrary to popular opinion, the Japanese economy was not, by Western notions and standards, highly efficient during the period in which the so-called economic miracle took place. Rather, the sustained rapid economic growth of that period, which actually occurred in the face of substantial “inefficiencies,” was made possible in part by significant “offsetting” effects. The latter inefficiencies, cited (but not described) in relation to the American economy in the summary of Essay 5 above, were inefficiencies by Western standards only and arose from the presence of cultural traits quite distinct from those of the West. A major objective of Essay 7, “The Role of Optimization in Economics,” is to expose the linkage between the appearance and employment of the mathematics of optimization 11 in economics on the one hand, and the cultural forces that significantly influence individual behavior in Western economies on the other. The essay begins by describing the history and early use of both constrained and

11 As used here, “optimization” is essentially the same as the term “maximization” employed earlier.

14 Introduction unconstrained optimization as mathematical tools in economics. It then identifies three broad categories of current applications of optimization in economics, namely, (i) in explanations of economic behavior; (ii) to define societal ideals; and (iii) as a convenience in formal argument. Finally it considers two questions in relation to these categories: why has optimization become so important in economic analysis, and why is it likely to remain so? In addressing the latter questions, the essay argues, in part, that optimization has become important in explanation through a combination of two facts. These are first that explanatory models using optimization were created by Western economists primarily to explain economic behavior in a Western cultural environment. And second, it is a consequence of the peculiarities of Western culture that optimization is highly relevant and significant in explanations of Western economic behavior. The essay also suggests that since societal ideals depend to a considerable degree on cultural values, a parallel argument arises with respect to the use of optimization in defining them. In these respects, the presence of mathematical optimization in economics has been culturally driven. Essay 8, “Economics and the Principle of Uniformity,” contends that both aspects of the principle of uniformity, namely, substantive uniformitarianism or the statement that the rates of change of processes are constant through time, and methodological uniformitarianism or the more general assertion that the laws governing phenomena are invariant over time and across space, although originally applied to explain the physical world, have no place in economic explanation because they preclude the consideration of significant components of economic reality. In particular, cultural differences are likely to cause economic laws to vary across space, and the realities of historical time are likely to render both economic laws and rates of change nonconstant through time. The essay is included in this collection because the latter declaration concerning cultural differences implies that, accepting the present book’s basic argument that cultural variation needs to be accounted for in economic explanation, economic laws must necessarily differ in meaning and significance from laws in the physical sciences. The last two essays provide examples of formal models of, respectively, the firm and the consumer that do not rely on rationality or maximization as the driving force behind decision-making and behavior. The first of these (Essay 9), entitled “Cultural Variation in the Theory of the Firm,” presents a formal model of the firm that includes the possibility of firm and employee-on-the-job decision-making based not only on profit and utility maximization, but also on other criteria as well. The alternatives, of course, are relevant and significant when explaining firm activity in cultural environments in which self-interest is not the primary force driving human behavior. In the context of this single model, three types of firms are defined and their properties compared: the Western firm in which profit and utility maximization propel individual behavior, the Japanese firm in which profit maximization and utility maximization are absent, and the clan. The third of these types is a combination of the first two. Contrasts among the three kinds of firms are drawn with respect to both their structures and efficiency.

Introduction 15 “Choice and the Explanation of Choice Behavior” (Essay 10) shows how alternative, culturally derived motivational impulses can be substituted for maximization in the theory of choice. Several possibilities are considered, including the replacement of preference optimization by such propellants as the selection of the “second best” or the “central” option. It is argued that although all choice behavior, even that consistent with the alternatives considered, can ultimately be understood as satisfying the criterion of rationality, richer and more meaningful explanation is obtained by focusing on culturally significant different motivations when the latter turn out, in particular environments, to be more important than self-interest. The thrust of the essays of this volume, then, may be summarized as follows: if explanation in economics is to be correct, culture must figure significantly into the determination of both its structure and content. In particular, expositions and models developed to explain economic behavior in one culture need be neither appropriate nor relevant in attempting to explain behavior in another. Although this may mean that self-interest (rationality or maximization) has to be discarded as the driving force behind economic behavior in some cases, it is possible to construct expositions and models based on alternative forces – even if it means that, because they may not be invariant over space, the laws so obtained do not have the same properties as physical laws. It is obvious, of course, that explanation in economics has to be correct if it is to have any meaning or importance. Incorrect explanation provides no insight into the workings of reality. It can result in policy recommendations that induce substantial harm rather than provide help. It is a sterile exercise that wastes valuable resources. To avoid incorrect explanation it is, in part, necessary, as has been and will continue to be said over and over again in this book, to be sensitive to the cultural backgrounds of the individuals or economy under investigation.

References 1. Fehr, E. and S. Gächter, “Fairness and Retaliation: The Economics of Reciprocity,” Journal of Economic Perspectives 14, no. 3 (Summer, 2000), pp. 159–181. 2. Gudeman, S., Economics as Culture (London: Routledge & Kegan Paul, 1986). 3. Guiso, L., P. Sapienza, and L. Zingales, “Does Culture Affect Economic Outcomes?” Journal of Economic Perspectives 20, no. 2 (Spring, 2006), pp. 23–48. 4. Henrich, J., R. Boyd, S. Bowles, C. Camerer, E. Fehr, H. Gintis, and R. McElreath, “In Search of Homo Economicus: Behavioral Experiments in 15 Small-Scale Societies,” American Economic Review 91, no. 2 (May, 2001), pp. 73–78. 5. Katzner, D.W., Analysis without Measurement (Cambridge: Cambridge University Press, 1983). 6. ——, Unmeasured Information and the Methodology of Social Scientific Inquiry (Boston: Kluwer, 2001). 7. Nisbett, R.E., The Geography of Thought: How Asians and Westerners Think Differently … and Why (New York: Free Press, 2003). 8. Pálsson, G., The Textual Life of Savants (Chur: Harwood, 1995).

16 Introduction 9. Schultz, T.W., Economic Crises in World Agriculture (Ann Arbor: University of Michigan Press, 1965). 10. Smith, A., An Inquiry into the Nature and Causes of the Wealth of Nations (New York: Random House, 1937). 11. Stigler, G.J. and G.S. Becker, “De Gustibus Non Est Disputandum,” American Economic Review 67 (1977), pp. 76–90. 12. Stiglitz, J.E., “Rational Peasants, Efficient Institutions, and a Theory of Rural Organization: Methodological Remarks for Development Economics,” The Economic Theory of Agrarian Institutions, P. Bardhan, ed. (Oxford: Oxford University Press, 1989), pp. 18–29. 13. Tylor, E.B., Primitive Culture, v. 1 (London: John Murray, 1920).

2

Western economics and the economy of Japan1

An important issue that is relevant to all economic inquiry concerns the ways in which the culture of any particular society affects the economic behavior of individual agents in it. Equally important is the matter of the extent to which the cultural backgrounds of the investigators making the inquiry can and do influence the explanations and understandings of the workings of the economy of that society as developed by those investigators. Generally, a continuum of possibilities can be imagined. At one extreme would be the idea that, like the physical world, there are “universal” laws governing all economic behavior which do not vary across cultures and, once they are fully understood, would be readily accepted by all investigators, regardless of their cultural backgrounds.2 At the opposite extreme would exist a notion of “full” cultural determinism of both economic behavior and explanations generated to understand that behavior. From this latter perspective, neither explanations of behavior nor the behavior itself would be, as a rule, transferable in unaltered form from one cultural setting to the next.3 Without attempting to identify a particular point on the continuum at which economic inquiry might actually be placed, this essay argues that cultural factors can play a significant role in economic behavior as well as our understandings of that behavior.4 It follows, then, that an investigator who ignores 1 Journal of Post Keynesian Economics 21, no. 3 (Spring, 1999), pp. 503–521. © M.E. Sharpe, Inc. Reprinted in slightly modified form with permission. The original draft of this essay was written while the author was visiting the Institute of Social and Economic Research at Osaka University in the Spring of 1993. He would like to thank Tatsuo Hatta, Charles Yuji Horioka, Alfred B. Hudson, Hajime Miyazaki, Takashi Obara, and Douglas Vickers for many helpful comments and suggestions, and absolve them of any responsibility for the result. 2 See, for example, the quotation attributed to K. Takeuchi and translated by Morris-Suzuki [30, p. 1]. 3 Kirby [22, pp. 409, 414, 417] hints at the existence of such a position. See also Greif [10]. 4 It should be borne in mind that to argue in this way is not to favor cultural relativism in its most general sense. The latter is a broad philosophical approach that applies to all social and behavioral inquiry in all contexts. In particular, it provides one answer to the question of what is, from an epistemological standpoint, the best way of obtaining knowledge of human behavior in general (see, for example, Harris [12] and Sahlins [38]). But the only point at issue here is much narrower and more specific, viz., that culture may make a difference in generating and understanding the economic behavior of individuals from one environment to the next. A similar position has been taken by, for example, North [32, p. 4] and Gudeman [11].

18 Western economics and the economy of Japan these factors can expose his investigation to incompleteness, irrelevance, and even error.5 Now culture may be defined as a system of meaning or orientation that provides a guide to perceiving reality, and, in addition, furnishes standards for interpretation, judgment, and action. It might also suggest a vision of the way the world ought to be. As such, culture clearly impinges on economic investigation and, at the level of the individual agent, on the determination of both economic objectives and the motivating forces behind economic behavior. Its influence therefore permeates all economic analysis and all economic activity. In particular, it plays a significant role in the answering of the fundamental “what?” “how?” and “for whom?” questions posed in all economies, along with the derivative queries surrounding such things as income distribution, employment, investment, and growth. Because much of contemporary economic analysis was constructed by Western economists to explain Western economic phenomena, the importance of cultural factors has tended to remain hidden from view. Living in the same cultural milieu that forms the background of the economy he is studying, a Western economist building a model to explain economic observations is unlikely to perceive as noteworthy the cultural imperatives that drive the economic behavior he observes. It is only when he attempts to employ those thought forms to explain observed economic behavior set in a different cultural environment that a dissonance suggesting the power of culture emerges. Thus, for example, Leff and Sato [25] have discovered that the proposal of Stigler and Becker [43] to assume homogeneous preferences in constructing economic models does not seem to be appropriate when attempting to explain observed variations in aggregate savings and investment behavior over cultures. Lau and Yotopoulos [24] have argued that country “idiosyncrasies,” which may, in part, be culturally attributable, do matter in estimating a production function for world agriculture. And De Long [7] has shown that the convergence of aggregate productivity levels across nationalities, and hence cultures, although expected in theory due to the public-good nature of modern technology, has not occurred in practice. To enhance the visibility of such tensions, then, present argument is couched in terms of the application of Western Economics to analyze an economy that exists in a cultural tradition quite foreign to that of the West, namely, Japan. A reason for focusing on Japan itself is the existence in the literature examining the Japanese economy of an implicit debate on similar and related issues that occasionally bubbles to the surface.6 On one side are the neoclassical economists who unabashedly apply Western neoclassical economics to the study of Japanese economic life. To them the behavior of economic agents, regardless of cultural backgrounds, is understood to be rational in that, among other things, it is interpretable in terms of, say, repeated maximizations subject

5 The same point has been made in a more specific context by van Wolferen [47, p. 8]. 6 This debate has been described, in part, by such writers as Aoki [1, pp. 2–3, 300–306] and MorrisSuzuki [30].

Western economics and the economy of Japan 19 to market-, endowment-, and technologically determined constraints. Ignoring rationality, the other side explains Japanese economic behavior with respect to such cultural factors as the interactions within and among small groups, the effort to achieve group consensus, employee loyalty and employer paternalism, and the pervasive influence of the workplace on private lives. There are, in addition, a growing number of in-between positions such as those of Aoki [1, p. 3] and Gerlach [9, pp. 19–20, 25–27]. Subsequent discussion enters into this debate in that it attempts to indicate why cultural factors are important, and why, in the case of understanding the Japanese economy, the use of Western economic thought forms without modification can run into difficulties. In light of Japanese history, it is reasonable to suppose, and such an assumption is quite adequate for present purposes, that within nineteenth and twentieth century Japanese society, cultural continuity has prevailed in that certain cultural elements have modified only very slowly over time, and insignificantly across society’s members.7 Hence variation in these elements cannot generally play a large explanatory role in intra-societal dynamic and cross-sectional economic analyses of Japan during this period. In particular, their use in explanations of such phenomena as, say, the high savings rates in post-World-War-II Japan, are not very satisfying since, in spite of the more or less constancy of the referred-to cultural elements throughout the twentieth century, pre-war Japanese savings rates are much lower.8 Of course, one might argue that cultural factors such as these generally contribute to a predisposition towards high savings rates in Japan,9 but this is quite different from asserting that they are a primary cause of the higher postWorld-War-II rates. In any event, cultural effects can be recognized as potentially having a major impact on economic affairs in general only when the same kinds of economic activities, although capable of arising in the face of sharply different values of specific cultural variables, do not occur in the same qualitative way or to the same degree as the values of those culture variables modify. And under such conditions, the central issue is the extent to which culture matters, or is thought to matter, in determining and understanding the economic behavior that has been observed to emerge. The essay begins with a brief discussion of the justification for using a broad swath of Western economic analysis in explanations of Western economic behavior. Then, in demonstrating the collapse of that justification in reference to explaining economic behavior in Japan, it is shown how Japanese culture can impose a significant barrier to the outright transfer of this analysis into the Japanese economic context. Although, to be concrete, attention focuses almost exclusively

7 This is not to deny the presence of cultural change over time or cultural differences across individuals. But at the level of discourse entered into here, such change and differences have been, and continue to be, relatively minor with respect to the components at issue, and can be safely ignored. 8 Ito [14, pp. 262–264]. 9 Some of the investigations summarized by Horioka [13, pp. 54–60] can be interpreted in this way. See also Bellah [2], the relevant parts of which are outlined in Section II.

20 Western economics and the economy of Japan on the Western notion of self-interest as it arises in neoclassical microeconomic theory, the argument itself applies quite generally.10

I Western Economics, or the thought-forms employed by mostly Western scholars attempting to understand and explain various facets of what they perceive to be the “Western” economic world, covers an immense variety of subjects and relies on a rather broad array of techniques. But at its core sit a very small number of visions of the way a Western economic world or Western economy might operate.11 These visions tend to be, at the same time, both independent and overlapping, and each gives rise to its own unique collection of models of economic behavior that serve, in part, to articulate the vision from which the models spring. Among them, for example, is the well-known Walrasian vision of the microeconomy. Regardless, all such visions and models share certain characteristics that, in a vague sort of way, define their “Westernness.” One of the Westernness-defining characteristics of Western economic visions is the methodological employment of the rationality of constrained maximization. This has led the visions to focus substantially on a very particular notion of self-interested economic behavior. Generally, however, self-interest is a motive. It determines an agent’s objective, say, to maximize his income, or to minimize certain costs. The self-interest motive itself may be purely selfish, altruistic, or may include elements of both. (For example, self-interest need not necessarily imply maximizing one’s own income and, alternatively, maximizing someone else’s income may result in maximizing one’s own.) Rational behavior, broadly considered, is behavior which (i) is designed to achieve an objective determined by self-interest and (ii) takes place in an environment or model in which certain axioms are satisfied. In this sense, rationality is a reflection of self-interested behavior. But Western economics tends to look on self-interest in a very special and narrow sense, namely, in terms of the pursuit of what, in the mind of the individual, is most preferred. And this leads quite naturally to the representation of self-interest in Western economic models as the rationality of constrained maximization mentioned earlier. Of course, no one would deny that, in actuality, other motives leading to the pursuit of less preferred alternatives may play a role in generating

10 The experiments of Kachelmeier and Shehata [19] have suggested that, at least in laboratory markets with oral double auction rules, self-interested behavior extends across the cultures of Canada, China, and the US. These results, however, do not contradict the argument that culture matters in determining and explaining individual behavior for at least three reasons: First, it is entirely possible that self-interest is a motivating force behind individual behavior in Canada, China, and the US, but not in Japan. Second, the oral double-auction rules may create an environment in which cultural factors are rendered inoperative. That is, the experiments may be too coarse to detect the impact of cultural differences. And third, the individuals participating in the experiments may not be representative of their societies at large. 11 Of course, specific details of these visions and their applications may vary across scholars of different Western cultures, and even among scholars within the same culture.

Western economics and the economy of Japan 21 economic behavior. But the point of economic analysis is to abstract from the real world and thereby to focus on what are considered to be reality’s most important features.12 And in that context, self-interest depictable as constrained maximization is almost universally taken to be the primary motivating force governing individual behavior in Western economies. In what follows it will be convenient, in drawing attention to the Western approach to these matters, to equate the meanings of the terms “self-interest,” “rationality,” and “constrained maximization” with each other. This, however, is a pedagogical device only. It is not intended to imply that agents who do not maximize subject to constraint are, in the more general senses of the terms, neither rational nor motivated by what they perceive to be their own self-interest. In models of Western economies, self-interested behavior is expressed in socalled “postulates of rationality,” namely, for example, the assumptions that each consumer makes purchases and sells endowments so as to maximize utility subject to his budget constraint, and that each firm hires inputs and produces and sells outputs so as to maximize its profit. However, Caldwell [5, pp. 156–157] has argued that, as a practical matter, it is not possible to test empirically the rationality postulate of constrained utility maximization by consumers either directly as it is stated or indirectly through one or more of its implications. For any such test can only be set in a partial equilibrium framework defined by ceteris paribus assumptions. And since observations are made through time, the latter assumptions include such requirements as “preferences are constant and information is perfect” across the period over which the observations are taken. Thus a disconfirming instance can always be judged undamaging to the postulate of rationality on the grounds that “ceteris was not paribus” or, in other words, that, say, preferences might have changed from one observation to another, or information might not have been perfect. A parallel argument could be made for profit maximization on the part of firms, as well as for other postulates of rationality appearing in models that emanate from the same or alternative visions. Hence formal corroboration of rationality postulates in economic models is beyond the economist’s means. How, then, can the appearance of self-interested behavior in the visions of the Western economy and the expression of that behavior in terms of the various postulates of rationality be justified? There is, after all, no single, universal, prescriptive scientific methodology generally acceptable to economists, such as logical empiricism or falsificationism, that would provide the basis for such a justification.13 In the past economists have attempted to vindicate their use of rationality postulates with reference to terms and phrases like “‘self-evident propositions,’ ‘axioms,’ ‘a priori truths,’ ‘truisms,’ ‘tautologies,’ ‘definitions,’ ‘rigid laws,’ ‘rules of procedure,’ ‘resolutions,’ ‘working hypotheses,’ ‘useful fictions,’ ‘ideal types,’ ‘heuristic mental constructs,’ ‘indisputable facts of experience,’ ‘facts of immediate experience,’ ‘data of introspective observation,’

12 See, for example, Marshall [27, p. 461]. 13 Caldwell [5, pp. 5, 89–90, 223, 244].

22 Western economics and the economy of Japan ‘private empirical data,’ ‘typical behavior patterns,’ and so on.”14 Clearly, these characterizations are not all equivalent to each other. But what matters here, according to Machlup [26, p. 17], is not so much the logical role played by the postulates of rationality themselves in the structure of any particular argument. Rather, it is the fact that most Western economists can introspectively conceive of “sensible” persons or firms acting, at least sometimes and in some situations, in the manner that the postulates prescribe. In other words, rationality or selfinterested behavior is a legitimate assumption because it seems reasonable to Western economists investigating the Western economy. This conclusion, however, only pushes the issue back to another level. For it is still necessary to ask why self-interested behavior “seems reasonable to Western economists investigating the Western economy.” A plausible answer to the latter question can be found in reference to the cultural heritage of Western economists. That is, the pursuit of self-interest makes sense to Western economists as a motivating force for economic behavior because it is a part of the cultural imperatives of their own existence. In the conduct of their private lives, Western economists often see themselves as motivated, in part, by self-interest. They also expect it of, and assume it is present in the behavior of others. This is the way they have learned to understand their private worlds; it is a consequence of their enculturation into Western society. It is therefore quite natural for them, as scholars, to be comfortable with the generalization that, at least as a crude approximation, the primary motivation for economic behavior is self-interest. Now it is clear that in all societies, even the most primitive ones, whenever scarcities relative to needs and desires are perceived by members, self-interested behavior on the part of individuals is a possibility, if only as a means for survival.15 But such random or unsystematic perceptions of scarcity and acts of self-interest are not the issue here. What matters, insofar as general explanations of behavior are concerned, is that these perceptions and acts are accepted by the members of a society as having become systematic and institutionalized, and hence exist as a part of the cultural milieu within which explanations of behavior have to be constructed. And such acceptance or institutionalization has come about in the West only relatively recently.16 Some years ago, Max Weber [45] described how the institutionalization of self-interested behavior in Western economies might have occurred. Weber was actually concerned with the causal relationship between Protestantism and the rise of capitalism. And it is his sweeping assertion that the latter generally emerges from certain forms of the former that has been debated and, to some extent, controverted in the critical literature.17 But one part of Weber’s

14 15 16 17

Machlup [26, p. 16]. See, for example, Sahlins [37, pp. 13, 14, 204, 205]. Polanyi [33, pp. 66, 67]. For example, Samuelsson [39]. See also Samuelsson’s survey of the critical literature in his first chapter.

Western economics and the economy of Japan 23 argument that does not rest on the validity of this general assertion calls attention to the trigger that led, under the conditions of the time, to the institutionalization of self-interest into Western culture. For Weber, self-interested behavior is included as part of a broader concept that he referred to as the “spirit of capitalism.” Although, in its pre-institutionalized form, that spirit was present across the world and throughout the ages, his argument suggests that it was catapulted into the heart of Western culture during the seventeenth century through the dogmas of what Weber understood to be Christian asceticism. These dogmas, reflecting the rediscovery of the significance of the individual during the Reformation, and the simultaneous collapse of the social and economic constraints formerly imposed upon him by the church, rationalized the pursuit of wealth and self-interest by making them a means to a higher religious end. With the consciousness of standing in the fullness of God’s grace and being visibly blessed by Him, … [any] man [living at the end of that century], as long as he remained within the bounds of formal correctness, as long as his moral conduct was spotless and the use to which he put his wealth was not objectionable, could follow his pecuniary interests as he would and feel that he was fulfilling a duty [to God] in doing so. [45, pp. 176–177] Gradually, over the years since, the religious goal has been forgotten, and efforts to justify the pursuit of wealth and self-interest on religious grounds have been abandoned [45, pp. 181–182]. The means have become the ends, and the pursuit of self-interest, in particular, has gained acceptance in Western culture as a goal by itself. In Western economics, the introduction of self-interested behavior dates at least to Adam Smith who wrote, “It is not from the benevolence of the butcher, the brewer, or the baker, that we expect our dinner, but from their regard to their own interest” [42, p. 14]. Although it is an open question as to the extent to which utilitarianism may have been influenced by motives other than pure selfishness, Bentham provided a language for expressing the pursuit of self-interest by the consumer: “Nature has placed mankind under the governance of two sovereign masters, pain and pleasure. It is for them alone to point out what we ought to do, as well as to determine what we shall do. On the one hand the standard of right and wrong, on the other the chain of causes and effects, are fastened to their throne. They govern us in all we do, in all we say, in all we think: every effort we can make to throw off our subjection, will serve but to demonstrate and confirm it” [4, p. 1]. And, in the vocabulary of Bentham, Jevons gave an explicit statement of a postulate of rationality for the consumer: “The theory which follows is entirely based on a calculus of pleasure and pain; and the object of Economics is to maximise happiness by purchasing pleasure, as it were, at the lowest cost of pain” [15, p. 23]. It should also be noted that by this time a postulate of rationality for the firm, namely, the setting of price to achieve “the greatest possible profit,” had already been proclaimed by Cournot [6, p. 56]. And in any case, from here on

24 Western economics and the economy of Japan the spread of these and other rationality postulates throughout Western economics was quite rapid.

II Consider, now the case of Japan. For all their relevance and usefulness in thinking about Western economies, self-interest and postulates of rationality would appear to be less suitable hypotheses for understanding Japanese economic behavior and the Japanese economy. Although models involving maximization may still have pertinence to the Japanese experience, such maximizations, were any of these models to be invoked, would be inappropriate as representations of rationality or self-interest. This is because, unlike explanations of Western economic behavior, the cultural basis for the construction of analyses founded on self-interest is simply not there. Indeed, there are not now, nor have there ever been, religious dogmas rationalizing the pursuit of wealth and self-interest waiting to be institutionalized or catapulted, in a Weberian sense, into the center of Japanese culture.18 Although an analogous institutionalization of religious dogmas did take place in Japan during the eighteenth and nineteenth centuries, the latter dogmas, according to Bellah [2], centered on certain political values that had little to do with self-interest. And like self-interest in the West, those political values became the central motivating forces driving individual behavior. Chief among the political values that were to be institutionalized in eighteenth and nineteenth century Japan were service and loyalty to the collective, whether it be family, fief, or state. Along with service, achievement through the performance of that service was also included. Membership in a collective was symbolized by loyalty to its head. Harmony, willingness to compromise, and unaggressiveness were significant derivative values because their absence not only implied disloyalty, but could also interfere with the attainment of the collective’s goals. In such a setting, economic activity was suspect unless it was seen as a contribution to the functioning of the collective. Work as an end by itself was not important; but work as an expression of the selfless devotion to the collective was highly regarded. The presence of these political values has roots in Japanese antiquity and the values themselves, though not yet institutionalized, were subsequently reinforced and sharpened with the merging of Confucianism into the religious system during the middle ages. By the start of the Tokugawa era in 1600 their primacy was well established in the religious dogmas of Japanese society. Institutionalization, however, was still to come. Generally, it was mostly Buddhism and Confucianism that, in Japan, provided a context of meaning for the central values of society, and attempted to meet the threats to those values that arose from human frustrations. Although during the Tokugawa era (1600–1868) no new religious orientations developed, the strains and tensions that emerged from the increasing complexity of Japanese life resulted

18 See Weber [46, pp. 266–272].

Western economics and the economy of Japan 25 in considerable intensification and splintering of the religious system. Rather than attempting to weaken old values and replace them by new ones, religious movements developed that redefined the sacred by imposing more forceful and acute forms of the old. The main values affected were the political ones identified above which now became the focus of fervent religious activity. These religious movements spread to such an extent that their ideas became institutionalized. Thus, by the early years of the Meiji era that began in 1868, the religious and political had merged with the emperor at the center. Family, nation, emperor, parents, and superiors all had acquired a sacredness about them in a religious hierarchy. Obligations to the emperor had come to take precedence over all other obligations. Work became the means of fulfilling obligations to superiors. The obligation to maintain honor required frugality and the keeping of consumption to a minimum, and honesty, credit worthiness, and the provision of quality in business. Fulfilling such obligations was enthusiastically pursued because it meant becoming identified with the ultimate religious reality. The impact of the institutionalization of these values, in both their general form and in their specific economic manifestation, was felt throughout Japanese society. For example, their infiltration into the Japanese system of labor-management relations during the twentieth century has been described by Kinzley [21]. In this case, certain elites consciously looked to the institutionalized values as the best means for achieving industrial harmony. It was their efforts to steer and reshape those values so as to meet new needs that eventually led to such labormanagement innovations as lifetime employment, worker participation, docile enterprise unions, and corporate paternalism. None of this is to argue that the Japanese do not engage in self-interested behavior (in the narrow, Western sense) on numerous occasions. But it does suggest that self-interest is not the primary force that motivates them. By extension of the argument of Section I, then, analyzing the Japanese economy would seem to require, if focus on the most significant features of Japanese actuality is to be maintained, the setting aside of those features of Japanese reality that direct attention towards self-interest. Thus there is little justification for introducing self-interest or rationality as a major assumption in explanations of Japanese economic behavior. Moreover, to use models of utility and profit maximization that represent rationality or self-interest as explanations of the behavior of Japanese agents is to impose a cultural imperative of dubious relevance. Such things as the fulfillment of obligations and the furtherance of group harmony apparently play a more central role in motivating their behavior than the pursuit of self-interest.19 Indeed, competition for social rank among firms rather than profit maximization often determines which investments are undertaken and, instead of being paid according to the market value of their contributions to the productivity of the firm, an activity which, in Western Economics, is a necessary

19 See also Benedict [3] and Nakane [31]. In Sen’s language [40, pp. 326–329], agent behavior in Japan generally emerges from commitment rather than from personal welfare considerations.

26 Western economics and the economy of Japan implication of various assumptions including profit maximization, employees are paid according to their seniority.20 A lack of cost minimization for specified levels of output by firms, another necessary requirement for profit maximization under Western assumptions, has also been suggested in another context by Nakane [31, pp. 82–83]. There is another obvious way to arrive at the conclusion that the employment of Western models to understand Japanese economic behavior might not be culturally warranted. For it is a part of the cultural environment of Japan that long-term relationships among individuals within groups are highly regarded and sought after. Indeed, the manifestation in the Japanese economic sphere of the desire for long-term relationships in terms of lifetime employment, the maintenance of the Keiretsu enterprise groups and other networks of firms, and the loyalties among particular manufacturers, wholesalers at various levels, and retailers, has often been observed.21 To the latter might also be added the loyalties of many Japanese consumers to specific retail outlets. Since the pursuit of these long-term relationships often overrides self-interested considerations in decisionmaking, Western models of profit maximization by firms and constrained utility maximization by individuals, along with the “efficient” allocation of resources secured through the operation of the competitive price mechanism in models of general equilibrium, are once again compromised. Moreover, to the extent that, to preserve long-term relationships, buyers knowingly do not buy from sellers offering the lowest price, the price mechanism is further diverted from allocating resources in the same efficient manner as typically envisaged by the models of Western economists. One might think that, with the decline in economic growth since the speculative bubble in land prices burst in 1990, the Japanese economy has and will continue to become “more Western.” That may be so. But it does not in any way vitiate present argument. Since the sixth century ad the general pattern of Japanese history has been one of periods of relative isolation or destruction followed by periods of borrowing or copying from abroad. Some periods of isolation were longer than others; in some periods, or parts of periods, the isolation was more intense than in others. But in all cases, the isolation or destruction left Japan far behind the advanced civilizations of the world in technology, institutional development, and personal skills in a wide variety of areas including, in particular, the economic. The subsequent periods of borrowing and copying allowed her to catch up. More importantly, borrowed or copied elements were always adapted so as to fit into what was already in Japan and, once adapted, they, like the civilization as a whole, advanced on their own. In this way, a unique culture and a unique way of life has both evolved and been preserved for over 1,500 years. It is safe to assume, then, that the same will apply to whatever modifications are introduced to make the Japanese economy more Western in the current situation. This, of

20 Nakane [31, p. 90] and Ito [14, p. 211]. 21 For example, Ito [14, pp. 177–196, 214–226] and Gerlach [9].

Western economics and the economy of Japan 27 course, means that, without an overhaul of prevailing institutionalized cultural mores and norms (which is not, at present, anticipated), efforts to Westernize the Japanese economy will, as in the past, be inhibited by their presence. There is reason to expect, then, that in the long run these mores and norms will assert themselves once again and lead to a unique adaptation of the newly introduced Western economic characteristics into the Japanese system. Thus the fulfillment of obligations, the furtherance of group harmony, loyalty, etc. will most likely remain more significant in Japanese culture than self-interest. In that event, the cultural basis for self-interest or rationality will still not exist, and models that include rationality to explain Japanese economic behavior will still not be appropriate. Clearly, models that replace the imperative of self-interest by, say, the subjugation of individual preferences to group harmony, are likely to have a different character from those that employ standard postulates of rationality. Such models would have to include, in the case of the example cited, descriptions of how the subjugation of individual preferences occurs, along with the nature of the individual behavior that emerges from it. The latter behavior, obviously, would have to be based on some specific motivating force other than selfinterest. Nevertheless, like their rational-behavior counterparts, these models would still contain concepts, assumptions, and derivations. The deduction of “empirically falsifiable propositions” would remain possible,22 and the models, or their appropriate parts, could be fitted to, and hence used to explain, observed data. But in contrast to models of rational behavior, the latter constructions would reflect the subtleties of a different cultural environment. To illustrate the basis upon which such models might be constructed, consider the following example modified from Sen [41, p. 501]. Imagine an individual, who prefers more cake to less, is presented with several pieces of cake from which he is asked to choose one piece for himself. Suppose also that the individual lives in a cultural environment in which it is not proper to select the largest piece. Thus, in spite of his preference for the largest, he always takes the next-to-largest piece from any assemblage of pieces presented to him.23 Against previously stated criteria, this, evidently, is nonrational behavior since the choice made is never utility-maximizing. To model it requires a different set of assumptions from those employed to model rational behavior.24 In any event, from the perspective of the significance of culture in understanding Japanese economic behavior elucidated here, Western economic models, as well as any other model based on individual self-interested behavior, would not seem to be especially relevant for understanding how the Japanese economy works.

22 It should be noted that models based on assumptions of rational behavior can still generate empirically falsifiable propositions even though the postulates of rationality themselves cannot, as a practical matter, be empirically tested. 23 By varying the cultural requirement, alternative decision rules such as the third largest, etc. are possible. 24 See Katzner [20] appearing as Essay 10 below.

28 Western economics and the economy of Japan Parallel conclusions have been reached by others in different contexts. For example, they are implicit in Prestowitz [34, p. 23] who complains about the inability of American trade negotiators to recognize that the Japanese economy neither functions nor is oriented like that of the US. More explicitly, Fallows [8] sees the tensions of the last 40 years between the US and Japan as stemming from divergent perceptions of reality and approaches to explanation that arise from contrasting historical experiences. Thus the American past has left Americans with the notion that the purpose of an economy is to provide for individual material well-being, whereas Japanese (and, more generally, Asian) experience has resulted in the Japanese (Asian) belief that the purpose of the economy is to further political ends. Temin [44] suggests that Japanese industrial practices are partly the result of the collective orientation of its culture. And Johnson [16], arguing the inadequacy of neoclassical economic theory (which he regards, at least in part, as left over ideological baggage from the Cold War [18]), explains the Post-World-War-II Japanese economic “miracle” as a consequence of successful government direction of Japanese economic development.25 In spite of the foregoing, and in opposition to the argument of this essay, it should be pointed out that there are studies relying on self-interest as the primary motivating force behind Japanese action. A well-known illustration is Ramseyer and Rosenbluth’s [35] explanation of individual and group behavior of the oligarchs who ran Japan in the 65 years or so after the Meiji Restoration of 1868. In part, Ramseyer and Rosenbluth are concerned with the failure of the oligarchs as a group to maintain themselves at the center of power in the Japanese government. For Ramseyer and Rosenbluth, the behavior of individuals everywhere is driven mainly by self-interest. This meant, with respect to the oligarchs, that each individual would do his utmost to stay in office. The oligarchs as a group failed to maintain their power and position because individual selfinterest led to noncooperative behavior amongst themselves. Cooperation would have required them to maximize their collective political rents and thereby enhance the power, wealth, and status of the group. But as in a cartel, each was often able, by cheating, to increase his own share of political rents beyond what he would receive by acquiescing to the group. Thus the pursuit of self-interest by each oligarch individually resulted in the loss of power and eventual destruction of the group itself. It is also interesting that, as partly suggested at the outset, Japanese investigators studying the Japanese economy are themselves split on the usefulness and

25 Fallows [8, p. 4] essentially accuses Americans (including American economists) of hubris partly for their tendency to “… view the world as one vast potential extension of their own culture ….” Johnson [17, p. 49] is more caustic: “In talking about economic issues in the Asia-Pacific region, one must begin by largely discounting or ignoring the opinions of economics professors in Englishspeaking countires. This is because they have failed to anticipate the growth of great wealth in the area, still cannot adequately explain it, and are today concerned more with defending their arcane theories than with studying the nature and potentialities of Asian capitalism.”

Western economics and the economy of Japan 29 applicability of Western economic ideas. There are many reasons why this division might have developed, including the cross-fertilization of ideas and the intermixing of cultures that often occurs among members of the international scholarly community. But regardless of how it came about, on one side Morishima, for example, has asserted [28, pp. 2–3] that Countries may share a capitalist economy, but their historical experience and cultural traditions differ, and the lives, beliefs and modes of behaviour of their people are certainly not the same …. [Thus] countries can … react in a different manner to the same stimulus. Where the difference in reaction is not more than a matter of degree it is possible to handle these economies with models of the same type using different numerical values for the coefficients (parameters) constituting the framework, but where the reactions are qualitatively different they have to be analysed using quite separate models. In another place, Morishima has also said the assumption that “the individual always acts in a rational manner” is “far removed from reality” in Japan [29, p. 202]. Similar thoughts have been expressed somewhat differently, and in a Weberian context, by Murakami: There are two paths by which the nexus of values for industrialisation may be created. One is the path which reinforces instrumental activism with individualism. This is achieved through the mediation of Protestanttype religions which, while separating man from God, place one in direct communion with the other. The alternative path is a path of collectivism where the power of the political system is used to direct society as a whole towards industrialisation.” 26 On the other side are people like Koike [23, pp. 47–48] who argue that, in spite of cultural pressures, the behavior of Japanese workers in large corporations is actually a manifestation of rational behavior. Thus, for example, firm competition for social rank, and consumer and employee activity that fulfills obligations or leads to group harmony, all derive from the maximization of appropriate, Japanesespecific, profit and utility functions that take the roles of rank, obligation, and loyalty into account. And, because those with more experience tend to be more productive, or because delayed compensation (i.e. underpaying the young and overpaying the old) to encourage loyalty balances out in the long run, compensating employees according to seniority is, in effect, the same as paying them the profitmaximizing, market value of their contributions to firm productivity. From such a vantage point, then, the presence of cultural variation does not mitigate the imperatives of self-interest and rationality postulates as such, but alters instead the means by which that self-interest is expressed, along with the specifics pertaining to

26 Translation by Morris-Suzuki [30, p. 186].

30 Western economics and the economy of Japan the conditions and characteristics of the rationality postulates.27 In this view, selfinterested behavior, and the need to employ rationality postulates to understand it, are universal across all cultures. But the fact that debates of the foregoing kind have occurred cannot be taken to controvert the principal argument of this essay. For, as has already been observed, it is not possible to develop tests to refute the universality of self-interest and rationality on empirical grounds. Acceptance or rejection of it as a basis for explanation can only be founded on the cultural norms underpinning that which is the object of explanation. And since those norms can vary widely from one society to the next, meaningful explanation requires that under certain conditions, in particular those that arise in Japan, self-interest and rationality be rejected. Moreover, even if an argument such as Koike’s were accepted, it would still be necessary to admit to the existence of a different resonance to, and a different meaning of, self-interested behavior in Japan from that which appears in the West. For self-interested behavior, although a driving force in Western culture, does not appear in Japanese culture in the same way. Indeed, the primary motivating forces in Japanese culture have little relation to it. Models of rational behavior must therefore have a different explanatory significance when they are applied to understand individual behavior in one economy as opposed to the other. It should be pointed out, however, that to say models of rational behavior applied to explain economic activity in the Japanese context have a different explanatory significance from those applied to explain Western economic behavior implies the claim, contained in the argument of this essay, that models of rational behavior, as a Western economist might understand them, are less relevant for analyzing the Japanese economy. And given the scarcity of resources available for scholarly research in economics, it is probably wise to use those resources in pursuing avenues along which the deepest and most significant meaning, as defined with respect to the culture of the society in question, is likely to be attainable. This, too, leads to the rejection of self-interest and rationality in explanations of Japanese economic behavior. Regardless, accepting the absence of self-interest as a motivating force behind individual behavior in the cultural environment of Japan, the dangers of employing Western thought forms that rely on self-interest to understand the Japanese economy are brought more sharply into focus. Apart from building explanations on foundations of limited cultural relevance, one would expect economists raised in a traditional Japanese environment to have, as a rule, a natural uneasiness (as expressed by Morishima and Murakami in the quotations cited above) with those constructs. For, in general, if explanation and understanding of observed

27 Reed [36, pp. 4, 25, 26, 47] has made this argument in a more general setting: Japanese behavior can be understood in terms of the same overall motivating forces that are used to explain Western behavior. But in reference to Japan, these motivating forces have to be applied to situations or environments with fundamentally divergent characteristics from those that arise in the West. And it is differences in situations that account for observed differences in behavior.

Western economics and the economy of Japan 31 behavior is the primary purpose of analysis, then when an investigator reared in a particular society undertakes an investigation of economic behavior in that society, there must be a confluence of the cultural background of the investigator with that implicit in the explanations of behavior in the models he constructs such that, given that culture, the investigator is comfortable with and believes in the efficacy of the models. Without this, the models will not seem reasonable to him, and he is unlikely to accept them. It is also clear that, to avoid irrelevance, an outsider has to respect the cultural environment of the economy he is studying. Thus, in these kinds of ways, the culture of society impinges on the economic behavior that takes place in it and, as a consequence, necessarily has to influence the understandings and explanations of that behavior constructed by the scholars who study it.

References 1. Aoki, M., Information, Incentives, and Bargaining in the Japanese Economy (Cambridge: Cambridge University Press, 1988). 2. Bellah, R.N., Tokugawa Religion (Glencoe: Free Press, 1957). 3. Benedict, R., The Chrysanthemum and the Sword (Boston: Houghton Mifflin, 1946). 4. Bentham, J., An Introduction to the Principles of Morals and Legislation (New York: Hafner, 1948). 5. Caldwell, B.J., Beyond Positivism: Economic Methodology in the Twentieth Century (London: Allen & Unwin, 1982). 6. Cournot, A., Researches into the Mathematical Principles of the Theory of Wealth, N.T. Bacon, trans. (New York: Kelley, 1960). 7. De Long, J.B., “Productivity Growth, Convergence, and Welfare: Comment,” American Economic Review 78 (1988), pp. 1138–1154. 8. Fallows, J., Looking at the Sun (New York: Pantheon, 1994). 9. Gerlach, M.L., Alliance Capitalism: The Social Organization of Japanese Business (Berkeley: University of California Press, 1992). 10. Greif, A., “Cultural Beliefs and the Organization of Society: a Historical and Theoretical Reflection on Collectivist and Individualist Societies,” Journal of Political Economy 102 (1994), pp. 912–950. 11. Gudeman, S., Economics as Culture (London: Routledge & Kegan Paul, 1986). 12. Harris, M., Cultural Materialism: The Struggle for a Science of Culture (New York: Vintage Books, 1979). 13. Horioka, C.Y., “Why is Japan’s Household Saving Rate So High? A Literature Survey,” Journal of the Japanese and International Economies 4 (1990), pp. 49–92. 14. Ito, T., The Japanese Economy (Cambridge: MIT Press, 1992). 15. Jevons, W.S., The Theory of Political Economy (New York: Kelley, 1965). 16. Johnson, C., MITI and the Japanese Miracle (Stanford: Stanford University Press, 1982). 17. ——, “History Revisited: Japanese-American Relations at the End of the Century,” Pacific Economic Relations in the 1990’s, R. Higgott, R. Leaver, and J. Ravenhill, eds. (Boulder: Lynne Riener, 1993), pp. 39–61. 18. ——, “The Foundations of Japan’s Wealth and Power and Why They Baffle the United States,” Japan: Who Governs? (New York: Norton, 1995), pp. 96–112.

32 Western economics and the economy of Japan 19. Kachelmeier, S.J. and M. Shehata, “Culture and Competition: a Laboratory Market Comparison between China and the West,” Journal of Economic Behavior and Organization 19 (1992), pp. 145–168. 20. Katzner, D.W., “Culture and the Explanation of Choice Behavior,”Theory and Decision 48 (2000), pp. 241–262. This volume: Essay 10. 21. Kinzley, W.D., Industrial Harmony in Modern Japan (London: Routledge, 1991). 22. Kirby, S., “The Reception of Western Economics in the Orient,” Quarterly Journal of Economics 66 (1952), pp. 409–417. 23. Koike, K., “Skill Formation Systems in the U.S. and Japan: a Comparative Study,” The Economic Analysis of the Japanese Firm, M. Aoki, ed. (Amsterdam: North-Holland, 1984), pp. 47–75. 24. Lau, L.J. and P.A. Yotopoulos, “Do Country Idiosyncrasies Matter in Estimating a Production Function for World Agriculture?” Journal of Economic Development 13 (1988), pp. 7–19. 25. Leff, N.H. and K. Sato, “Homogeneous Preferences and Heterogeneous Growth Performance: International Differences in Saving and Investment Behavior,” Kyklos 46 (1993), pp. 203–223. 26. Machlup, F., “The Problem of Verification in Economics,” Southern Economic Journal 22 (1955–56), pp. 1–21. 27. Marshall, A., Principles of Economics, 8th ed. (New York: Macmillan, 1948). 28. Morishima, M., The Economics of Industrial Society (Cambridge: Cambridge University Press, 1984). 29. ——, Why Has Japan‘Succeeded’? (Cambridge: Cambridge University Press, 1984). 30. Morris-Suzuki, T., A History of Japanese Economic Thought (London: Routledge, 1989). 31. Nakane, C., Japanese Society (Berkeley: University of California Press, 1970). 32. North, D.C., “Institutions and Economic Theory,” American Economist 36, no. 1 (Spring, 1992), pp. 3–6. 33. Polanyi, K. “Our Obsolete Market Mentality,” Primitive, Archaic, and Modern Economies, G. Dalton, ed. (Garden City: Doubleday, 1968), pp. 59–77. 34. Prestowitz, C.V., jr., Trading Places (New York: Basic Books, 1988). 35. Ramseyer, J.M. and F.M. Rosenbluth, The Politics of Oligarchy (Cambridge: Cambridge University Press, 1995). 36. Reed, S.R., Making Common Sense of Japan (Pittsburgh: University of Pittsburgh Press, 1993). 37. Sahlins, M., Stone Age Economics (Chicago: Aldine Atherton, 1972). 38. ——, Culture and Practical Reason (Chicago: University of Chicago Press, 1976). 39. Samuelsson, K., Religion and Economic Action (New York: Basic Books, 1961). 40. Sen, A.K., “Rational Fools: a Critique of the Behavioral Foundations of Economic Theory,” Philosophy and Public Affairs 6 (1976–77), pp. 317–334. 41. ——, “Internal Consistency of Choice,” Econometrica 61 (1993), pp. 495–521. 42. Smith, A., An Inquiry into the Nature and Causes of the Wealth of Nations (New York: Random House, 1937). 43. Stigler, G.J. and G.S. Becker, “De Gustibus Non Est Disputandum,”American Economic Review 67 (1977), pp. 76–90. 44. Temin, P., “Is It Kosher to Talk about Culture?” Journal of Economic History 57 (1997), pp. 267–287. 45. Weber, M., The Protestant Ethic and the Spirit of Capitalism, T. Parsons, trans. (New York: Scribner, 1958).

Reply to Thiruvadanthai 33 46. ——, The Sociology of Religion, E. Fischoff, trans. (Boston: Beacon Press, 1963). 47. van Wolferen, K., The Enigma of Japanese Power (London: Macmillan, 1989).

Reply to Thiruvadanthai1 In his comment on my essay, “Western Economics and the Economy of Japan” (Katzner [2]), Thiruvadanthai [4] raises three issues of considerable importance for economic analysis. The first concerns the significance of the relationship between the cultural background of the investigators and the culture in which the object of the investigators’ analysis is set. The second has to do with the meaning of the term “rationality” or “rational behavior” as it is used in economics. And the third, which relates, in a fundamental way, to the purpose of economic analysis itself, is about the meaning and role of explanation in economics. I disagree with the position taken by Thiruvadanthai, or that implicit in his discussion, on each of these issues and, in what follows, will try to set out the reasons why.

I In my essay, I argued that when analyzing economic behavior, it is necessary to both understand and account for the peculiarities of the culture in which that behavior takes place. As a corollary to that proposition, I also suggested that when an investigator is analyzing economic behavior in his own culture, he will not be comfortable with his own analysis unless that analysis is consistent with what he understands as the imperatives of his own cultural experience. Thiruvadanthai accepts the general proposition, but not the corollary. He seems to think that were the corollary to be true, then economists of different cultures would not be able to communicate meaningfully with each other, and that general principles underlying economic behavior that are independent of culture, even if sustainable, could not be discovered. I believe that both reasons for rejecting the corollary are invalid. At the outset, it should be mentioned that, in the field of anthropology, there are distinct approaches to analysis based on the view one has towards the significance of culture in understanding behavior (e.g. Gudeman [1, p. 28]). On the one hand the “universalists” claim that analysis transcends culture and that, in all instances, one model may be applied by everyone to understand the same kind of behavior everywhere. On the other hand, “localists” or “relativists” assert that individuals are capable of understanding only in terms of their own cultures and hence that, in the absence of far-reaching cultural similarities, a person of one culture cannot accurately and fully understand behavior that arises in another. I doubt whether

1 Journal of Post Keynesian Economics 22, no. 4 (Summer, 2000), pp. 653–658. © M.E. Sharpe, Inc. Reprinted with permission.

34 Reply to Thiruvadanthai Thiruvadanthai would accept either of these extremes in the context of economics. Nor would I. My own position is that in spite of significant cultural differences, enough interaction among human beings of dissimilar cultures can actually occur to permit an individual of one of the cultures to comprehend the other culture and the behavior that occurs in it. Now my statement above that “… when an investigator is analyzing economic behavior in his own culture, he will not be comfortable with his own analysis unless that analysis is consistent with what he understands as the imperatives of his own cultural experience …” says nothing about an investigator in one culture studying behavior that might be observed in another. But my essay does argue, and I reaffirm that argument here, that to properly understand the behavior of another culture, such an understanding has to be consistent with the imperatives of that culture that drive the behavior in question. Moreover, to the extent that an investigator does not meet the latter standard when studying economic behavior that takes place in the cultural environment he identifies as his own, he is likely to be uneasy with his analysis and look for ways to “improve” it. That is because his sense of the manner in which such things work in his own world is heavily dependent on his cultural background. Of course, it is always possible for an investigator in his professional activity to adopt the standards and perspectives of another culture, and that may provide him with some degree of comfort in analyzing his own economy from a vantage point different from that which he might otherwise take. But even so, for reasons already set out in “Western Economics and the Economy of Japan” and to be expanded on in Section III below, I maintain that it is absolutely essential when studying human behavior to capitulate to the most significant and relevant imperatives that propel behavior in the cultural environment in which that behavior occurs. If this implies that to understand the behavior of, say, a Japanese consumer it is necessary to drop the assumption of constrained utility maximization in favor of some other motive, then so be it. And in my essay, I showed why it is necessary to do precisely that. The fact that there are Japanese economists that continue to use constrained utility maximization to understand the consumption behavior of individual Japanese is irrelevant. Surely, the truth or falsity of an argument does not depend on who it is that supports or denies it. Turning to the second reason why Thiruvadanthai rejects my corollary, the general statement that culture matters in determining behavior certainly does not prevent the possibility that, to partially quote my earlier phrase, “general principles underlying economic behavior that are independent of culture” might turn out to exist. It might be that a particular principle is consistent with all known cultures and observed behavior in them. And I take such a principle to be discoverable because I hold, as indicated above, that cross-cultural communication and understanding are possible. Thiruvadanthai himself states that one such principle is that “… individual economic ‘behavior is the outcome of appropriate deliberation’” [4, p. 648]. The question is, how does he know this? Nowhere does he present an argument to show that such a principle is consistent with all cultures and all observed behaviors. It is entirely possible to believe that such a universal proposition is true. But then it is also entirely reasonable that someone else with the same

Reply to Thiruvadanthai 35 or different cultural background as Thiruvadanthai will disagree. In an indirect way, I will have more to say about Thiruvadanthai’s principle in the next two sections.

II In my reading of Thiruvadanthai’s comment, there seems to be considerable narrowness in the meaning he gives to the term “rationality.” Now the notion of rationality as determining economic behavior, existing as it does on the level of mental ideation, is not itself directly observable. It may or may not be present in the mind of an individual taking action and, even if there, it may or may not be the primary force that motivates his behavior. But in any case, it is most certainly a concept that has been created or invoked by economists for analytical purposes. In “Western Economics and the Economy of Japan” I defined rational behavior as behavior that is consistent with a prespecified set of axioms and that is designed to achieve an objective determined by self-interest (this volume, p. 20). However, on the basis of the first part of this definition alone, with or without a commitment to the objective-determining role of self-interest, it would be possible to specify alternative sets of axioms, call them A and B, each of which would generate different behaviors. I would call action consistent with either A or B rational. More precisely, behavior consistent with A would, to my mind, be rational with respect to axiom set A, while that consistent with B would be rational with respect to axiom set B. Clearly, because it omits the demand for a connection to self-interest, the second notion of rationality is more general than the first. After providing the first definition of rational behavior above, it was convenient for argument in the remainder of “Western Economics and the Economy of Japan” to compress the meaning of the concepts of self-interest and rationality so that they became identical to that of constrained maximization. But it is the second definition of rational behavior, the one that does not require a link to self-interest, on which I wish to focus here. Thiruvadanthai appears to use several different phrases to represent rationality in addition to the term “rationality” itself: “the outcome of appropriate deliberation” [4, p. 648] that I mentioned above, “procedural rationality” [4, p. 649], and “logical calculus” [4, p. 650]. All of these expressions are rather restrictive in that they are explicitly related by Thiruvadanthai in some way to constrained maximization. And the thrust of his argument would seem to imply that, in his view, the only way, and the proper way, of approaching rationality in economics is through such connections to constrained maximization. The second definition I gave above, however, is more general. According to it, a person could still be thought of as rational if his behavior emerged from constrained maximization. But behaviors motivated by such forces as custom, loyalty, obligation, or friendship that have no intrinsic relation to constrained maximization could also be considered rational with respect to an appropriate set of axioms even if those axioms were not connected to self-interest. (The fact that behaviors motivated in this way can always, but erroneously, be understood as the outcome of constrained

36 Reply to Thiruvadanthai maximization is discussed in the next section.) Thus, my second concept of rationality, being more cross-culturally relevant, is more widely applicable than Thiruvadanthai’s. Moreover, it is only in terms of broad notions such as these that there is any hope at all of discovering “general principles underlying economic behavior that are independent of culture.”

III Turn now to the issue Thiruvadanthai raises concerning the meaning and role of explanation in economics. At the end of his comment [4, p. 650] Thiruvadanthai claims that all deliberative economic behavior can be explained in terms of constrained maximization where, as he indicates earlier [4, pp. 648–649], cultural factors influence the objective function. Now it turns out that in another essay (Katzner [3]), I have actually given a formal proof of that proposition in the context of the theory of choice. This proof takes behavior that is based on, say, custom, and that violates the implications of constrained preference maximization, and adds realizations of the customs in question to, for example, the quantities of goods among which the individual chooses. Thus in selecting a basket of goods, the individual is also selecting a certain adherence to custom. By having the individual choosing baskets together with the adherences to custom intrinsic to those baskets, the seemingly non-preference-maximizing selections of goods are embedded in the outcomes of a constrained-preference-maximization problem. As in Thiruvadanthai’s claim, this also suggests, and I have no doubt that it can be formally proved in general, that all observed economic behavior everywhere can be viewed as the outcome of constrained maximization. But why, it should be asked, would anyone want to view all behavior everywhere in such a way? And it is this question that brings discussion to the meaning and role of explanation in economics. It is best to begin with some definitions. a phenomenon or event is said to have been explained when the place it occupies in the meaning of the universe of which it is a part has been exhibited or demonstrated. That is, to explain something is to show how that something fits in or contributes to an overall understanding of what is going on. In the context of economic behavior, explanation requires that reasons be given for that behavior that describe the forces that generate it, and that mesh with the general vision of the manner in which the society in which that behavior occurs operates. Explanations of behavior can be “right” or “wrong” in the following sense: an explanation is right or correct if it accurately reflects the forces that actually drove the individual to that action. It is wrong or erroneous if it does not. I take the primary purpose of economic analysis to be the correct explanation of economic events. Of course, prediction, or the forecasting of future events may be a by-product of correct explanation. But prediction could also emerge from erroneous explanation. For example, sun-spot activity might be a very good predictor of the movement of the Dow-Jones Industrial Average, but it is likely to be erroneous as an explanation of movements of that Average. Correlation does

Profit maximization and the Japanese firm 37 not imply causality. Prediction alone, that is, cannot rule out the possibility of spurious coincidence of events and predictors. Seen in this light, it is clear that to explain a person’s behavior (economic or otherwise) as the outcome of constrained maximization, when in actual fact that behavior was generated by some other force, is to provide erroneous explanation. My argument is that erroneous explanation is of little use to economists in understanding economic reality. For it can only lead us astray, causing us to think that we know something when, in fact, we do not.

References 1. Gudeman, S., Economics as Culture (London: Routledge & Kegan Paul, 1986). 2. Katzner, D.W., “Western Economics and the Economy of Japan,” Journal of Post Keynesian Economics 21 (1999), pp. 503–521. This volume: Essay 2. 3. ——, “Culture and the Explanation of Choice Behavior,” Theory and Decision 48 (2000), pp. 241–262. This volume: Essay 10. 4. Thiruvadanthai, S., “A Comment on Katzner’s ‘Western Economics and the Economy of Japan,’” Journal of Post Keynesian Economics 22 (2000), pp. 647–651.

Profit maximization and the Japanese firm: A reply to Coffey and Tomlinson1 In a recent series of essays published mostly in this Journal (Katzner [5], [6], [7], [8]), I have tried, among other things, to make the case on cultural grounds against employing the assumption of profit maximization to explain the behavior of that analytical entity referred to as the “Japanese firm.” The argument may be summarized as follows: The primary purpose of economic analysis is to provide correct explanation of observable economic phenomena, that is, in the present instance, to furnish accurate representations reflecting “true” reasons why firms behave in certain ways. All firm behavior is derived from human thought processes, including motivations, that rely heavily on the cultural backgrounds of the individuals making the decisions that generate that behavior. Moreover, significant cultural differences across some societies (or even some groups within a society) are present and manifest themselves in the dominance of different cultural traits within those societies (or groups) as the main forces propelling individuals’ actions. This, by and large, is true of the differences in Japanese and Western cultures, and that fact alone implies that the Japanese are generally motivated by different cultural imperatives than Westerners. When constructing correct explanations of firm behavior, then, it is necessary to abstract from the many features of reality and build a model of the firm that, among other things, 1 Journal of Post Keynesian Economics 26, no. 4 (Summer, 2004), pp. 745–750. © M.E. Sharpe, Inc. Reprinted with permission. The author acknowledges with appreciation and thanks the help of Charles Yuji Horioka and Douglas Vickers.

38 Profit maximization and the Japanese firm fully respects the dominant cultural traits of the society in question. In the US, for example, one of those traits is self-interest as it relates to the benefit of the agent, and the representation of that trait in models of the firm often appears in the form of profit maximization. But self-interest is not included among the dominant cultural characteristics in Japan and, as a consequence, its representation as profit maximization cannot appear in models purporting to provide correct explanation of Japanese firm behavior. To assert or assume that the Japanese firm maximizes profit when in fact, as a rule, it does not, is to engage in erroneous explanation that is irrelevant to the realities of Japanese economic activity. As was noted in Katzner [5], none of this is to say that the Japanese never engage in self-interested behavior. Certainly, in the face of bankruptcy, for example, it is reasonable to expect that firms will take action that enhances profits in order to survive. But acts undertaken in desperation or near desperation are not necessarily driven by the cultural standards that guide normal activity. And, in light of the cultural heritage of Japan as partly described above and as will be further detailed below, self-interested efforts of this or any other sort are clearly unsystematic in Japan, if not random. Furthermore, just because a firm takes action that moves it in the direction of increased profit does not mean that the motive behind that action is the maximization of profit. The main point here is that for a cultural trait to be dominant in a society and hence relevant in correct explanation, it has to have been institutionalized in some way so as to have become an integral part of the thought processes that produce the behavior of typical individuals. And, as previously indicated, that has not happened with respect to self-interest in Japan. The dominant cultural characteristics in present-day Japan along with the process of their institutionalization during the Tokugawa era (1600–1868) have been described by Bellah [2]. Those characteristics include service and loyalty to appropriate groups, the pursuit of harmony and compromise within them, and the honoring of the complex networks of obligations that have, partly in relation to the promotion of loyalty, harmony, and compromise, emerged. With respect to the management of the Japanese firm, these traits arise, in part, with respect to the seeking of the highest ranking possible for the firm (and hence its employees) in the Japanese social hierarchy. Thus, in probable violation of any assumption of profit maximization, competition among firms for social rank often determines which investments are undertaken (Nakane [9, p. 90]), labor is not necessarily hired up to the point at which the value of its marginal product equals its wage (Katzner [7]), and inefficient and unproductive employees are seldom fired (Nakane [9, p. 83]). Members of boards of directors, when such boards are present, are drawn largely from inside the firm or any Keiretsu group (or groups) with which it is affiliated and, as a result, the firm tends to be administered and run more for the benefit of its employees and Keiretsu group (or groups) than to provide maximum return for its owners (Abegglen and Stalk [1, pp. 185–186]). Within the Keiretsu structure itself, member firms are obligated to help each other even at the expense of profitability. Thus a firm will continue to buy a significant portion of the parts it needs from its Keiretsu-group suppliers even if it could purchase the same parts at a lower price elsewhere. Even firms unaffiliated with Keiretsu groups, out of

Profit maximization and the Japanese firm 39 loyalty to long-term suppliers, will often do the same, thus maintaining the longterm relations so highly valued in Japanese economic interactions. All of these elements considered, then, profit maximization does not seem to be a primary goal of the Japanese firm. Coffey and Tomlinson apparently do not accept this view. In their paper, “Globalization, Vertical Relations, and the J-Mode Firm” [3], they argue, in part, that the large Japanese companies have, in the last 50 years “… engaged in a series of cost-contingent subcontracting arrangements with low wage firms in the domestic economy, that have now given way to global production strategies that seek wage advantages abroad, unhindered by ‘commitments’ to small Keiretsu firms” [3, p. 135]. This outsourcing of production through “… overseas affiliates – at the expense of domestic sites – has served to both generate and then exacerbate a ‘hollowing out’ of Japanese industry” [3, p. 137]. And the latter, in turn, “… has contributed to lower economic growth and higher unemployment in Japan” [3, p. 137]. Up to this point, I have no quarrel with them. However, they then go on to infer from this that “… the cumulative evidence points toward behavior that improves the profit position of Japan’s corporate sector realized through conduct easily construed in ‘Western’ terms” [3, p. 140]. Combine this with their assertions that “… it is difficult not to support … the general applicability of a maximizing calculus when applied to the study of the behavior of the Japanese firm …” [3, p. 119] and “The … position taken in this paper is that … Japan’s current economic crisis is at least in part a consequence of the profit-seeking activities of its giant transnational producers …” [3, p. 142], and it is clear that Coffey and Tomlinson are, at least implicitly, attempting to cast doubt on my claim, set out in summary form above, that the behavior of Japanese firms is best understood in terms other than profit maximization. It seems to me that the (implicit?) Coffey–Tomlinson assertion that the evidence “points toward” profit maximization on the part of Japan’s large firms is a non sequitur and that the implication casting doubt on my claim is unwarranted for the reasons given below. First, profit is a function of many variables, not only that portion of labor that is outsourced to foreign economies in order to economize on wages. To say that a firm has moved production abroad for that purpose means only that, with respect to this one particular portion of its labor input, the firm is lowering its labor cost per unit of that labor. It carries no implication that the firm is, in general, locating in that country or region where labor costs are lowest, or that the firm is employing its optimum input mix by minimizing its overall production cost for each level of output. Nor does it mean that the firm is necessarily hiring the profit-maximizing quantity of labor abroad, or that the firm is necessarily employing the profitmaximizing quantities of domestic labor (for, say, administrative operations at home) and other inputs. Profit maximization also depends on the manner in which the firm finances its production and the expansion of its productive capacity – in particular, on the mix of equity versus debt capital it chooses. Therefore, to say only that a firm has moved its production abroad to take advantage of lower labor costs, does not say very much about whether the firm is actually maximizing profit or not.

40 Profit maximization and the Japanese firm In light of the preceding discussion, it is clear that there are grounds other than profit maximization that might lead the Japanese firm to move production to countries outside of Japan. Indeed the expansion of productive operations abroad may be propelled by the same non-profit-maximizing, culturally generated forces that govern the firm’s domestic activity. Thus, for example, the firm may feel that it needs to outsource some of its production simply to maintain its social status by doing what its competitors are doing to enhance the growth of their firms and create “presences” in foreign countries. The second reason why the Coffey–Tomlinson argument does not follow is, as partly suggested above, that even if, in the unlikely event it could be empirically demonstrated that a particular collection of activities undertaken by a firm achieves greater profit than any other possible collection of activities in which that firm may engage, the choice of that profit-maximizing collection by itself still cannot imply a motive of profit maximization. Nor can it imply that the firm’s actions in the future will continue to be (fortuitously?) profit maximizing. That is, the firm may arrive at a profit-maximizing position adventitiously in the process of consciously setting out to satisfy other objectives as described earlier. Of course, in the absence of any reason why the firm might not be maximizing profit and without actually asking the firm’s decision-makers what is motivating their decisions, one might hazard a guess that the firm is indeed driven by profit maximization. But in the case of Japan there is good cause, as indicated with respect to Japanese culture previously, to conclude that the main propelling force behind the behavior of Japanese firms is not profit maximization even though the outcome of a particular firm’s behavior might in fact result in the maximization of its profit. To assert profit maximization under these circumstances is to leap to a largely unjustified conclusion and to court the very real possibility of erroneous explanation. Third, the Coffey–Tomlinson argument centers primarily on the Japanese automobile industry with some discussion extending to large companies and Keiretsu groupings. But this focus covers only a small portion of the actual Japanese economy: the automobile industry generates less than 2.5% of Japanese GDP,2 while large firms (300 or more employees) in the manufacturing sector in general contribute about 9.4%.3 In addition, the number of large firms accounts for between only 0.1 and 0.2% of all private establishments.4 The number of firms of any size linked in Keiretsu groupings has been estimated at 22,000,5 whereas the

2 2003 Annual Report on National Accounts (Japanese Ministry of Finance Printing Office), pp. 212–213. 3 This follows since the value added by large firms in the manufacturing sector for the year 2000 was 43.3% of Japanese GDP (2000 Census of Manufactures, Japanese Ministry of Economy, Trade and Industry), whereas the value added by the entire manufacturing sector as a whole in 2000 amounted to 21.8% (2003 Annual Report on National Accounts, Japanese Ministry of Finance Printing Office, p. 209). 4 1996 Establishment and Enterprise Census (Japanese Ministry of Public Management, Home Affairs, Posts, and Telecommunications). 5 Ito [4, p. 206].

Profit maximization and the Japanese firm 41 total number of firms in the manufacturing sector alone has been counted at about 341,000.6 And the latter figure, of course, leaves out large chunks of the Japanese economy, including the approximately 1.4 million small retail establishments that dot the Japanese economic landscape.7 My assertion about the absence of profit maximization is a general statement that is intended as a guide to correct analytical explanation of the Japanese firm as an abstract entity. Such a statement cannot be refuted or declared irrelevant by referring only to one very small portion of the Japanese economy. Finally, it should be noted that, as also recounted in another essay (Katzner [7] – the numbers there are drawn from other sources), there is some suggestive, though by no means conclusive empirical evidence that many Japanese firms do not make profit-maximizing decisions. The absence of profit maximization is alluded to by taking the US as the basis for comparison of the Japanese economy with a competitive economy in which firms maximize profits, and by observing that, in terms of adjusted accounting earnings, between 1967 and 1988 the rate of return on capital in Japan was considerably less than that in the US. In addition, comparative productivity data again using the US as the competitive, profit-maximizing base, further suggests a lack of profit maximization in Japan over roughly the same period by hinting at the possibility that many firms overhire labor, that is, they hire a quantity of labor that is greater than that at which the value of labor’s marginal product equals its wage. Regardless, the general conclusion remains that an assumption of profit maximization under cultural conditions, like those in Japan, where that assumption is not warranted is likely to be both misleading and preclusive of correct explanation.

References 1. Abegglen, J.C. and G. Stalk, jr., Kaisha, The Japanese Corporation (Tokyo: Tuttle, 1985). 2. Bellah, R.N., Tokugawa Religion (Glencoe: Free Press, 1957). 3. Coffey, D. and P.R. Tomlinson, “Globalization, Vertical Relations, and the J-Mode Firm,” Journal of Post Keynesian Economics 26 (2003), pp. 117–144. 4. Ito, T., The Japanese Economy (Cambridge: MIT Press, 1992). 5. Katzner, D.W., “Western Economics and the Economy of Japan,” Journal of Post Keynesian Economics 21 (1999), pp. 503–521. This volume: Essay 2. 6. ——, “Reply to Thiruvadanthai,” Journal of Post Keynesian Economics 22 (2000), pp. 653–658. This volume: Essay 2. 7. ——, “Explaining the Japanese Economic Miracle,” Japan and the World Economy 13 (2001), pp. 303–319. This volume: Essay 6. 8. ——, “‘What Are the Questions?’” Journal of Post Keynesian Economics 25 (2002), pp. 51–68. This volume: Essay 3. 9. Nakane, C., Japanese Society (Berkeley: University of California Press, 1970).

6 2000 Census of Manufactures (Japanese Ministry of Economy, Trade, and Industry). 7 1999 Census of Commerce (Japanese Ministry of Economy, Trade, and Industry).

3

“What are the questions?”1

In a 1977 article in the Journal of Economic Literature, Joan Robinson [17] asked the same question. She suggested in that paper that the questions economists ask, as well as the answers they provide, are determined more by political outlook and ideology than anything else. “In economics,” she wrote [17, p. 1318], “arguments are largely devoted, as in theology, to supporting doctrines rather than testing hypotheses.” She then went on to criticize the questions posed and addressed within the mainstream conventional wisdom of the economics establishment of the time. One might contend that Robinson’s general criticism still applies with all of its force today. But political outlook and ideology emerge, in part, from elements of human existence that have themselves direct influence on the questions economists ask without passing through the filters of political outlook and ideology. One such element, and that which is the main focus of attention here, is culture. Indeed, like Robinson’s, the present essay may also be viewed as a criticism, some quarter of a century later, of the mainstream conventional wisdom of the economics establishment, but now for its apparent reluctance to consider seriously the impact of culture on economic behavior, rather than for its political and ideological narrowness. Of course, it is always possible to claim that economists’ standard analyses of consumer and firm behavior already account for culture in that cultural variation, when present, is reflected in, say, preference differentiation (e.g. Schultz [20, pp. 20–21]), or an absence of sufficient information to behave rationally (e.g. Stiglitz [23, p. 23]). But this merely sweeps the “dirt” of cultural effects under the proverbial rug. For, as will be argued below, to take cultural variation seriously is to recognize that the very structure of models and explanations may have to be modified across differing cultural environments. In particular, when,

1 This is a revised and considerably expanded version of the paper with the same title published in the Journal of Post Keynesian Economics 25, no. 1 (Fall, 2002), pp. 51–68. © M.E. shapre, Inc. Reprinted with permission. Thanks are due to Samuel Bowles, John W. Cole, Paul Davidson, Charles Yuji Horioka, and Douglas Vickers for considerable help with this essay.

“What are the questions?” 43 for cultural reasons, the motivations behind agent behavior vary, meaningful and correct explanation of that behavior must, of necessity, be based on different decision criteria. Retaining the same criteria and condensing cultural variation to alternative manifestations of preferences or informational availability marginalize and avoid the driving force of the behavior, and can fool the economist into thinking that he has explained something when actually he has not. There are, not surprisingly, exceptions to the avoidance of cultural issues as they relate to economic analysis. Greif [8], Hampden-Turner and Trompenaars [11], Henrich et al. [12], Katzner [13], North [14], Roth et al. [19], and Sen [21] have all suggested that culture matters in determining, and that it needs to be accounted for in explaining economic behavior. But by and large the mainstream view is that one collection of models, namely, those developed to understand Western economic activity, satisfactorily explain all economic behavior in all cultural environments everywhere. Speaking for many, Becker [1, p. 14] not only put it succinctly, but extended the idea well beyond the confines of mere economic action when he wrote that “… all human behavior can be viewed as involving participants who maximize their utility from a stable set of preferences ….” The following paragraphs put forth some of the reasons why this approach is inadequate and inappropriate when attempting to understand or explain large categories of economic phenomena. The essay begins with a description of the manner in which culture influences the thoughts that result from mental processes and shows how those thoughts, were they to derive from different cultural environments, could lead to different economic behaviors. It then illustrates how American cultural characteristics implicitly appear in models of American economic behavior and suggests as an example that, because the cultural foundation is not the same, similar models are irrelevant to the understanding of, and hence should not be applied to explain Japanese economic behavior. Finally it demonstrates how, in light of the role of culture and cultural change in molding human thoughts, people of different cultures can arrive at different questions and come up with different answers in relation to economic matters. It should be emphasized that the argument of the present essay is concerned only with explanation, not with prediction, in economic analysis. To explain is to clarify, to render intelligible, and to provide reason for. Prediction, or the forecasting of future events, may be a by-product of explanation, but it can never be used as a substitute for explanation. This is because prediction by itself cannot rule out the possibility of a spurious coincidence of events and predictors.

I Culture has to do with belief structures, ethical and behavioral norms, interests, the development of knowledge, the extent of intellectualization, and so on. Its determinative components include significant parts of both the current and historical experience of the society whose culture is under consideration. Thus, for example, past and present immigration has often introduced, and continues to introduce, new elements into the cultural complex, brought in by people arriving

44 “What are the questions?” from foreign lands. Prior and present social, political, and economic institutions have frequently contributed, and continue to contribute, to cultural modification. And previous and current technological change and internal population movements have also influenced, and continue to influence, cultural development. Culture, of course, interacts with the mind, not with the brain. On the one hand, the latter is a physical mechanism within which thoughts and other activities, both physical and mental, take place. The mind, on the other hand, only exists in relation to the brain. It is that component of personhood that has the capacity to think and to recognize and respond to feelings. Subsequent argument suggests the ways in which culture influences the activity of the mind. But the mind also contributes to both the maintenance and the development of culture. Culture is maintained through its transmission from, for example, parent to child and group to individual. Both of these kinds of transmissions involve mental processes. Cultural development or change can occur in at least two respects. First, intellectual history and the memory of inherited social and political institutions (e.g. a democratic system), both, in part, products of the mind, can generate behavior that leads to cultural modification. Second, the same is true of the emergence of new ways of looking at things and new ways of perceiving and ordering relations between various aspects of reality. The maintenance and development of culture have been considered more fully by Boyd and Richerson [5] and Cavalli-Sforza and Cavalli-Sforza [6, pp. 207–215]. In addition, culture is generally not homogeneous across society. Different groups or regions often have their own “subcultures” (which may not even be homogeneous within themselves), and the contrasts and contradictions between such subcultures can be quite sharp and extensive. But within any culture or subculture, it may be said that the manner in which the meaning and determinants of that culture or subculture are understood gives rise to a method for, or a guide to perceiving reality and, in addition, furnishes standards for interpretation, judgment, and action. There are, however, a number of additional characteristics of culture that are worthy of special consideration. Two of these discussed by Geertz [7, pp. 44–46, 75–76, 214–218] are of particular interest here. 2 First, in the process of individuals’ minds interacting both with and within a cultural complex, certain symbols like words or national flags, which are descriptive in a summary sense (i.e. not exhaustive descriptions but, rather, short-hand expressions) of various aspects of that complex, emerge as prominent. These symbols are public in that they lie outside of the individual. Culture patterns, which are collections of elements of the cultural complex and are not, as previously indicated with respect to culture in general, necessarily homogeneous across society, likewise project their own organizations and aggregations of possibly varying symbols. Of course, culture patterns, even within the same society, may conflict. And the cultural complex itself, along with any of its subcultures, may be encapsulated, again in summary

2 Subsequent discussion presents a slightly different view than that projected by Geertz.

“What are the questions?” 45 form, as the accumulated and appropriate totality of these nonhomogeneous and possibly conflicting culture patterns. It is with respect to the symbols in terms of which they summarize and describe, that culture patterns give meaning to social and psychological reality, both by furnishing a basis for models of that reality or the means to understand it, as well as by providing a foundation for models that deal with that reality or prescriptions for behaviors in it. Clearly, then, culture is fundamental to human understanding, explanation, and action. More will be said of the roles of symbols and models below. The second characteristic of culture relevant for present discussion is that, in addition to the above, culture may also be regarded as a control mechanism for governing human behavior. That mechanism clearly relates to such things as the ethical mores and canons of acceptable behavior mentioned earlier. Furthermore, it is possible to view individuals as highly dependent on this sort of control. Unlike lower animals, human beings have a much greater capacity and freedom to generate behavioral responses to the various stimuli they encounter. Lower animals react more on the basis of programmed sequences of learned, instinctive actions. For example, in the face of fear they might threaten first and then flee. But an individual, in a similar circumstance, has the flexibility to choose one or more options from among many possible acts, and each option can be exercised in varying degrees. He might, for example, alternate between blustering and placating, with the force and energy behind each dependent on the previous response received and the subsequent reaction that is anticipated or hoped for. Thus the actual sequence of action in answer to a given stimulus may exhibit a wide range of complexity and subtlety. Moreover, the nature and sequencing of such acts across the situations in which they are used are governed by culture rather than simply by preset instinctive behavioral sequences. Without the direction provided by culture patterns, then, human experience would have a much lower degree of form and substance and, as a result, human behavior would be in danger of becoming chaotic and pointless. According to Geertz [7, pp. 46–52], these characteristics of culture, along with cultural complexes in general, unfolded in relation to the physical development of humanity over the ages. In particular, Geertz claims that, subject to a qualification to be mentioned shortly, as he evolved from prehistoric time to the present day, man’s physical and cultural maturation occurred more or less simultaneously and each influenced the advancement of the other. Along with his learning to organize hunting and gathering, to create and use tools, and to live in families, man’s reliance on organized collections of symbols necessarily grew. The latter growth, which meant greater dependence on culture patterns, forced alteration in the criteria for survival, and this, in turn, led to the emergence of new practices and physical traits. As the process continued, his physical and cultural structures became increasingly complex. More had to be learned at each stage; and the sorts of things that had to be learned became more complicated. What makes man so different from other animals is not only his remarkable ability to learn, but also the quantity and complexity of the things he has to learn in order to function at all in the community in which he lives. Thus culture may be seen as providing

46 “What are the questions?” the link between a person’s innate capacities and his actual behavior, between what he is capable of becoming and what he actually becomes. Regardless, at each stage of the development process just described, the required learning could have taken place only within a framework based upon the organized symbols of a culture pattern. And because man may be thought to have evolved with no organized collection of symbols genetically implanted in his brain, the mind, in that case, is completely dependent on importing them for its operation. Those imports, or cultural resources, once brought in, become vitally involved in his mental activity. It follows that culture is necessarily an intrinsic component of human thought. While such an evolutionary-historical scenario as Geertz has proposed in the manner of the foregoing does not necessarily provide a definitive explanation of man’s cultural status, it is nevertheless true that, in the act of thinking, the mind makes use of and manipulates the symbols associated with the culture pattern or patterns with respect to which that thinking occurs. Therefore, it may be concluded that, although human thought is private due to the sanctity of personhood, it is still influenced by, and necessarily takes into account, the public symbols on which it depends. Now symbols, recall, are recognizable formulations, such as words, pictures, and sounds, of notions abstracted from experience.3 They are embodiments of beliefs, attitudes, ideas, etc. Symbols are separate from what they represent and impose meaning on it. They are presented, as he passes through life, to the individual by the community in which he lives. They may modify during his lifetime and, in whatever form they assume, they are left with the community after his death. He uses these symbols to orient himself in the community and to understand, through the construction of models, his surroundings and experiences. The manner in which those latter understandings come about in relation to models is considered momentarily. Symbols symbolize; that is, they function to mediate meanings, by matching the states and processes of symbolic models against the states and processes of reality. They are outside sources of information with respect to which human life can be patterned and which provide the mechanisms for the perception, interpretation, judgment, and action mentioned earlier. Thus, as previously suggested, culture patterns, which project organized combinations of symbols, are templates for rendering social and psychological phenomena systematic and coherent, and they permit man to make sense of that which he experiences. Culture patterns are greatly needed too, because, as has been previously indicated with respect to the underlying symbols, such guides to behavior are not instinctive, and they are not determined solely by the genetically established circuitry of the brain. Consider, further, the activity of the mind in relation to culture. Three observations appear to be relevant. First, the mind contains, in the sense that it is conscious of, “dispositions” that manifest themselves in certain actions and certain things (Geertz [7, p. 58]). These dispositions include such elements as skills,

3 Subsequent discussion is derived from Geertz [7, pp. 44–46, 208, 214–218].

“What are the questions?” 47 tendencies, habits, and capacities, and are culturally influenced in their reliance on the symbols of the relevant culture patterns. (Two further examples of such dispositions are provided subsequently.) Second, the very nature of personhood is such that human decision and action, which originate in the mind as a result of the directed reasoning to be discussed momentarily, are determined by both physical capacities of the brain and psychic inputs. That, of course, is a necessary and universal characteristic. It follows from previous argument, however, that the nature and range of decision-making activity, along with the results it achieves, are dependent upon, and in varying degrees determined by, the cultural complex within which it occurs. Generally speaking, thoughts vary widely from culture to culture, with different perceptions, interpretations, meanings, and significances. Hence, the behavior derived from those thoughts is likely to vary also. Third, as a particularization of the previous claim that the physical and cultural development of man occurred simultaneously and interactively throughout his evolution, Geertz [7, pp. 67–69] argues that the same assertion can be made concerning cultural development, the development of the mind, and the development of the brain. However, he continues, and this is the qualification cited earlier, since the emergence of modern man at the end of the age of glaciation, the physical evolution of his brain has slowed while his cultural and mental development seems to have increased, possibly at an increasing rate. Clearly, human thought has both intellective, in the sense of directed reasoning, and affective or emotional aspects (see Geertz [7, pp. 78–81]). The purpose of the former is to solve problems, or to transform something that is obscure and opaque into clear and coherent form. That transformation is accomplished by constructing a model with the symbols at hand. The aim of the latter is to sort out and clarify feelings. This rests on symbols and models too. To illustrate, for a hand raised in a certain way to provoke fear (as, for example, in a prelude to a physical assault) requires a model, with its symbolic foundation, indicating what that raised hand is likely to do and the consequences that might ensue. A hand raised by a student participating in a discussion in a classroom, however, does not instill fear because the symbolic interpretations of both the manner in which the hand is raised and the context in which it is raised are different. But, in any event, it is clear that both intellective and affective tasks can be accomplished only by using the cultural resources imported into the mind from the culture patterns of the community that surrounds the individual. Motivations like greed and frugality are additional examples of dispositions of which the mind is conscious. They are persisting tendencies, or chronic inclinations to perform certain acts and experience certain feelings in certain situations. Motivations emerge from the integration of directed reasoning and emotion, and impart direction and orientations that point to particular ends. It follows that, to a significant extent, motivations depend on culture due to the previously described cultural grounding with respect to symbolic interpretation of the directed reasoning and emotion out of which they are constituted. Indeed, culture patterns are capable of arousing in individuals powerful, pervasive, and enduring motivations.

48 “What are the questions?” Evidently, then, human beings search, cope, adjust, understand, and explain through the construction of models. And the construction of models rests, in turn, on the culture patterns or organized collections of symbols imported from the community in which the individuals live. This is the point of departure for Gudeman [9, pp. 37–43] and Gudeman and Rivera [10, pp. 14, 15] in their investigation of the role of models in human thinking and the interaction of those models with culture. The term “model,” as used by them, is meant to refer to a culturally constituted relationship between the thing that is being modeled and the schema that makes up the model. Thus a model provides a means for interpreting, understanding, explaining, knowing, or even doing. (For example, a plan for going to a restaurant and eating a meal may be viewed as a model.) It is only a partial construction in that it selects just those pieces from the totality of experience or potentially possible experience on which attention is to focus. Models are often constructed in relation to fundamental analogies. These analogies are so basic to the thought forms of a person that they define reality for him. They constitute the dominant vocabulary with respect to which experience is understood and expressed. Pepper [15, pp. 91, 319] describes the process of building a model from a fundamental analogy in the following way: a person seeking to understand a particular phenomenon searches for a clue to its comprehension. Out of the mass of raw, pre-analytical cognitive material of which he is conscious, and which has been obtained from his past experience and knowledge (and expressed in terms of the symbols of appropriate culture patterns), he focuses on some part that catches his eye, and tries to understand the phenomenon in question in terms of it. The part on which he focuses is what he recognizes as his fundamental analogy in relation to the phenomenon. Next he sorts out the structural characteristics of the fundamental analogy, and thereby determines the categories that constitute the basic concepts of his explanation of the phenomenon. He then attempts to interpret the facts of the phenomenon in terms of the basic concepts. While engaged in these latter efforts, he may find he has to qualify or otherwise adjust the categories initially derived from his fundamental analogy. Thus the categories employed may modify as his thinking develops. At any rate, the end result is the model he uses to understand the original phenomenon. The process of model building may also begin with a search to find previously developed models of past experience (which may have been derived from a still earlier model or from the fundamental analogy) that might, at least in part, be applicable to the thing currently being modeled. Thus past wars or past economic depressions may have considerable impact on the interpretation of, respectively, a current war or a current depression. Models may be arranged according to the extent of what Gudeman [9, p. 41] refers to as ontological commitment to them. At one extreme is the “as if ” formulation in which a model is regarded as a heuristic device or a convenient fiction. In this case, however, the model, although usable for predictive purposes, cannot necessarily provide any explanatory information about what is actually happening in the real world. The “as being” formulation at the other extreme presents the model as an accurate representation of what exists in reality. In either

“What are the questions?” 49 case, the individuals using models usually require their justification or validation, in some form, on the basis of personal experience. In this regard, it should be noted that an “as if ” model can closely conform to the observations of personal experience without providing the real explanation, or even an approximation of the real explanation, of how those observations actually came about. Models are built, as has been described, on the basis of past experience and knowledge. They are often shared by members of the same community and serve as the foundation for communication among them. Indeed, they are frequently the product of discussions and arguments (about human interactions and practices) among communities of individuals fashioned over long periods of time. Models also vary in generality and their development is an ongoing process. They are remade over and over again through repeated use and in response to cultural change. Moreover, models relating to human and group behavior are social models in that they are both derived from, and applied to, social activity. Thus the use of a Western model, say, to explain the behavior of an individual of a non-Western culture runs the risk of denying the capacity of that individual to understand his own environment and behavior in his own way, unless it is regarded as a total “as if ” fiction irrelevant to what is actually going on. That is because the Western activity from which it is derived does not necessarily belong to the same category of activity as that to which it is applied. The economic behavior of individuals consists of conscious action taken by them that is intended to have specific economic consequences. That action, like all action, relies, as has been seen, on the construction of appropriate models in the minds of the individuals under consideration.4 Moreover, since it is entirely conscious and undertaken with intent, that action also, of necessity, springs from their motivations. And because both models and motivations are, as previously indicated, culturally dependent in their reliance on culturally generated symbols, so, too, is economic behavior. The previous argument that economic behavior is generated in part by models and motivations that are culturally dependent is not meant to belittle the role that an individual’s situation can play in determining such behavior. As has been suggested by Reed [16] and Ross and Nisbett [18], individuals with similar models who are motivated in the same way can still behave differently because they are behaving in different environments. But the focus of attention below is on motivational dissimilarities that arise from differing cultural backgrounds. In the present context, then, individuals in comparable situations will behave differently if their cultural backgrounds have instilled in them different motivating forces.

4 It should be noted that the word “model” as used here (and earlier in this section) refers to a structure of possible explanatory relations between real world entities and the historical emergence of events. The descriptions of what it is that constitute such relationships are in no sense necessarily quantifiable in the manner in which, in accordance with well-established conventions, economists have tended to imagine.

50 “What are the questions?”

II In plying their trade, economists often engage in two broad categories of analytical endeavor: explanation and prediction. It is frequently suggested that the economic models which are used only for predictive purposes need have no relation to what is actually happening in the real world. The only concern, in that case, is the accuracy of the forecast that emerges, and “as if ” models are sufficient for that purpose. But explanation is another matter. To explain an action means, as indicated above, to clarify it, to render it intelligible, and to provide relevant and accurate reason for its occurrence. Explanation, that is, requires an “as being” model because it attempts to assert what it is in reality that actually produced the action or the outcomes associated with it. Moreover, models built to explain may always be used as a basis for prediction; but for many economists it is not necessary that models developed only for prediction have explanatory power. Regardless, since culture is an important determinant of actual economic behavior in the manners suggested above, it is clear that economists’ models purporting to explain that behavior need to take culture and the ways in which it influences behavior into account. The means by which such an end is to be accomplished can be expected to depend on a variety of perspectives and preferences among scholars. The objective of subsequent discussion is not to offer detailed suggestions on that plane, but rather to investigate what is seen as a highly important preliminary question; namely, the reasons why new levels and styles of thinking among economists are necessary in these respects, and the reasons why, as a result, commitment to established thought forms can be potentially and disastrously misleading in the formulation of economic explanation and policy. In that sense, what follows is a prolegomenon to economic model-building and to the understanding of conventional procedures. To illustrate how culture is reflected in economists’ explanations of events, attention will focus initially on the traditional modeling of American economic behavior constructed by American economists. Consider motivations first. In attempting to explain the actions of individuals, American economists frequently invoke what are referred to as traditional “postulates of rationality.” Thus, for example, American consumers, in arriving at their demand behavior, are usually assumed to choose baskets of commodities that maximize their utility functions subject to their budget constraints. And American firms are taken to hire inputs and produce outputs so as to maximize either short-run profits or long-run capital value, or both. The cultural basis for these postulates relates, in part, to an amalgam of the American values of the sanctity of the individual, self-reliance, and success.5 In American society, it is generally the case that the persons who are the most successful are also the ones who are the most looked up to and receive the most respect. Of course, one of the more common ways to achieve what might be considered evidence of success is by amassing wealth. But in attempting to accumulate wealth, the individual is typically thought of as completely alone.

5 For a discussion of these values and their place in American culture, see Bellah et al. [3, Chs. 1–3].

“What are the questions?” 51 The only one he can truly count on is himself. All decisions are his to make. He must collect the information he needs, marshal the requisite resources, and take all of the responsibility. Thus, in the quest for wealth, the individual of necessity has to behave in his own self-interest with respect to the purchases and investments he makes and the endowments he sells. Clearly, then, the pursuit of self-interest, which is, by itself, another feature of the American cultural scene, is intimately intertwined with the American cultural values of the sanctity of the individual, self-reliance, and success.6 And it is self-interest that often generates, when translated into American economists’ models of consumers and firms, the traditional postulates of rationality.7 Cultural characteristics are mirrored in other ways in models built by American economists to explain American economic behavior. A frequently employed method of measuring economic success in America is in terms of the material objects with which one is able to surround himself. As a rule, the bigger the house, the more opulent the furniture, the more expensive the car, the greater the evidence of success. This suggests that the larger an individual’s income, the more he will consume. Such a relationship between income and consumption appears in American economists’ models in at least two contexts. First, in the traditional theory of consumer demand, a good having the property that an increase in the consumer’s income results in, other things being equal, a larger quantity being demanded by him, is referred to as, using a value-laden term, “superior.” Most goods are thought to be superior. Goods for which increases in income lead to declines in quantities demanded are pejoratively labeled “inferior.” The latter goods are considered to be less typical and more pathological. Moreover, it is a conclusion of the theory that raising the consumer’s income (again, other things held equal) necessarily results in increased demand for at least one good. A second and perhaps more significant example arises with the macroeconomic consumption function in which aggregate consumption is taken to be an increasing function of aggregate disposable income. This is an explicit statement of the conclusion derived from the cultural considerations set out above. Moreover, if the level of aggregate taxation were initially taken to be fixed as aggregate income, and hence disposable income, varies, and if that level of taxation were then “permanently” reduced, consumption would rise at all values of aggregate income. The latter, too, is suggested by the same cultural considerations, indicated above, that lie behind the consumption function. A third aspect of the American consumption function that could be attributed to cultural effects is the form the short-run function takes in presenting a declining average propensity to consume

6 The process by which the pursuit of wealth and self-interest was institutionalized in Western European culture during the seventeenth century has been described by Weber [24]. Of course, the motivation to pursue wealth and self-interest came to American shores with the arrival of various European immigrants. Weber’s argument has been summarized by Katzner [13] which appears in this volume as Essay 2. 7 The beginnings of this translation dates at least to Smith [22, p. 14]. The fact that behaviors generated by non-self-interested motivations, such as altruism, could be considered rational is ignored.

52 “What are the questions?” as disposable income rises. Since a falling average propensity to consume implies a rising average propensity to save, and since proportionately greater savings now generally means still greater wealth in the future provided income levels are maintained, individuals, like Americans, focused on the pursuit of wealth would have a natural tendency towards a lower average propensity to consume with higher disposable income. These properties of the consumption function clearly have implications of considerable importance for models in which the consumption function appears and for policy conclusions drawn from them. Thus, as an illustration with respect to what will be called the (short-run) Keynesian-cross American model below, if consumption were an increasing, linear function of disposable income with marginal propensity to consume between 0 and 1, if aggregate income were partitioned into consumption, autonomous investment, and government expenditures only, and if the level of government expenditure were also assumed fixed independently of aggregate income (as may also be assumed of the levels of aggregate taxation and autonomous investment), then decreasing aggregate taxation would, by enlarging consumption at all levels of aggregate income, increase the equilibrium level of aggregate income. That is, reducing aggregate taxation becomes, ceteris paribus, one means of attempting to stimulate the economy and moderate the danger of apparent tendencies to recession or depression. Since, as has been suggested, models that economists use to explain economic behavior have to account for and mirror the cultural environment in which that behavior takes place, it is evident that economic behavior arising in a culture in which the characteristics of American society described above are absent may require different models to explain it. Elsewhere [13] it has been shown that, at the level of the individual, variation in parametric values of behavioral functions or the introduction into the relation itself of a parametric variable identifying culture is insufficient to properly account for cultural differences. The argument may be summarized as follows: all models constructed to explain behavioral relations employ assumptions pertaining to the forces that drive or generate the behavior described in the relations. Often those assumptions are, in the case of the consumer, (constrained) utility maximization, and, in the case of the firm, profit maximization (subject to production possibility constraints). These assumptions, along with others (such as those that characterize the properties of utility and production functions), imply that the behavioral functions so derived necessarily have their own particular properties. It is the latter properties that, in principle, can be used as the basis for empirically testing the viability of the assumptions. But in practice, it is never possible to refute such assumptions because any disconfirming test can always be attributed to changes that occurred during the time between observations (such as modifications in preferences, technology, and expectations) and that are beyond the control of the economist. Therefore the viability of these assumptions can only be based on the judgment of the investigator that his suppositions accurately reflect what is actually going on. And since cultural patterns, as has been seen, largely determine the motivations of individuals, the only way to accurately

“What are the questions?” 53 reflect what is actually going on is if the assumptions relating to the driving force of behavior truly represent the most important imperatives of the cultural environment in which the behavior in question takes place. That, in turn, means that where maximization, say, is a culturally inappropriate assumption in relation to consumers and firms, it has to be discarded in favor of some other behavioral propellent, and hence that parametric variation in behavioral functions, even if a separate parameter indicating culture is included, is likely to be insufficient to properly record the impact of cultural differences that render maximization reasonable in one case but not in another. At the macro level, where the degree of abstraction is considerably greater, parallel arguments apply. When the analytical image of the consumption side of the economy is based on the so-called “representative consumer,” and/or the production side emerges from the “representative firm,” the above argument may be used in much the same form – although the meaning of the phrase “changing preferences” as a rationalization of the presence of disconfirming observations is not as clear. (Exactly whose preferences are changing, and in what way?) When macro relations, such as, for example, the consumption function, are introduced as descriptive of what exists without prior explanation of their generation or source, there are no properties that can serve as the basis for tests of viability. Thus a straight line may be fitted to a scatter of observed points in consumption-income space to yield C = a + bY in the short run or C = bY in the long run, where C denotes consumption and Y income, and a and b are appropriate parameters. But since correlation does not imply causality, how do we know that such functions actually play fundamental roles in determining macroeconomic behavior? We may assume that they do, but on what basis do we accept such assumptions? Previous discussion suggests that the acceptance of these and other assumptions rests entirely on the cultural heritage of the economy in question. In the case of the US economy and based on what is known to be culturally important to Americans in directing their behavior, assuming consumptions functions of the above form make good sense. That is, American cultural patterns suggest, as described above, that a person’s consumption is determined in large part by his income. It is therefore quite natural and appropriate to apply that fact in the abstractions that attempt to describe the macroeconomy. But with respect to economies whose cultural heritages do not propel behavior in the same or similar directions, it would not be correct to employ such relations even if they could be empirically extracted from scatters of points. Not only may the relevant consumption functions have forms different from those of the American economy indicated above, but it may be appropriate to think of consumption itself as dependent on a variable or variables other than income.8 Once again, parametric variation is insufficient to account for the cultural variation that may be present.

8 Of course, consumption in the US economy is more generally thought to depend also on variables such as the rate of interest, etc. But a similar argument could be made with respect to these variables too when cultural differences warrant their omission as determinants of aggregate consumption.

54 “What are the questions?” One may conclude, then, that to the extent that cultural values such as the sanctity of the individual, self-reliance, success, and the pursuit of self-interest are lacking, and to the extent that there are no other cultural elements that would lead individuals to the quest for wealth or to consume more as their incomes rise, models of individual behavior based on postulates of rationality and macro Keynesiancross models (with short- or long-run consumption functions) which, as has been suggested, are relevant for explanations of American economic behavior, do not apply.9 Such may, in fact, be the case with Japan. Now the significant cultural values in Japan traditionally raise the importance of the group (e.g. close family, close friends, employees of a company, and society in general) far above that of the individual. They also include the fulfillment of individuals’ obligations to family, superiors, and others who have done favors for them, loyalty, the maintenance of honor, and harmony within groups and with society and nature. In the economic sphere, these values have come to imply hard work, frugality, and the restriction of consumption to a minimum, and honesty, credit worthiness, and the provision of quality in business.10 Japanese values are not the same as those held by Americans and, as a result, the symbols and models employed by the two societies are different. This is not only true in the present, but extends far back into the past. Evidently the two cultures have evolved over the years in strikingly different ways, and the contribution to contemporary culture that has been historically made by economic, political, social, and intellectual forces has been dissimilar in each case. It follows that directed reasoning by the Japanese is likely to proceed along different lines and toward different ends than that of Americans.11 The emotional elements that impinge on thought are likely to be dissimilar too. Since, as has been argued, motivations are the result of a coming together of directed reasoning and emotion, one would expect the Japanese to be motivated differently than Americans. Indeed, the pursuit of wealth and selfinterest does not generally seem to be the driving force behind Japanese action. Thus models based on traditional postulates of rationality would not appear to be relevant in explaining the economic behavior of consumers and firms in Japan. Moreover, the Japanese tend culturally and emotionally to be more withdrawn into themselves than Americans. This is so for at least two reasons. First, the search for harmony in society and nature by individuals in Japan is primarily an inwardlooking quest. By that it is meant that the complex tangle of historical and, to some extent, religious and philosophical influences prevalent in Japan has created in the Japanese mind an internally developed disposition to adjust individual behavior in such a manner as to harmonize with the interests of external groups and society, and with the physical environment. Second, except for intimate family, the people

9 Experimental evidence that self-interest-based models are not relevant for explanations of individual behavior in a variety of cultures has been given by Henrich et al. [12]. 10 A description of how some of these cultural values emerged in Japanese society has been provided by Bellah [2]. Bellah’s argument has been summarized by Katzner [13] (this volume, Essay 2). 11 This is also suggested by the laboratory evidence cited earlier in this volume on p. 6.

“What are the questions?” 55 with whom Japanese persons work closely, and small numbers of friends (and the latter two categories often overlap considerably), the Japanese are wary of interacting with other Japanese. On the one hand, they do not want to increase the heavy burden of obligations that they and others already carry. On the other, they go to great lengths to avoid taking actions and making mistakes in relation to others that would bring possible dishonor both to themselves and to the others involved.12 Monetary saving, too, is an act of withdrawal. It conforms to, and is consistent with, these other actions that lead to withdrawal. It is also derived from the general value of frugality mentioned above. However, such an inclination to withdraw combined with, or as a part of, a general cultural priority to save would tend to have a weakening effect on any relation between consumption and income that might be thought to exist in parallel to that relevant for Americans and American society as described above. For if consumption is a function of income (or disposable income), then since saving is the residual income after consumption is finished and taxes are paid, saving must also depend on income. But in the Japanese case, as has just been indicated, a relation between saving and income, and hence between consumption and income, is, on theoretical grounds, less clearly established and more problematical, and therefore less predictable. And, in the same way as, for cultural reasons, Americans are outgoing (and decidedly not withdrawn) in their pursuit of wealth and self-interest, so, due to the presence of different cultural patterns, the Japanese do not seek economic success in that form. Nor do they measure success, however it is defined, in terms of material consumption. The postulation of consumption as an increasing function of income might not therefore be appropriate in reference to Japan. It should be noted that while there is some evidence suggesting an empirical relation between savings and income in Japan, the extent to which that relation might be spurious is not known. It might be thought, on the surface of things, that the consumption function as conventionally articulated is, in fact, relevant to both Japanese and American economies as to its general form, differing only in its parametric specification and possibly in its explanatory variables. But what is at issue here is the more fundamental reality that, as indicated above, vastly different motivations and criteria of behavior enter into the determination of Japanese economic decisions and actions, so much so as to make direct transliteration of models from one economy to the other irrelevant. It follows that an economist who uses the models of American economic behavior described above to explain Japanese economic behavior is exposing himself to a significant possibility of error and irrelevance. Such error and irrelevance, should they be present, would not only lead to the misunderstanding of what is actually happening, but it could also result in mistaken policies and predictions. Thus, for example, a tax cut and other efforts to increase aggregate demand that, based on the Keynesian-cross American macroeconomic model, are intended to energize the economy and help it to emerge from recession might not

12 These ideas are discussed in detail by Benedict [4, Chs. 5–8].

56 “What are the questions?” have that effect. Rather, the result might be to increase saving and possibly, as a consequence, to worsen the actual situation. Although not constituting a proof of this last assertion, it is interesting that the Hashimoto tax cuts of the late 1990s and the “couponing” that followed were both ineffectual in stimulating the Japanese economy.13 It could be argued that those tax cuts and coupons had no effect because they yielded only temporary increases in income. However, the latter increases did not even produce much in the way of temporary expansions of consumer spending – contrary to what would have been predicted by the Keynesian-cross American model’s consumption function were that model actually operative. Most of the funds that individuals received from the tax cuts and coupons simply found their way into greater savings. (In the latter case, the coupons were spent, and the money that would have been used had the coupons not been issued was saved.) It is not clear what would have happened if these increases in income had been “permanent.” But in any case, the relevance to Japan of the assumption that consumption depends on income is still brought into question. The lack of response to the Japanese tax cuts and couponing, if it were to be interpreted in terms of the Keynesian-cross American model, could be said to be “perverse” and unexplainable. But in light of the argument of this and the preceding section, it can be asserted that the model itself is inappropriate for providing policies to end recessions in Japan. Previous discussion suggests that what is needed is a model that accounts for the role of a very different culture in determining Japanese economic behavior and that has the potential of giving some indication of the kind of policies that are likely to succeed.

III Some questions that arise within the context of a particular discipline seem to emerge inevitably from that discipline without regard to the cultural setting to which it is applied. Thus, for example, as any student of introductory economics is likely to “know,” the questions (not the answers!) of what is to be produced, how that production is to take place (i.e. with what inputs), and who is to obtain what is produced, can be asked in reference to any economy, regardless of its culture. Still, in light of preceding argument, it is necessary to consider the possibility that the nature, meaning, and significance of the discipline itself may vary from culture to culture, thereby raising doubts about just how independent of culture such questions and their importance really are. But before generalizing further and taking up the issue of the relation of economics as a discipline to the culture of the

13 Hashimoto cut individual income taxes by 2 trillion yen on two occasions during 1998. The total tax reduction was 4 trillion yen and taxes went back up to their old levels in 1999. But during the latter year, Hashimoto’s successor Obuchi issued 0.7 trillion yen worth of coupons to households with at least one child under 16 years of age and to low-income, elderly individuals. The coupons were monetary gifts that could only be used for purchasing goods.

“What are the questions?” 57 economy under investigation, it is both appropriate and convenient to introduce Gudeman’s distinction [9, p. 28] between what he refers to as the “universalist” approach to explanation on the one hand and the “localist” approach on the other. Briefly, universalism means that economic explanation transcends culture. It implies that one explanation or model applies to the description of economic reality in all cultures everywhere. From the localist perspective individuals explain only on the basis of their own understandings of the reality that surrounds them and, as has been seen, those understandings are defined in terms of the cultural patterns of the environment within which the individuals live. This is taken to imply that explanation necessarily varies from culture to culture (a conclusion supported by previous argument) and that, in the absence of significant and extensive cultural similarities, it is not possible for a person submerged in one culture to understand accurately and explain the behavior of a person embedded in another. Clearly, in light of the argument of Sections I and II above, universalism seems overwrought as a basis for explanation and has to be rejected.14 On the other hand, were localism accepted in its entirety, then the discipline of economics itself, including the questions it asks, would vary from culture to culture. Moreover, without major and far-reaching cultural similarities, there would be no common ground on which to describe the intellectual content of the discipline of economics, and persons of one culture would neither recognize nor understand it as it arose in a foreign cultural environment. But causal observation of interactions among people of very different cultures suggests that it is often possible for a person of one culture to gain some appreciation and understanding of a person, and his behavior, of another. In parallel with universalism, then, localism also appears to be too extreme an approach to explanation and it, too, is rejected. The position taken in the present essay falls between that of universalism and localism as defined above. Succinctly stated, it is that in spite of the fact that cultural differences produce different behaviors and require different explanations for understanding those behaviors, sufficient human interaction takes place among people of different cultural backgrounds to permit translations of meanings across cultural boundaries, however imperfect, to be made. Thus the discipline of economics within any culture can be both recognized and understood by scholars of foreign cultures as long as the appropriate cultural discrepancies are acknowledged and carefully accounted for. Furthermore, to the extent that the symbols people use and the standards they employ for perception, judgment, and action vary across cultures, one would expect peoples of different cultures to understand their worlds differently and to hold different features relating to them as important. One would also expect them to ask at least some different questions in thinking about their worlds and to provide answers to questions in general that are often unique to their

14 This is not to say that explanation of international trade among countries with contrasting cultures is not possible. Indeed, it is part of the argument of this paper that any such explanation has to be sensitive to the cultural differences that are present.

58 “What are the questions?” own cultural environment. To return to the example of the “what?,” “how?,” and “who?” questions mentioned earlier, those questions can certainly be asked of, and answered for, any economy. But the answers may reflect differences in cultural settings, and the questions themselves may have a different resonance and importance from one culture to another. That is, with respect to the latter, other questions may be thought to be of greater significance in the discipline of economics. It is useful to consider the mental process that generates questions which, in the present instance, have possible relevance to economic decisions. A person observes something in the real world or thinks of something in relation to it. That observation or thought is framed in terms of the organized collections of symbols of the culture patterns derived from that person’s environment and etched in his mind. Mental manipulation and directed reasoning then act upon those symbols to interpret and compare the observation or thought with respect to the fundamental analogy and the models the individual has previously derived from that analogy, i.e. the interpretations and comparisons are based on what the individual knows, or thinks he knows, from past experience. If the observation or thought fits seamlessly into the analogy or one or more of those models with no loose or jagged ends, then the individual feels that he has explained and understood it, and there is nothing to stimulate the asking of a question. But if there are contrasts, contradictions, or gaps with respect to the fundamental analogy or one of the models, then understanding and explanation are not satisfactory and a mental tension arises that results in at least one question. Viewing the determination of questions in this manner, it is clear that culture plays a fundamental role in constructing and formulating the questions that are asked about economic activity. It also suggests that the greater the tension, the more significant and the more important the questions emerging from that tension are likely to be regarded. Moreover, since questions are asked in context, and since context exists only in reference to symbols in the mind, culture influences context as well. Based on the discussion of Sections I and II, parallel statements may be made about the role of culture in generating answers to questions, and about the manner in which cultural variation across societies can give rise to different, culturally specific answers to the same question. Indeed, everything relating to the questions asked for the purpose of explaining economic action rests in some way on culture. Thus the answer to the question, “What are the questions?” cannot be complete without at least an implicit cultural referent. Returning to the contrast between explanations of American and Japanese economic behaviors and their cultural determinants described in Section II, one would expect that questions asked about the American economy would be different from those inquired of the Japanese economy. Since American culture is, to a predominant extent, focused on individualism, economic questions relating to the well-being of the individual ought to be of paramount importance when studying the American economy. But when thinking about Japan, where cultural imperatives direct that groups are much more important than individuals, it is economic questions that bear on group well-being that matter. Thus, rather than asking what

“What are the questions?” 59 the firm has to do to maximize the rate of return to its shareholders, a question that is certainly relevant and significant for understanding the American economy, the appropriate question in reference to the Japanese economy might be how can the firm best serve its employees and the community at large. And rather than asking, as would again be suitable in the case of explaining American economic phenomena, how the individual might best pursue his self-interest and accumulate the most wealth, in relation to Japan it might be more appropriate to ask how the individual can best fulfill his economic obligations and thereby enhance the employment group to which he belongs and the harmony within it. The nature of the economic questions that are asked, then, depends in a very fundamental way on the cultural environment in which the economic behavior to which those questions are addressed takes place. The objective of this essay has been that of clarifying the significance of appropriate questions and appropriate methodological approaches to culturallydetermined economic contexts and behaviors. Since many current economic questions have been raised and many currently existing economic models have been constructed for economies with Western cultural backgrounds, there is little difficulty in posing those questions and applying those models in similar cultural circumstances. But for economies whose cultural heritages differ significantly from that of the West, the detailed task of constructing suitably specified explanatory models lies ahead, and the relevant procedures will clearly be as various as the cultural complexes in which they are assumed to apply.

References 1. Becker, G.S., The Economic Approach to Human Behavior (Chicago: University of Chicago Press, 1976). 2. Bellah, R.N., Tokugawa Religion (Glencoe: Free Press, 1957). 3. Bellah, R.N., R. Madsen, W.M. Sullivan, A. Swidler, and S.M. Tipton, Habits of the Heart (Berkeley: University of California Press, 1985). 4. Benedict, R., The Chrysanthemum and the Sword (Boston: Houghton Mifflin, 1946). 5. Boyd, R. and P.J. Richerson, Culture and the Evolutionary Process (Chicago: University of Chicago Press, 1985). 6. Cavalli-Sforza, L.L. and F. Cavalli-Sforza, The Great Human Diasporas, S. Thorne, trans. (Reading: Addison-Wesley, 1995). 7. Geertz, C., The Interpretation of Cultures (New York: Basic Books, 1973). 8. Greif, A., “Cultural Beliefs and the Organization of Society: A Historical and Theoretical Reflection on Collectivist and Individualist Societies,” Journal of Political Economy 102 (1994), pp. 912–950. 9. Gudeman, S., Economics as Culture (London: Routledge & Kegan Paul, 1986). 10. Gudeman, S. and A. Rivera, Conversations in Columbia (Cambridge: Cambridge University Press, 1990). 11. Hampden-Turner, C. and A. Trompenaars, The Seven Cultures of Capitalism (New York: Doubleday, 1993). 12. Henrich, J., R. Boyd, S. Bowles, C. Camerer, E. Fehr, H. Gintis, and R. McElreath, “In Search of Homo Economicus: Behavioral Experiments in 15 Small-Scale Societies,” American Economic Review 91, no. 2 (May, 2001), pp. 73–78.

60 “What are the questions?” 13. Katzner, D.W., “Western Economics and the Economy of Japan,” Journal of Post Keynesian Economics 21 (1998–99), pp. 503–521. This volume: Essay 2. 14. North, D.C., “Institutions and Economic Theory,” American Economist 36, no. 1 (Spring, 1992), pp. 3–6. 15. Pepper, S.C., World Hypotheses (Berkeley: University of California Press, 1942). 16. Reed, S.R., Making Common Sense of Japan (Pittsburgh: University of Pittsburgh Press, 1993). 17. Robinson, J., “What Are the Questions?” Journal of Economic Literature 15 (1977), pp. 1318–1339. 18. Ross, L. and R.E. Nisbett, The Person and the Situation (Philadelphia: Temple University Press, 1991). 19. Roth, A.E., V. Prasnikar, M. Okuno-Fujiwara, and S. Zamir, “Bargaining and Market Behavior in Jerusalem, Ljubljana, Pittsburgh, and Tokyo: An Experimental Study,” American Economic Review 81 (1991), pp. 1068–1095. 20. Schultz, T.W., Economic Crises in World Agriculture (Ann Arbor: University of Michigan Press, 1965). 21. Sen, A.K., “Choice Functions and Revealed Preference,” Econometrica 61 (1993), pp. 495–521. 22. Smith, A., An Inquiry into the Nature and Causes of the Wealth of Nations (New York: Random House, 1937). 23. Stiglitz, J.E., “Rational Peasants, Efficient Institutions, and a Theory of Rural Organization: Methodological Remarks for Development Economics,” The Economic Theory of Agrarian Institutions, P. Bardhan, ed. (Oxford: Oxford University Press, 1989), pp. 18–29. 24. Weber, M., The Protestant Ethic and the Spirit of Capitalism, T. Parsons, trans. (New York: Scribner, 1958).

4

An analytical vision of the workings of the United States economy1

The preceding essays have established the principal thesis that engages present discussion, namely, that an understanding of the relation between culture and economic behavior is necessary to a clear and adequate grasp of the functioning of a given economic system. In this essay and the next, that thesis will be illustrated by a comparative investigation of the workings of the American and Japanese economies, respectively. Both economies are, of course, monetary economies in that money operates as a medium of exchange in market activity and as a store of value over distance and time. In market trading, goods exchange for money and money exchanges for goods, but usually goods do not exchange for goods. Further, the savings that are made out of incomes are generally deposited in savings and other financial institutions which, in turn, lend those funds in the form of money to firms and individuals for investment purposes. In so doing, a price for the use of monetary funds, expressed as a rate of interest, is chargeable to borrowers and payable to savers. Finally, there exists in each economy an apparatus of banking institutions that participates in various ways in the saving-investment process and in the provision of the economy’s money supply. There are many differences that relate to the functioning of money in the United States and Japanese economies. But for present purposes it will be convenient, without loss of generality insofar as contrasting the two economies is concerned, mostly (there will be a few exceptions) to set aside monetary features and forces. Present interest is in comparisons of the main determinants and their implications of economic behavior at the levels of the consumer of produced goods and the firms who produce them, focusing, as has been indicated, on the impact and influence of cultural elements. Turning to the United States itself, an American, when thinking about himself and his position in the vast economy in which he participates, is typically concerned with a number of central issues. He cares about his economic status and welfare (including that of his family). He wonders how secure his job might be. And he evaluates the extent to which the economy rewards his economic contribution

1 The help of Douglas Vickers is gratefully acknowledged.

62 An analytical vision of the workings of the United States economy in relation to what he thinks he deserves. Each of these issues gives rise to a whole host of specific questions relating to employment, wages, prices, economic growth, the distribution of income and wealth, the power of labor unions and various industries, and so on. In what follows, various characteristics of the American economy are described and a widely held vision is articulated that purports to provide an explanation of how that economy works. In the process, a particular understanding emerges of the manner in which the concerns listed above are addressed. Now in elucidating the vision of the workings of the US economy, it is necessary to explain how certain economic decisions, behaviors, and consequent outcomes come about. And in the view presented here, even in its most idealized form, the latter outcomes depend on the distributions of income, wealth, and talents of individuals existing as they enter the economic process. Thus, for example, an individual starting out poor and with little skill, is likely, though not necessarily, to remain so at subsequent stages at which economic outcomes obtain. It is an open question, then, as to how satisfactory the answers provided by the subsequently displayed view of the American economy are to the concerns raised above. That view, of course, does not speak to all matters thought to be important by every person; nor does it delve in great detail into those matters it does consider. But the image that emerges is sufficiently abstract and general, and resonates with a sufficiently large and broad population, to serve for present purposes as a “standard” vision of the functioning of the US economy. As such, it will provide a basis for comparison in the next essay with the corresponding portrait of the workings of the Japanese economy. Before beginning, several caveats are in order. First, in presenting the vision of the operation of the American economy that follows, it will be useful to bring briefly and informally into focus a slightly abstract or technical approach to economic explanation. For that purpose, a number of purportedly explanatory assumptions are made. These assumptions imply what is generally referred to by economists as “perfect competition.” For example, in the more technical mode, it is assumed that all participants in all markets possess full knowledge of all relevant economic elements and configurations and all possible outcomes. Along with this is the supposition that the conjuncture of market forces in operation at any moment is such as to lead to outcomes associated with the full employment of all of the economy’s resources, including the resource labor. These are, of course, rather severe restrictions. The real world of industrial and economic affairs clearly departs from what they intend or imply. Moreover, and in partial recapitulation, although the perfectly competitive vision is assumed here to be associated with automatic harmonization of all economic forces, in actual fact the United States economy has a natural tendency to disruption due to the continual presence of intervening and destabilizing pressures. Those tendencies to disruption may be due to imperfect estimates and anticipations of commodity supplies and demands (including overinvestment or underinvestment in inventories), to intertemporal distortions between saving and investment, to excessive concentrations of market power in the hands of monopolies and oligopolies, and to the impact on markets of government regulating, purchasing, taxing, and transferring activities. It will

An analytical vision of the workings of the United States economy 63 therefore be necessary, as subsequent argument proceeds, to indicate instances where the relaxation of the assumptions of perfect competition may, in fact, lead to different, more realistic, behaviors and possible outcomes. Second, and in the same spirit that leads to an accounting of some aspects of the aforementioned forces of disruption, it will also be necessary at certain points to call attention to particular intertemporal considerations that might result in the overriding of static or fully-disconnected-from-time conclusions. For the vision presented below is primarily static in nature. It generally speaks to the determination of economic outcomes independently of time. But some economic decisions, notably those relating to business firms’ production and investment activities, frequently take into account possible sequences of outcomes and events across time. In the case of investment decisions in particular, firms would be concerned, in part, with the income stream realized over future time periods from possible investments, and this, in turn, implies the need to allow for the risks and uncertainties that might arise from potential instabilities in those streams. Such considerations, along with the degree of acceptable risk that firms are prepared to take on, influence, through the decisions that come forth, the allocation of resources and the distribution of outputs in the economy. While all of this has considerable real-world significance, it will for the most part, and except for occasional reference, be (like the monetary features previously mentioned) passed over in the following. Finally, it should be noted that, in actual fact, decisions and resulting behaviors in America are often influenced by conceivably small numbers of opinion formers. Opinion formers project their ideas in newspapers, magazines, and books, and on radio, television, and the internet. They can be instrumental in shaping the understandings and expectations upon which decisions are based. But while opinion formers may play a role in the generation of decisions and behaviors, their presence in itself does not affect what is described below as the structure of market interactions among consumers and firms, or buyers and sellers, throughout the economy.

I The United States economy, broadly speaking, is made up of consumers, firms (including financial institutions), and markets. Consumers are individuals who buy goods and services from firms, sell labor (and sometimes sell or rent land and buildings) to firms, and make their savings available to firms by purchasing newly issued stocks and bonds either directly or through financial intermediaries in which their monetary savings have been deposited. Old issues of stocks and bonds may be bought with savings, but that will not directly represent additional funds made available to firms by consumers. The latter may occur indirectly, however, if the sellers of the old issues dispose of the funds they acquire by purchasing new issues of corporate capital. Consumers’ incomes, of course, derive primarily from the sale of labor and the returns accruing from earlier purchases of stocks and bonds, and, ignoring tax payments to the government, are either used to purchase goods and

64 An analytical vision of the workings of the United States economy services or saved. Firms are institutions that hire inputs and use them to produce outputs. They also obtain funds by, in part, issuing stocks and bonds to finance the acquisition and expansion of their productive capacity and by retaining annual earnings for that purpose. Firms receive revenue from the sale of their output and use it to pay for inputs, cover interest payments on their debt, and provide a return to their owners. Markets are separate institutions whose purpose is to facilitate the buying and selling of goods and services (including labor, stocks, bonds, and industrial and consumer assets). It is the buying and selling decisions of consumers and firms that generate most of the economic activity in the economy. The making of these decisions requires taking into account such elements as expectations, long-term and shortterm strategies, and the use of money for transactions purposes and as a store of value. But, as previously indicated, the latter elements, although significant for the individual’s perception of his place and prospects in a monetary economy, are, with a few brief exceptions, set aside in order to focus on those features of the American economic system that are more relevant for present purposes. Other matters like those relating to consumer credit and the banking system are also ignored. The government, a third important player in the economy after consumers and firms, will be introduced later on. In their economic lives, as in other aspects of their daily existence, Americans are directed, in part, by the values they hold. As has been described elsewhere (for example, pp. 50–51 above and Katzner [1, Ch. 5]), these include the sanctity of the individual, self-reliance, success, freedom, and justice. The first three values on this list merge and emerge in what is often understood as the fundamental driving force of American economic behavior, namely, that Americans generally engage in the pursuit of their own well-being. It is up to each person, as an individual, to succeed. In the push towards success, the only one the individual can rely on is himself. And the success he actually achieves is frequently understood to be measured by the physical possessions he accumulates. This is one of the reasons why Americans place so much emphasis on material consumption. Regardless, all actions are based on calculations whose purpose is to determine the best way to achieve personal goals. In this sense, the individual or consumer always behaves in his own self-interest. That self-interest, of course, may include looking out for family or other group interests, or individuals whose well-being affects his own. It is self-interest that leads to his concerns relating to economic status and welfare, job security, and the rewards for contributing to economic activity mentioned at the outset of this essay. And it is the pursuit of self-interest that is always pushing the individual to contemplate his best next move. In terms of the more technical vision explaining economic action, the facts that self-interested choices are often tempered by habit or impulse and influenced by fashion and advertising are ignored. Thus self-interest means that, within the constraints imposed by the economic environment in which he operates, including the resources at his disposal, each person buys the commodities (and in the amounts) he wants most; that he employs himself where and to the extent that is, to him, most satisfactory, given his abilities and his desire to earn

An analytical vision of the workings of the United States economy 65 income for consumption purposes (the technical vision assumes that the relevant employment opportunities are available); and that, when able, he engages in financial transactions, like the purchase of stocks and bonds, to maximize his expected future income and wealth. Within this calculus it is, respectively, the preferences he has among various baskets of commodities, among consumption possibilities in general and leisure time in particular, and among numerous combinations of risks and returns, that impart concreteness to his goals and the actions he undertakes to meet them. In each case, the prices of the things in question (goods, services, and assets) and the income necessary to forgo in order to have leisure, along with whatever income he already has available, or would have available to spend (as set in relation to his wealth) determine the particular opportunities among which he is able to choose. Similarly, decisions on whether to attend college would turn on the possible costs of attending various institutions as compared with the expected income streams that an education at those institutions would secure. Once the various opportunities in any instance are delineated, the individual selects the most preferred among the options he faces. Firms in the United States differ widely in size. But regardless of size, they often hire persons in mid-career from other firms and they pay similar wages and salaries for similar jobs performed, irrespective of seniority. Employment in small firms is as “permanent” or “temporary” as that in larger firms. Pricing policies, which often involve marking up from unit cost depend more on the competitive environment in which the firm operates than on the size of the firm. However, large firms tend to be owned more by institutions and smaller firms more by individuals. Because firms are owned by individuals, or institutions run by individuals, who are rewarded (in the form of dividends, or with higher stock prices if earnings are retained) when the firm profits from its activities, and because individuals, and hence institutions, seek to maximize their expected wealth and income, selfinterest is also reflected in the operation of firms. Since it is the only way to satisfy their owners, firms must continually strive to produce outputs and hire inputs so as to maximize profits, although there are instances when a firm may forgo shortterm profits for greater long-term gains. And, in the context of the technical vision employed here to explain economic activity, neither size nor mark-up pricing policies play a role. Moreover, under ideal (i.e. perfectly competitive) conditions to be specified below, profit maximization has a number of implications, including (without going into the reasoning behind them) the following: First, the firm needs to make certain that it always obtains the largest output technologically possible from any collection of inputs it employs. This is referred to as productive or technological efficiency. Second, the combination of materials and labor used in production must be such as to minimize the production cost of each volume of output. That is an aspect of economic efficiency, further features and implications of which will be noted to below. Thus firms would have to be sure that the levels of output they produce could not be obtained with smaller expenditure on inputs. Alternatively, they would have to endeavor to combine inputs in such a manner that, keeping the total cost of inputs the same (and assuming no change

66 An analytical vision of the workings of the United States economy in input prices), increasing the quantities used of some inputs and decreasing those of others would not yield larger output. Third, each input employed would have to be hired in quantity up to the point at which the value to the firm of the last unit used is the same as its cost, that is, the last unit of labor hired is paid what it is “worth.” Were a smaller quantity of input hired, the value to the firm of the last unit would, in that case, be greater than its cost. Were a larger quantity hired, the latter cost would be greater than its value to the firm. And, fourth, outputs would have to be produced in quantities such that the cost of producing the last unit equals its selling price. In parallel to the situation with respect to inputs, no unit of output, were it the last to be produced, should cost more than the price at which it is sold. (Somewhat analogous considerations determine when the firm borrows funds or uses retained earnings to expand its productive capacity, or uses its profit to retire previously accumulated debt. From here on, however, it will be convenient to simplify matters by disregarding the purchase and sale of stocks and bonds.) Were any of these conditions not met, it would be possible, by adjusting input and output quantities, to increase the profitability of the firm. For consumers to engage in the activities of buying goods from firms and selling labor to them, and for firms to hire inputs from consumers and other firms and to sell output to them, requires, in each case, and as suggested earlier, that self-interested decisions or choices be made. In making these choices, American values dictate that consumers and firms should have complete freedom, subject to minimal legal and other institutional constraints, to do as they wish. In addition, the American sense of justice (another of these values) necessitates that each individual consumer or firm has equal opportunity in making his or its choices. Without freedom of choice and equal opportunity, an individual may be prevented from achieving what might otherwise be possible. Through no fault of his own, his chances of success would be lessened, and the sanctity of the individual would be improperly tempered. Impingement on freedom of, and equal opportunity of, choice occurs, for example, when a labor union, discrimination by an employer, or a lack of skills prevents an individual from entering on a particular career path. In addition, high concentrations of economic power in an industry may block a potential investor from entering the industry or hinder buyers from securing the kind of products they want. In any case, both freedom of choice and equal opportunity are to be maintained and protected to the fullest extent possible by the government. It is clear that both consumers and firms need information to make their choices. This is certainly true with respect to real world activity. It is also true to a much greater extent in relation to the more technical vision of explanation invoked above. Focusing on the latter, to know what to produce and how to maximize profit, firms need to know what consumers want. They also have to know the technical features concerning how inputs can be combined to produce outputs, how much inputs cost, and for how much outputs can be sold. The technical features of production may be assumed to be already known by firms. Information about what consumers want is obtained by observing the relative prices of the various possibilities. Loosely speaking, the more a good is desired, the higher its price although, of course,

An analytical vision of the workings of the United States economy 67 a price may be high due to unrelated factors such as seasonal variation in supply, speculation, changes in expectations, and the presence of a transient fad. Given knowledge of the technical features of production, then, prices act as rough signals to producers that help them to determine what (and how much) to produce, and what inputs (and in what quantities they) should be used in that production. Thus, as a rule, the higher the price of an output arising as a consequence of greater demand, the more that will be produced because, by increasing production, profits will be enhanced. Similarly, the more expensive an input, the more firms will try to find cheaper substitutes for it. The use of cheaper substitutes would mean lower production costs for each level of output, and greater profits. In the same way, and as suggested earlier, assuming a consumer is aware of his preferences, knowledge of prices permits him to calculate the goods and services (and how much of them) he should buy, and the kind and amount of labor he should sell so as to achieve his most preferred position. Prices, in turn, are generally determined by the interaction of consumers and firms in markets as they pursue the decisions they have made regarding what to buy (demand) and what to sell (supply). Real world markets, of course, are highly specific to location. In a supermarket, for example, firms provide products for consumers to buy. There are many products and their prices need not be identical to those of the same commodities sold in other markets. But the explanatory technical vision in focus here abstracts from place, the multiplicity of goods in the same market, and the possibility of differentiated pricing. As a rule, each good or service has a single market and each individual market handles quantities of only one good or service and functions so as to establish a single price for that good. The well-known demand-supply framework is often employed as the basis for understanding how actual prices are established in markets in general. For imagine a market with unchanging conditions of demand and supply that establish an actual price at a particular moment in time. Were a price to exist that was below the actual price, consumers and firms would want to buy more of the market’s good than is produced. The lower price would then rise and more of the good would be produced and supplied on the market. The gap between what is demanded and what is supplied would thereby be reduced. Were a price to exist that was above the actual price, consumers and firms would wish to buy less of what is available, the higher price would fall, some of what was supplied would be taken off of the market, and again, the gap between quantities demand and supplied would decline. In the first instance, sellers raise the price because there are persons or firms willing to pay more; in the second, sellers lower the price to induce buyers to purchase more. This activity is often referred to as competition because, on the one hand, buyers are competing among each other to buy “scarce” goods and, on the other, sellers are competing among each other to sell their wares to “scarce” buyers. In the technical vision, movements of prices in response to shortages and excesses, and what that implies for the achievement of profits and the reduction of the gap between amounts demanded and supplied, will occur until marketclearing (that is, actual) prices are established everywhere such that consumers

68 An analytical vision of the workings of the United States economy and producers are satisfied with the outcomes.2 At that point, the allocation of goods, services, inputs, and outputs among all consumers and firms in the economy will be consistent with maximal consumer satisfaction out of what is available to them and the maximization of profits by firms. As was suggested earlier and as will be noted again below, concentrations of economic or market power may inhibit the flexibility of price movements and influence allocations. But, under the ideal conditions still to be detailed below, as long as prices are free to move up or down, and consumers and firms are free to try to buy and sell what they wish, everyone will, in the end, be able to buy and sell exactly what he wants at prevailing prices. In particular, anyone who wants to work at the going wages will be able to find a job. Of course, the ideal conditions under which this last result obtains may not arise in actual fact because of wider macroeconomic forces that cause cyclical fluctuations in economic activity. But like concentrations of market power, this possibility is, for convenience, set aside to concentrate on other fundamentals of market operations. Thus the economic system, which consists of self-interested consumers, profit-maximizing firms, and markets, determines what goods and services will be produced, how or with what inputs they will be produced, and who will get what has been produced. The quantities so determined along with the prices established in the various markets represent the outcome of the economic activity of the economy’s buyers and sellers. There are, as has been suggested above, implicit assumptions in the preceding description of price determination and the allocation of goods and resources in the economy. One of these is the presence of full employment of labor and all other resources. Another is the absence of “interference” from behavior with respect to money. To illustrate the latter, were people suddenly to use part of their incomes to increase their holdings of money rather than spend it on the purchases of goods and services, then aggregate spending in the economy could decline and, along with it, opportunities for the full employment of labor. Regardless, when either or both of these assumptions are violated, as they might well be in reality, the referred-to descriptions need not hold. If it were assumed that economic markets generally functioned as previously described under the ideal conditions and assumptions alluded to (and shortly to be spelled out in further detail), then, with respect to the outcome of economic activity, the price of a produced good or service may be thought of as measuring its value or benefit to society at large. Furthermore, the cost of producing the last unit of that good tends to reflect or measure the value of the opportunities that society forgoes in order to acquire it. Since, as has already been indicated, these two values equalize, the economic system, in its production and allocation of commodities, tends to balance the benefits of each good against the value of opportunities that might have been secured. This property, too, requires the implicit assumptions cited immediately above to be in force.

2 The ideal situation contemplated includes the supposition that the market-clearing price of labor will be at or above the level required for labor’s subsistence.

An analytical vision of the workings of the United States economy 69 Continuing with the explanatory technical vision of this essay, any impediments to the free movement of, say, prices, such as minimum wage laws, which prevent wages from falling when too much labor is offered to firms by workers, or rent control laws, which prevent rents from rising when too few rental properties are available to renters, will negate some portion of the requisite ideal conditions and, hence, interfere with the process of price determination and allocation of goods, services, inputs, and outputs by the economic system. And when this happens, not all consumers and firms will be able to buy and sell what they want at prevailing prices. Under these latter conditions, nonmarket schemes like “firstcome-first-served” and the emergence of black markets will supplement the market mechanism in the price-determination and allocation process, and the outcome will be quite different than would be the case if the economic system were left to function freely on its own. The environment in which a market that is free from such impediments operates depends on certain ideal structural characteristics that determine the manner in which the buyers and sellers in that market interact. This environment sets the tone of the competition among buyers and sellers and also places limits on the price and quantity outcomes that can be obtained in the market. The structural characteristics required of a market to have competition in its purest (or perfect) form are first, that there are a sufficiently large number of buyers and sellers and all buyers and sellers are sufficiently small so that no individual buyer or seller has any influence in determining the price that emerges as a consequence of buyerseller interaction; second, that all units of the product supplied are similar enough so that no buyer cares from which seller he purchases the good; third, that there are no barriers (such as a patent or licence, limited ownership of a scarce resource, or prohibitive set-up costs) preventing entry into and exit from the market by either buyers or sellers; and fourth, that all information relevant to the sale and purchase of the good is available to, and possessed by, all market participants. In the more technical vision of economic explanation, the ideal economic system, or what is referred to in the next essay as the competitive ideal, is one in which the above form of competition prevails in all markets everywhere, in which consumers are driven by self-interest and firms by profit maximization under the additional conditions set out previously, and in which certain other standards that it is not necessary to identify here are met. It should be mentioned that the assumptions underlying the ideal economic system carry certain implications regarding the technical concept of efficiency, as that is understood both at the level of production by firms and in reference to the global outcomes created by the economy as a whole. With respect to the latter, the ideal system efficiently delivers, in a sense to be clarified subsequently, the goods and services, and in the exact quantities, that people want. Only goods and services that are scarce, and hence that have to be purchased at nonzero prices, are considered, and it is assumed that no social or cultural opprobrium is associated with them. It is also assumed that no “snob-value” attaches to any good in the sense that it may be purchased because its price is high or as a means of achieving a certain social status. The benefits of the ideal system, along with some of its limitations, such as

70 An analytical vision of the workings of the United States economy the fact that the outcomes it produces depend on the distribution of income and previously accumulated wealth with which the system begins, will be considered later on. Although it is possible in that discussion to do so, the outcome of the ideal economic system will not be evaluated with respect to either the general welfare it provides society or the extent to which it might be thought to provide economic fairness and justice. When considering the structure and functioning of the American industrial economy, it is important to observe more explicitly the large number of respects in which the specialized assumptions of the ideal economic system as exhibited up to this point do not, in reality, apply. By introducing appropriate modifications in those and other assumptions, it is possible to move the technical vision explaining how the economy works closer to what might actually take place. Non-ideal forms of market competition are considered first. In one form all but the second of the above four structural characteristics of market environments obtain. That is, there are still a large number of small buyers and sellers, entry and exit barriers are absent, and the requisite information is available. But now, each seller in the market is offering, at least in the minds of some buyers, a slightly different commodity or service. These products are sufficiently distinct that buyers are not indifferent when choosing among them. The variety of goods proffered can be quite large, as is the case in the market for, say, breakfast cereals or automobiles. None of the goods, however, is different enough to be classified as a product in a separate market. Nevertheless, the market is still characterized by a degree of competition that is less than the ideal, and under such conditions the price at which the seller offers his output is usually determined as a mark-up from unit cost. The amount of that mark-up, moreover, will not be very substantial due to the intense competition among the large number of firms in the market. Competition becomes even more confined when, regardless of whether market products are distinct or similar, buyers and (or) sellers are larger and there are fewer of them. Such circumstances, which will be referred to as situations with small numbers of large buyers and (or) sellers, could arise for many reasons, including the existence of barriers that restrict entry to and exit from the market. In any event, one or more of the buyers or sellers, in particular those that are larger, are able to exert influence over the price established in the market, and it is the competition between these limited number of buyers or sellers that, in such special circumstances, determines the established price in relation to the sellers’ unit costs. Here, because there are fewer competing firms involved, the mark-up is likely to be greater than in the situation with larger numbers of small firms described previously. Finally, at the less-competitive end of this spectrum, lies the case in which there is only one buyer or one seller. In the latter instance (a monopoly), it is the seller alone who, subject to market demand conditions, sets the market price as a mark-up from unit cost – without the pressure of competition to act as a restraint. Casual observation suggests that the ideal or purest form of competition does not obtain in most markets of the US economy. However, Shepherd [2, p. 618] has estimated that in 1980 about three-quarters of domestic sales occurred in markets

An analytical vision of the workings of the United States economy 71 that were “effectively” competitive, at least with respect to the markets’ sellers. This may be interpreted to mean that, in the year in question, three-quarters of all US domestic output was sold in markets with competitive characteristics that approximate, to some reasonable degree, the purest form. But there are additional ways in which the US economy fails to measure up to the ideal system. Firms may charge different buyers different prices for the same good or service, they could use their economic power to deny buyers of their outputs and sellers of their inputs the opportunity to buy those outputs from them and sell those inputs to them, they could produce custom products for individual buyers and sell those products at specially contracted prices, and they could agree to “cost-plus” contracts with certain buyers. Two further instances deserve more detailed comment. Both involve nonconsumer, nonfirm institutions engaged in market activities that are, by dint of their relative size, powerful enough to exert an influence over the prices prevailing in the markets at issue. The first category of institutions is the governments at federal, state, and local levels. The US government buys many products from private industry, such as, for example, those required for national defense, where it is one of a very few, if not the only, buyer. As such, it has considerable influence over the price it pays for those commodities. Furthermore, government purchases have, assuming full employment of all resources, a significant impact on the allocation of inputs and outputs across the economy, since the production of those goods and services draws on inputs that would otherwise be used for different purposes. (Occasionally the government is also a producer and supplier of goods. But these latter cases are rare, and for present purposes, can be safely ignored.) In addition, the government imposes taxes and makes payments, such as subsidies and unemployment compensation, that also affect market prices and the distribution of goods and services, including inputs and outputs, throughout the economy. Similar statements can be made with respect to state and local governments. It should be noted, however, that government expenditure and taxation each amount to no more than one-third of gross domestic product, and hence the impact of government on the US economy is not as large as in many other industrialized countries, for example, in Sweden. Finally, the government also affects the economy through its power to regulate market activity. The second category of institutions consists of the labor or trade unions. Union membership in the US reached its peak in the 1940s and 1950s and has been slowly declining since. But in certain labor markets, unions remain strong enough to have a significant voice in setting market wages. Of course, unions can have such power only under less-than-ideal forms of competition, and firms may attempt to pass on to the buyers of their outputs any wage increases granted in union-management negotiations. In spite of these deviations from the ideal economic system, many Americans still think of the United States economy in ideal terms. There are at least three possible reasons why this is so. First, given the relative sizes of the government and the unions, and recalling the extent of effective competition indicated above, the actual economy in the most general of terms may be regarded as an approximation,

72 An analytical vision of the workings of the United States economy however distant, of the ideal system. Second, the ideal system provides a simple and concrete way of expressing specific characteristics of the US economy that may be thought to be both relevant and important. Some of these have already been described above; more will be considered momentarily. And third, the ideal system can be very satisfying and appealing because certain (though not all) American cultural values resonate strongly with it, and the goals that emerge from those values find clear articulation in its vocabulary.

II Because many Americans often express thoughts about their economy in the language of the ideal economic system of the explanatory technical vision described above, it is appropriate, at this point, to identify several particularities of that system that speak to its performance. Two such characteristics relate to additional aspects of the notion of efficiency, one with respect to the firm and the other in terms of the economy as a whole. Consider the latter first. In that context, efficiency is a feature of certain economic outcomes or, that is, a property of specific allocations of goods and services, including inputs and outputs, across the economy. Roughly speaking, an outcome or allocation is efficient if, within that allocation, people are obtaining the quantities of goods and services they want most, consistent with the consumption satisfaction achieved by others, and the goods and services are produced at minimum input cost. Here the phrase, “obtaining the quantities of goods and services they want most, consistent with the consumption satisfaction achieved by others,” is intended to reflect the idea that, given the levels of satisfaction from consumption secured by others, each person has the goods and services, and in the exact quantities, he most wants out of what remains. And, recall, the phrase “minimum input cost” means that, at prevailing input prices, any other combination of inputs that yields the same output would cost more. Another way to characterize efficiency is that an efficient allocation is one in which it is not possible through any reallocation of goods to make one person better off by providing him with what he considers to be a more preferred basket of goods, without making someone else worse off by forcing him to accept, for him, a less preferred basket. (As used here, the term “goods” is an abbreviation for “goods and services” and also encompasses both input and output quantities.) It turns out that these two approaches to the aspect of efficiency under consideration here essentially amount to the same thing. This is suggested in an informal and incomplete way by the following thought experiment: Suppose, for example, an allocation is inefficient in the first sense because not all production takes place at least input cost. Then by changing inputs appropriately and moving to minimum cost in only one instance saves funds that were formerly spent on inputs without reducing output. Next, employ the funds saved to purchase more inputs (which will be available because of the general inefficiency of the initial allocation), use those extra inputs to produce extra output of one good desired by one person, and give it all to him. Then that person has more of a good that he

An analytical vision of the workings of the United States economy 73 wants and nothing has been taken away from anyone else. He is therefore better off and no one else has been made worse off. The original allocation, then, cannot be efficient in the second sense. This argument illustrates the fact that efficiency in the second sense implies efficiency in the first sense. An argument could also be constructed to intimate the truth of the converse assertion that efficiency in the first sense implies efficiency in the second. One property of an ideal economic system, recalling that such a system implies full utilization of all resources and the absence of monetary interferences, is that, given the distribution of resource endowments and hence of income and wealth from which economic activity starts, it always provides efficient outcomes. That is, it furnishes the goods and services that people want most, consistent with the consumption satisfaction achieved by others, and at minimum input cost. Thus ideal systems are nonwasteful in that, with respect to the outcomes they provide, one person cannot be made better off without those gains having come at the expense of another. (This is not to suggest that, given one nonwasteful outcome, there might not be other nonwasteful outcomes achieved by starting at different income and wealth distributions that society might regard as superior.) Clearly, if one person could be made better off without hurting anyone, then economic waste would be present. Waste occurs, in particular, if the economy is not fully employing all available labor input. It is not necessary to delve into a more extended explanation of why ideal economic systems are efficient. But it should be pointed out that any such explanation has to be based on the fact, cited earlier, that ideal systems always balance the benefits of each good and service against the value of the opportunities that have been foregone to acquire them. Moreover, this efficiency property applies only to ideal systems as described here. These systems, as has been indicated, consist of self-interested consumers who always choose the most preferred option from any collection of opportunities presented to them, firms who always hire inputs and produce outputs to maximize profits, and markets that exhibit competition in its purest form. Any economic system that does not meet these standards, even if it violates them in only minimal ways is not ideal, and its outcomes, therefore, are not likely to be efficient or nonwasteful in the senses previously specified. Indeed, the presumption is that such outcomes are inefficient. It is clear that, according to American values as set out earlier, waste is to be avoided because it interferes with the pursuit of economic well-being. That is, the presence of waste means that individuals in the economy, as well as the economy as a whole, are unable to attain the level of preferred material well-being they could achieve were that waste eliminated. Thus, for example, fostering the spread of competition as close to its purest form as possible is desirable because, in the absence of off-setting circumstances (to be considered later), it tends to push the economy closer to the ideal and, hence, tends to eliminate waste. The second property relating to ideal economic systems is that any efficient allocation of goods and services can arise as the outcome of an ideal economic system by appropriately varying the resource endowments or, say, the wealth of distinct individuals or consumers in that system. Of course, wealthy people are likely to have greater incomes and thereby purchase greater quantities of goods,

74 An analytical vision of the workings of the United States economy and possibly different kinds of goods, than less wealthy people. But allocations of outcomes that derive from situations in which wealth is more or less equally distributed across individuals, as well as those that are based on some individuals being very wealthy while others are very poor, can all be efficient in the sense identified earlier. Each such allocation, given the distribution of income and wealth associated with it, provides the goods people want most, consistent with the consumption satisfaction achieved by others, and at minimum input cost. And each such allocation also appears as the outcome of an ideal economic system that is attended by the distribution of income and wealth on the basis of which that allocation can be achieved. That is, given each distribution of wealth, all buyers and sellers are able to buy and sell in the corresponding outcome exactly what they want at prevailing prices. Thus the ideal economic system is unbiased in that it is compatible with all efficient outcomes identified with the various distributions of wealth across individuals. It may be, of course, that society as a whole actually prefers some efficient outcomes to others. But, because it is unbiased, there is nothing in the ideal economic system that would, as a matter of course, lead that system to one of the more preferred outcomes. The outcome that in fact obtains depends on the distribution of resource endowments or of income and wealth across the population. Those persons with more income or wealth, and hence more money to spend, can buy more goods and services. Thus, contrary to what is believed in some quarters, leaving an ideal system to work things out on its own does not necessarily result in the best outcome from the perspective of society. Nor would it necessarily lead to more equality in the distribution of produced goods and services across individuals, or even to what society would regard as a fair distribution among those persons. Another way of saying this is that, due to the currently existing unequal distribution of income and wealth across American society, unbiasedness prevents the ideal economic system from guaranteeing as much as might be thought to be desirable of the equality of opportunity that is so central to the American value of justice. To some extent, Americans actually recognize this fact, even though the US economy is, in reality, some distance away from an ideal system. Government programs such as unemployment compensation and welfare, and the structuring of the income tax system to tax higher incomes at higher rates than lower incomes, are efforts to compensate for the unequal distribution of income and wealth in America and to secure a more socially preferred allocation of goods and services across individuals. A third property of ideal economic systems is unrelated to the nonwastefulness and unbiasedness described above. It arises at the level of the firm because under ideal conditions there are no barriers preventing buyers and sellers from attracting the requisite inputs (if needed), and entering and becoming participants in any market. As a consequence, if firms are not producing that volume of goods and services that yields minimum cost per unit of output (this is not the same thing as minimum input cost described earlier, which referred to the minimum cost of producing a given volume of output), then new production facilities will be created by other sellers who actively undercut the prices at which the original firms

An analytical vision of the workings of the United States economy 75 sell their outputs. Because products are identical within markets, in order to sell anything at all the latter firms will have to lower their prices and, eventually, their unit costs. This process eliminates a different kind of waste from ideal systems than that extracted by the first property described above. As such, it could be thought of as a further aspect of efficiency that is alternative to the notions of efficiency in the forms introduced earlier. To avoid confusion, however, the latter alternative is not referred to as efficiency here. Regardless, in addition to nonwastefulness and unbiasedness, the outcomes of ideal economic systems are all associated with production volumes at minimum unit cost. It has already been noted that, in fact, the US economy is only an approximation of an ideal system. To the extent that products are not identical within a single market, waste is created even in the presence of production at minimum input cost. Such waste arises because people are generally not obtaining what they want most, consistent with the consumption satisfaction achieved by others; nor are the firms in the market producing their outputs at minimum unit cost. But if the market still has a large number of small firms or sellers, this waste is more or less broadly scattered and relatively small. Of course, the existence of markets with differentiated products depends on the producers’ abilities both to design and invent competing products and to disseminate the relevant knowledge to consumers by advertising. With respect to the latter, when products within markets are not identical, the ability to advertise clearly enhances the variety of products that become available to consumers. Without that ability, since firms would have far more difficulty making the existence of new products known to potential buyers, it would become much more risky to produce them. Hence the number of products in any given market would be likely to be fewer. Furthermore, the availability of a wider variety of products within markets is a distinct advantage to American consumers because it expands the opportunities for consumption by furnishing many purchase options that would not otherwise be provided. Economic wellbeing is increased because more tastes, even if created, in part, by advertising, can be better served. Thus advertising, in spite of its costs, plays a major role in the economic well-being of Americans. When there are a small number of large firms in a market (with or without identical products), the distance from the ideal economic system is increased, and waste becomes greater, more concentrated, and more significant. Outcomes tend to be farther from what people want most, consistent with the consumption satisfaction achieved by others, and farther from minimum unit cost of production. The potential for abusing power also amplifies because so much economic power, arising from the great financial wealth these firms often amass and their considerable ability to influence market price, is centralized in the hands of so few. The departure from minimum unit cost, however, may be somewhat mitigated by the cost reductions that accrue from large-scale operations. In addition, advances in technology (for example, in computer chips), which are a major source of economic growth and improvement in the material well-being of US citizens, often occur in very large firms. That is because the revenues generated by large scales of operation frequently make it possible for firms to afford the research and

76 An analytical vision of the workings of the United States economy development (including the high capital equipment costs) necessary to invent and design new products. Thus the advantages and disadvantages of these more realistic but wasteful economic systems are decidedly mixed when evaluated with respect to American values. Even so, such departures from the ideal economic system are sometimes ignored by Americans when thinking about the US economy in broad, general terms. But the above list of departures from the ideal economic system does not tell the entire story. For there are further very specific and concrete departures that have yet to be identified. It is to the latter that attention now turns.

III In general, deviations of the American economy from the ideal economic system may be conveniently expressed with respect to the notion of market failure. A market is said to fail when there is a “malfunction” in it that precludes the outcomes that would emerge were competition in its purest form everywhere and the economic system ideal. It is the presence of market failure that, in the explanatory technical vision of this essay distances the real world from the ideal economic system. There are five main categories of market failure (the first has already been described above) and, in each case, remedies to correct the malfunction that caused it are suggested. Four categories point to malfunctions in individual markets and are as follows: (i) market competition that does not exhibit the first three characteristics of the purest form of competition listed earlier, namely, large numbers of small buyers and sellers, identical products, and the absence of barriers to entry and exit; (ii) externalities; (iii) public goods, and, in violation of the fourth characteristic of the purest form of competition; (iv) insufficient information. The final category is different from the above four in both scale and scope. For it refers to the frequent inability of the economy as a whole to provide full employment at stable prices, and may be viewed as a failure, in part, of the “economy-wide labor market.” Although little attention is given in what follows to the last of these categories of market failure, each is now considered in turn. First, as has already been suggested, waste results when competition in a market does not meet the standards of the first three characteristics, recalled immediately above, of that of the purest form. This waste derives from two sources. On the one hand, people are generally not obtaining what they want most, consistent with the consumption satisfaction achieved by others. On the other hand, production does not take place at minimum unit cost. Generally speaking, waste arises from these sources because, without the three characteristics of competition indicated (actually only the violation of either of the first two is all that is required to obtain the following conclusion), outputs will not be produced in quantities that equate the cost of producing the last unit of each good to its selling price. Indeed, it turns out that the selling price is always greater than that cost. Thus the benefits of each good are no longer balanced against the value of the opportunities of providing alternative goods that would have been produced with the same resources. As mentioned, this effect is a consequence of product variation within

An analytical vision of the workings of the United States economy 77 markets and/or the presence of small numbers of large firms, and it tends to be stronger the more difficult entry into the market and exit from it become. While the waste that emerges from product differentiation is not, as has also been previously suggested, considered to be a significant problem, that arising because firms are large and there are few of them is taken more seriously. There are two approaches to correcting the malfunction of small numbers of large firms. On the one hand, the behavior of firms in such a market can be regulated (for example, an upper limit can be placed on the price the firms are permitted to charge for their outputs) in order to bring the selling price more in line with the cost of the last unit produced. On the other, large firms can be broken up into smaller ones. The goal of each approach is the same, namely, to ensure that the outcome more closely approximates that of the ideal economic system. But the former approach, except for the effect it may have on the exit of firms from, or their entry into, the market, leaves the competitive structure of that market unaltered. Within this structure it attempts to force firms to produce goods and otherwise behave more like they would if competition in the market existed in its purest form. The latter approach constitutes an effort to transform the competitive structure itself, rendering the market more like one in which competition exists in its purest form, and therefore automatically yielding outcomes that are closer to the ideal. Attempts to correct the malfunction of small numbers of large firms in the US economy began towards the end of the nineteenth century and have persisted into the twenty-first century, although the energy put into these attempts has waxed and waned. And emphasis has also switched over time from one approach to the other. An externality, the second type of malfunction, is present in a market when a transaction in it involves a cost or benefit to an individual or a group that is not directly involved in the transaction itself. To illustrate, the purchase of electricity that has been produced by burning oil or coal results in the emission of chemicals and particles into the air that harms plant, animal, and human life elsewhere. Thus costs are imposed on people who have nothing to do with the production, sale, and purchase of that electricity. Alternatively, the restoration of the outside of a person’s house and the beautifying of his garden may please neighbors who formerly had to live next to what they regarded as an eyesore and a stain on their surroundings. Hence, although the neighbors did not pay for it, they still received benefits from the transactions that resulted in the restoration and beautification. Clearly, these extra, indirect costs and benefits of a transaction do not, like the direct benefits and costs to those participating in the transaction, flow through the market, and cannot, therefore, be subject to market discipline. That is, in the presence of an externality, market price no longer measures the value of all of the benefits that accrue to society from the production and sale of the good; and/or the cost of the last unit produced no longer encompasses all of the value of the foregone opportunities that are given up by its provision. The production of a good that is associated with an externality, therefore, necessarily involves waste. That waste is reflected in the fact that the quantity of the good (and its by-products) actually wanted by society is not provided by the economic system. In particular, it turns out that, as a rule, when competition exists in its purest form and when an

78 An analytical vision of the workings of the United States economy externality in the production of a good imposes additional outside costs, too much of the good will be produced relative to society’s desires. Were the externality to impose extra outside benefits, too little of the good would be produced. The remedies available to correct the malfunction that causes the waste of overproduction or underproduction in the presence of an externality include the use of taxes and subsidies, legal sanctions, and regulation. Alternatively, it is sometimes possible to introduce a new market in which the externality is “internalized” and brought under market control. For example, in the case of the production of electricity by burning oil or coal, suppose the possession of specific “rights to pollute,” or rights to emit certain amounts of chemicals and particles into the air, were required by law to actually release those pollutants into the environment, and suppose such rights could be bought and sold in a specially created market. Suppose also that each firm that produced electricity from oil or coal were initially issued a limited amount of rights to pollute. Then firms that had extra rights they did not need could sell them on the market to firms who did not have enough. In this way the cost of marginally acquired pollution rights would become a dollar cost that would be figured into the cost of producing the firm’s output. As long as the appropriate amounts of pollution rights were issued to firms at the outset, the latter production cost would include a portion of the actual cost of the externality. The balance between the value of the benefits of a good and the value of the opportunities foregone to have it would, in part, have been restored, and the waste associated with the externality would be reduced. The third category of market failure arises when there is a public good. A public good has two unique characteristics that distinguish it from ordinary commodities. First, when a unit of an ordinary commodity is consumed by an individual, that unit is no longer available for someone else to consume. In the case of a public good, however, one person’s consumption of any unit of it does not rule out or detract from the consumption of that unit by another, assuming that the latter is not prevented from consuming the good by crowds of people. Second, a public good cannot be produced for a single individual without making it available to everyone, free of charge. Common examples of public goods are street lighting and national defense. Of course, in order to be able to consume an ordinary good it is necessary to pay for it. But this is not so for a public good. That is, the consumer of a public good is not required to make payment for the specific units he consumes, although it is necessary that the actual cost of providing the good be covered, usually through taxation. Once street lights are turned on, say, all people on the street can use the light regardless of whether they pay for it or not. It follows that, since no one will pay for them, private firms will not attempt to produce and sell public goods. As a rule, then, without independent action, public goods will not be supplied by the private economic system. (One exception is wholly private radio broadcasting which, in most circumstances, is entirely paid for by the revenue from advertising.) Clearly, if society wants a public good to be available, then it has to make special arrangements for its provision. Such arrangements often take the form of federal, state, or local government contracting with a private firm for its production, subject to the relevant budgetary constraints and economic and social priorities.

An analytical vision of the workings of the United States economy 79 But the question remains of how much of the good society ought to have produced. In an ideal economic system, the correct amount of each good will, as has been described, automatically be provided, and waste will not exist. This is ensured, in part, by the forces of self-interest and profit maximization at the level of the individual consumer and firm, and by competition in the economy’s markets. But, with respect to a public good, a malfunction has occurred in that there is not even a market, let alone competition in any of its forms. The amount that ought to be produced, then, has to be set by other means. It might be thought that one way to determine how much of a public good should be produced, or even if the good should be provided at all, is to solicit information about what individuals are willing to pay for it. Willingnesses to pay could then be combined and balanced against the value of the alternative opportunities forgone by its production. But, from the perspective of the individual, each person may have significant incentives to conceal his actual feelings and provide incorrect information about what he is willing to pay. This is because, as long as the public good in question is produced, individuals can still enjoy the benefits of consuming it even if they say they are unwilling to pay anything for it. In addition, the production of a public good is usually so costly that its provision does not depend on whether one person pays for it or not. Hence, it is in the self-interest of each person, if he believes the good will be provided anyway, to claim publicly that he is unwilling to pay for it. The latter suggests that, unless individuals can be tricked into revealing their true willingnesses to pay, any decision to produce a public good in any quantity has to be based on different grounds. Regardless, an alternative method of solving this problem is by having the members of society vote either on the specific issues of whether and how much of the public good to provide, or for representatives who will be empowered to make the appropriate judgments and decisions. In either case, the tax burden of providing the public good needs to be considered. It should also be noted that there are other socially necessary goods that are often provided by the government that do not quite fall in the category of public goods because the private economy also supplies a portion of what is needed and because the government and the private firms producing those goods explicitly charge for them. A well-known example of government production under these conditions is the electricity generated and sold by the Tennessee Valley Authority. The fourth malfunction that results in market failure to be considered here is the situation in which buyers and/or sellers do not have sufficient information to fully evaluate the market commodity with respect to the price at which that good is bought and, respectively, sold. (That is, the fourth characteristic of competition in its purest form is breached.) The absence of information does not imply that transactions do not take place. Rather, it only means that the transactions that do take place will not necessarily live up to what the buyer and seller expect. To the extent that expectations are not met, outcomes involve waste because, had they known better, buyers and sellers would not have entered into those transactions in the first place. There are at least two ways in which insufficient information can be present. First, it is possible that either the buyer or seller has more information

80 An analytical vision of the workings of the United States economy about the good than the other. The buyer of old postage stamps may know that a particular stamp has much greater value because of a certain characteristic of which the seller is unaware. The seller of used cars may know that a particular car has serious mechanical problems of which the buyer is not aware. Were information sufficient in the former instance, the seller would insist on a higher price, and in the latter the buyer would not buy the car at the agreed-on price. But without that information, waste will occur because it will turn out that not everyone is obtaining what they want most, consistent with the consumption satisfaction achieved by others. Insufficient information is also present when a contract is entered into that (i) depends, at least in part, on the cost implications of the future behavior of one of the parties where that future behavior is generally unknown; and (ii) passes the cost of the future behavior on to the other party to the contract. For example, the cost to the seller of an automobile insurance policy rests partly on the behavior of the buyer after the contract is signed. The former would obviously have to cover the cost of such things as accidents and theft. And the latter might think that with his car fully insured, he does not have to concern himself with driving carefully or locking his car. In that case, waste is likely to appear because being unconcerned about such actions increases the likelihood of, respectively, accident and theft. And to the extent that accidents and thefts occur more frequently than general experience suggests, the insurance is not being produced at minimum input cost, where that cost now includes the cost of settling claims. With respect to this particular example, of course, the insured person may incur an indirect cost in the form of increased insurance premiums. The way to circumvent the waste that arises from insufficient information is to impose legal sanctions on those who would take advantage of the lack of information of others in both instances, and to write contracts that restrict the scope of the events that are covered under the terms of the contract in the second. The existence of legal standards for fair trade practices and to prevent insurance fraud illustrate the former; and fire insurance policies that require smoke detectors in covered buildings is an example of the latter. The last instance of market failure to be mentioned here is, as suggested earlier, related to the possible breakdown that occurs if reductions in the spending streams of the economy should lead to conditions in which some of the economy’s resources, including, notably, the resource labor, remain unemployed. There are a number of reasons why this can happen, and aggregate economic policies exist, such as fiscal and monetary policy, that the government can use to correct the difficulty. But the details are beyond the scope of the present essay. In all instances of market failure described here, it is the government that is looked to as the institution that is responsible for setting things right. In the American system, the general approach is to elect, by popular vote, representatives who are to make informed decisions on these matters for society at large. In many instances, they allocate tax funds for the purpose of hiring bureaucrats to carry out the details of the general laws and policies they enact. But it is important to recognize that just as malfunctions can cause markets to fail and create waste,

An analytical vision of the workings of the United States economy 81 so can the government, no matter how good its intentions, fail in its efforts to correct market failures. First, representatives usually want to be re-elected. Thus they may be willing to support something desired by a small number of people at the expense of society at large. And more generally, of course, they are always vulnerable to the lobbying of special interest groups. Bureaucrats, too, are often interested in preserving their jobs, and sometimes even in enhancing their power. They, like representatives, are also vulnerable to lobbying. Regardless, to have any government activity in any area at all, resources and the funds to purchase them have to be diverted from the private economy. To the extent that the government fails in its efforts to curtail waste in the private economy, waste of a kind that differs from that generated by market failure is introduced. The latter waste appears in the form of misused resources that could otherwise have provided additional goods and services for people to consume and improve their material well-being. Thus American society is always forced to choose between letting markets fail on their own and the less-than-perfect, sometimes wasteful, government efforts to rectify the difficulties. It is a choice that each generation of voters addresses and makes in its own way.

References 1. Katzner, D.W., Cherry Blossoms and Eagles: Some Comparisons between the United States and Japan (unpublished ms., 2004). 2. Shepherd, W.G., “Causes of Increased Competition in the U.S. Economy, 1939–1980,” Review of Economics and Statistics 64 (1982), pp. 613–626.

5

The workings of the Japanese economy1

As intimated in the previous essay, when explaining the way a particular economy works, economists are partly guided by those personal economic matters considered to be important by the people who participate in it. These concerns, of course, may vary from society to society or culture to culture. Thus, for example, the typical American cares about such things as his economic status and welfare, his job security, and the extent to which he is financially rewarded for his contributions to productive activity. By contrast, the personal economic concerns of a Japanese individual, which are, against their own criteria, as keenly felt as those of the American citizen, differ in numerous and significant respects. Economic status, although not having the same meaning in Japan as the US, is important, but only in relation to the social status that it confers. Economic well-being beyond that due to a given position in the social hierarchy is nice, but it is neither urgent nor expected. Job security matters, but in light of the loyalty between employer and employee (often expressed as an unwritten, life-time employment contract), most individuals have some measure of it. And the rewards bestowed by the economy are evaluated by the individual more in terms of the contribution he is able to make through his work to his company, to social harmony, and to society in general than with respect to income received.2 In all cases, explanations of economic activity, including both the models constructed to expound them as well as the visions that lie behind them, have to recognize these kinds of concerns if they are to be relevant to the real behavior that is being explained. And to the extent that such concerns modify across societies or cultures, the explanations, models, and visions to which they relate must, of necessity, reflect the relevant differences.

1 Crisis or Recovery in Japan: State and Industrial Economy, D. Bailey, D. Coffey, and P. Tomlinson, eds. (Chentenham: Elgar, 2007), pp. 9–30. © D. Bailey, D. Coffey, and P. Tomlinson. Reprinted in slightly modified form with permission. Thanks are due to Charles Yuji Horioka and Douglas Vickers for their help and guidance with this essay. 2 Of course, these statements of what is considered to be important by Americans and Japanese are generalizations that need not apply to the same degree (if at all) in every situation. They are still useful, however, in that they are relevant in many contexts and, as such, provide one way of understanding contrasts between the US and Japan.

The workings of the Japanese economy 83 In America, the economic concerns of the individual, framed with respect to American cultural norms and values, are often thought about and influence economic explanation in reference to a particular vision of Western capitalism. That vision, henceforth referred to as the vision of Western capitalism, expresses what may be called the (perfectly) competitive ideal. In this vision, frequently articulated by the so-called model of Walrasian or general equilibrium (informally set out in this volume as Essay 4), consumers and firms are motivated in their economic behavior by self-interest, and markets function freely and openly to determine, entirely through price competition, the prices of goods and services, and the quantities of them that are produced, bought, and sold. As a consequence, and continuing to confine attention to the competitive ideal, the prices that arise as outcomes of the economic process simultaneously reflect, in the case of produced goods, the values to society of the last units transacted and the opportunities forgone for those last units to be produced. Prices also serve more generally as signals that guide producers in the production of outputs and the hiring of inputs. Of course, it is quite clear that, in the operation of the actual American economy, divergences of varying degrees from the competitive ideal are observed, and with that, varying degrees of producers’ control over market selling prices arise. To the extent that such non-ideal competition among firms obtains and is accompanied by less than perfect competition in factor markets, the finely balanced marginal equivalences and optimization outcomes of the competitive ideal will not, in fact, be realized. Regardless, and in contrast to the perfectly competitive or Walrasian vision of the workings of the American economy, that is, the vision of Western capitalism as it applies to the American situation, there is no well-developed and generally accepted vision of the workings of the Japanese economy that explicitly addresses, from a uniquely Japanese perspective, the concerns of the Japanese citizen described above. To the extent that a vision of the objectives, structure, and functioning of that economy exists, it is substantially an imported vision, superficial and ill-suited to Japanese circumstances, of how a Western capitalist economy, like that of the US, operates. The latter vision, moreover, makes only a limited attempt to accommodate the unique features of Japanese economic and cultural life. Perhaps this is due to (i) the absence of an urgent necessity to justify and explain ideologically or otherwise the workings of a uniquely Japanese system; and (ii) the reality that many Japanese economists have been trained in Western traditions. But, in any case, the fact remains that although it is possible to speak of certain unique aspects of Japanese economic intercourse, there is, as yet, no comprehensive vision of the way in which that uniqueness coheres into a unified and interrelated economic system. The purpose of this essay is to explain why, and in what respects, the Japanese economy does not fit the imported mold derived from Western capitalism in general and the US economy in particular. Although, as has been acknowledged, the American economy does not exactly conform to that structure either (i.e. the vision purporting to explain the workings of the US economy provides an ideal picture that is only approximated by American economic reality), the Japanese economy

84 The workings of the Japanese economy differs from it to a far more significant degree. Indeed, the difference is so great that the use of the vision of the workings of the US economy to shed light on the workings of the Japanese economy must be called into serious question. Subsequent argument fits into a growing debate in the economics literature on the extent to which cultural divergence between the US and Japan is sufficiently strong and significant to warrant fundamentally different approaches to explaining economic behavior in the two societies. On the one hand, Fallows [6], Greif [7], Johnson [11], Katzner [12], [13], [14], and Prestowitz [18], to cite but a few, suggest or imply that it is; on the other, the so-called Chicago School as exemplified by Becker [2, especially p. 14], although not explicitly taking a position with respect to Japan, and writers like Koike [15], would not agree. The present essay falls clearly into the former camp. It begins (Section I) by describing the reasons why some of the basic culturally determined assumptions underlying the competitive ideal and the Walrasian model are irrelevant and inappropriate in explanations of the way in which the Japanese economy works. It then shows (Section II) how certain Japanese institutional arrangements (which could also be thought, in part, to be culturally derived), namely industrial structure, the labor market, the distribution system, and government participation in the private sector, drive the Japanese economy still farther, as compared with the US, from the competitive ideal. And, finally (Section III), it suggests that neither Western notions of efficiency expressed with respect to Pareto optimality and minimum unit cost, nor policy recommendations suitable for the US economy, need be apropos in Japan or even meaningful in the Japanese context.

I The Japanese economy, like the US economy, is made up of consumers, firms, and markets. Consumers purchase goods and services with income earned from the sale of labor to firms and from interest from past savings. The latter are mostly held in accounts at private banks and in an institution that, until recently, was part of the Post Office3 – not so much in the form of stocks and bonds. Firms hire inputs to produce outputs and obtain funds for expansion from both private banks and the government. Clearly the government is an important player here, and its role and impact on the economy is much more complex and extensive than that of the government in America. For, in addition to affecting the allocation of resources through purchases, transfer payments, the imposition of taxes, and monetary policy, the Japanese government, as has just been indicated, lends funds to private firms. Parallel to the US, the decisions of consumers and firms to buy and sell, supplemented, in the case of Japan, by government decisions to be discussed in greater detail below, provide the impetus for economic behavior.

3 In October of 2007, the Post Office was privatized and split into four parts. One of these was assigned the responsibility of collecting and handling consumer savings that would formerly have been deposited at the Post Office.

The workings of the Japanese economy 85 Now it is important to recognize, as has been argued elsewhere (Katzner [12], [14]), that in any given society the making of private economic decisions is based, to a considerable extent, on cultural norms and values. In America, the important norms and values are the sanctity of the individual, self-reliance, success, freedom, and justice, and these lead Americans, in part, to the pursuit of their individual self-interest. Moreover, one frequently employed method of measuring economic success in America (an important part of the value “success” listed above) is defined in terms of the material objects with which the individual is able to surround himself. As a rule, the bigger the house, the more opulent the furniture, the more expensive the car, the greater the evidence of success. This suggests that the larger an individual’s income, the more he will consume. It is also often thought that the greater the consumption, the greater the well-being (e.g. Bellah et al. [4, Chs. 2, 3]). But in Japan, the fundamental norms and values are different. They include service and loyalty to the collective, be it family, company, or society; the fulfillment of obligations to family, superiors, and others from whom favors have been received; the preservation of honor; the maintenance of harmony within groups and with nature; hard work; and frugality (Bellah [3], Benedict [5, Chs. 5–8], and Nakane [16, Ch. 2]).4 This list neither includes nor appears to result in the emergence of self-interest on the part of individuals. It may be supposed, then, that self-interest as it appears in the West, is not a significant element of Japanese culture and that it cannot, therefore, be an important motivating force behind Japanese economic behavior. The same can be said of the push to consume as a means of increasing well-being and of exhibiting the success that has been achieved. The absence of self-interest and the drive towards material consumption are two of the more significant features of the Japanese economic landscape that distinguish it from that of the US. What, then, propels individual consumer decision-making in Japan and leads to the purchases of goods and services, the supply of labor, and the act of saving? Although there is no generally accepted, Japanese-oriented answer to this question, several possibilities can be speculatively suggested. Whereas, in general, Americans (such as a cashier or an executive) work to further their interests and enhance their social statuses through the increased income generated by advancement in a career, the Japanese (here one might also cite a cashier or an executive) work to fulfill their obligations to themselves and those around them, and to serve the general community. In addition, decisions to supply various kinds of labor may have roots in the social status that different types of employment bequeath (more will be said of the latter below). The purchase of goods and

4 Like the statement of American and Japanese economic concerns at the outset, these norms and values (as well as those previously attributed to Americans) are generalizations that are not uniformly applicable to all Japanese (or, respectively, all Americans). But they are sufficiently significant to be useful in understanding, in an approximate way, motivations behind economic behavior in Japan (and, respectively, in America). Recall note 2 above.

86 The workings of the Japanese economy services may be determined largely by the kinds and quantities of them that are deemed appropriate to a person’s location in the social hierarchy. The place of purchase may relate to the presence of an already established long-term relationship with, and loyalty to, a particular supplier. And finally, the need to save may, to a considerable extent (though not entirely), be derived directly from the cultural norm of frugality mentioned above. Regardless, prices along with the amount of income the individual has to spend still define the available opportunities among which the individual must choose. It is worth digressing for a moment to suggest the complexity of the sources of these motivations. This can be done by exploring in some detail the need to save apart from the cultural norm of frugality. Culturally and emotionally the Japanese tend to be more withdrawn than Americans. This is so for at least two reasons. First, the search for harmony with society and nature by individuals in Japan is primarily an inward-looking quest. By that it is meant that the complex tangle of historical, religious, and philosophical influences prevalent in Japan, have created in the Japanese mind a disposition to adjust individual behavior in such a manner as to harmonize with the interests of relevant groups and society, and with the physical environment. Second, except for intimate family, the people with whom Japanese persons work closely, and small numbers of friends (and the latter two categories often overlap considerably), the Japanese are wary of interacting with other Japanese (Benedict [5, Ch. 8]). On the one hand, they do not want to increase the heavy burden of obligations that they and others already carry. On the other, they go to great lengths to avoid taking actions and making mistakes in relation to others that would bring possible dishonor both to themselves and to the others involved. Monetary saving, too, is an act of withdrawal. It conforms to, and is consistent with these other actions that lead to withdrawal. It is impetuses like these that combine with the general value of frugality in generating the motivation to save in Japan. In the American economy, firms are owned by individuals and institutional investors run by individuals who are driven by self-interest to seek the highest return on their investments consistent with the level of risk they are willing to bear. This translates into the pursuit of profit maximization on the part of firms with all of its implications for firm decisions in relation to the production of outputs, the hiring of inputs, and the borrowing of funds for continued operations and expansion. But in Japan, where firms are, by and large, primarily owned by institutional investors and other firms, and not by individuals, and where motivations other than self-interest are fundamental, the operation of firms in such a manner as to maximize profits is not a major concern. Thus, as a rule, firms do not necessarily mix inputs in production so as to minimize the production cost of each volume of output. They would not necessarily hire each kind of input up to the point at which the value to the firm of the last unit employed equals its cost (price, wage, or salary). And they would not necessarily produce a quantity of output such that the additional cost of producing the last unit of output is the same as its selling price. Rather, firms would tend to make production, input, and borrowing decisions in confluence with the Japanese cultural norms

The workings of the Japanese economy 87 and values listed above. Those decisions are made, that is, on the basis of how they best serve the community in general and the social interests of the firm and its employees. The firm’s decisions might also reflect an obligation to maintain appropriate relations with other firms with which it might be in some way affiliated, including the banks to whom it is indebted and from whom it might borrow funds in the future. It is clear that, Japanese cultural norms and values being what they are, the freedom from legal and institutional constraints required by Americans in the conduct of their economic activity is not needed or prized in Japan. Nor does the absence of equal opportunity of all persons in the making of choices violate the Japanese sense of justice as it does for an American. Moreover, the informational requirements of Japanese consumers and firms necessary to conduct their economic affairs are somewhat different from those of American consumers and firms, and the role that that information plays in the making of decisions is different. In Japan, of course, firms need to know the technical matters relating to the production of outputs from inputs and, as pointed out above with respect to consumers, both consumers and firms need to know prices to understand what opportunities exist. But unlike the situation that prevails in the US, the relationship between prices and what consumers want is murky in Japan because the dictates of a person’s social status may be more important for determining his economic purchases than what might otherwise be his personal preferences. Indeed, the pull between what a person might want and what he thinks he ought to have might lead to inconsistencies and instabilities in those purchases. Thus, in contrast to their role in the US economy, the functioning of Japanese prices as signals to firms indicating what they should produce is, to a significant degree, called into question. But even if Japanese prices could be used as signals to direct production and hiring decisions, it is unlikely that Japanese firms would always employ them for that purpose. That is because Japanese firms sometimes make investment decisions on the basis of (Nakane [16, p. 90]) “keeping up with the competition” or “doing what the competition does” in order to maintain their social status. (This is consistent with the overall goal described above of serving the social interests of the firm and its employees.) They also do not always rely on price information in relation to personnel decisions. Thus, for example, out of loyalty to them, employees are not necessarily fired when, at the margin, their value to the firm falls below their cost. As in America, the prices of goods and services are established in markets. But the competitive pressures of buyers forcing prices to rise when there are shortages and of sellers pushing prices down in the face of surpluses are not as strong. For even in retail markets, the producers of commodities retain considerable control over the prices at which their outputs are sold. And this is abetted by the loyalties of buyers to particular sellers even if it means paying higher prices for the same or similar products. There is intense competition among firms in Japan, but it does not always appear as price competition. Apart from price, Japanese firms may compete energetically with respect to the quality of the products they produce, the variety of products they make available, and the service in relation to

88 The workings of the Japanese economy those products they provide. And the most meaningful consequence of competition is the determination of the firm’s niche in the social hierarchy of firms – not profits (although that may be a secondary result). A high social rank is prized by both the firm and its employees. Indeed, the employee’s own social rank is determined by the rank he holds in the firm along with the rank the firm holds in the social order. It is social status, not monetary success or material surroundings, that is of primary importance to the Japanese. With respect to the individual, the economic status that provides and accompanies his social status, meaning by that the economic qualities of the job he holds, including the income and consumption opportunities that go with it, is of considerably less significance (Nakane [16, pp. 88–93]). In any case, the Japanese economy clearly answers fundamental economic questions – what will be produced, what inputs will be used in that production, and who will get what is produced – in a manner that diverges from answers provided by the US economy. In particular, it is neither necessary nor necessarily desirable in Japan that everyone is able to buy and sell exactly what he wants at prevailing market prices or that, in the production and allocation of commodities across the economy, the benefits of each good are balanced against the value of other opportunities that might have been procured from the same resources. It follows that, although the introduction of impediments to the free movement of prices in the US is thought to have serious consequences, additional blocking of the “free movement of prices” in Japan is of little concern since there is much less free movement of prices to begin with. It is implicit in the above discussion that the ideal (competitive) economic system in terms of which the actual US economy is analyzed and understood in Essay 4, and with respect to which, in that essay, its performance is measured, does not apply in Japan. The actual Japanese economy is so far from that ideal in so many ways that the employment of the ideal as a basis for evaluating the Japanese economy and understanding how it works is not very useful. This, of course, does not mean that scattered small bits and pieces of the ideal system might not be appropriate for explaining certain isolated fragments of Japanese economic life. For example, various versions of non-ideal competitive forms (e.g. monopolistic competition, oligopoly, and monopoly) that arise in practice in the US economy, and that contain some, but not all, of the characteristics of the ideal form, may have considerable resonance to actual Japanese economic experience. Those kinds of departures from the ideal competitive system, to the extent that they appear in Japan, are no different from those in the US. Rather than exploring them in greater detail, however, it will be more worthwhile to consider different deviations from the competitive ideal that do not appear in the American economy but which, at the same time, are significant features of the Japanese economic landscape. For it is deviations of this sort that tend to move the Japanese economy still farther (relative to the US position) from the competitive ideal than it already is. The specific deviations in relation to Japan discussed below arise with respect to industrial structure, the labor market, the distribution system, and the role of the government in economic affairs.

The workings of the Japanese economy 89

II Industrial structure. Consider the Japanese industrial structure first. Two important features of it are the focus of subsequent attention: the organization of firms into enterprise groups (called Keiretsu), and the distinction between “large” firms on the one hand, and “medium-size” and “small” firms on the other (e.g. Ito [10, Ch. 7]). There is substantial overlap between these categories in that one firm can belong to more than one enterprise group, and the same enterprise group can contain large firms as well as medium-size and small ones. An enterprise group is a collection of firms that are linked together through a network of mutually beneficial economic relationships. There are both horizontal and vertical enterprise groups. The horizontal groups are made up of companies in different major industries. The six largest groups combined employ about 5% of the Japanese labor force and earn about 15% of the Japanese economy’s sales and profits. A typical list of industries represented in them might include banking, insurance, construction, textiles, chemicals, mining, steel, electronics, automobiles, etc. Each group also contains at least one trading company that specializes in only exporting and importing. The same six horizontal groups share three common characteristics. (i) There is a collection of core group members that functions as a strategic decision-making unit for the entire group. The latter unit varies in importance from one enterprise group to another. (ii) Members of an enterprise group, including the banks, usually own shares of stock issued by other members. These shares tend not to be sold, and the amount of such shares held by group members can climb above 25% of total group shares issued. Such withingroup stock ownership is often very effective in preventing outsiders from gaining unwanted control of individual group members. (iii) Enterprise group members tend to raise significant quantities of the total funds that they borrow for business purposes from the banks and insurance companies in their respective groups. The percentage of borrowed funds in the group taken out from within-group banks and insurance companies can also be as high as 25%. (As with all firms, to the extent that a member’s use of borrowed funds, or debt capital, introduces a high debt-to-equity ratio in its financing structure or, that is, a high degree of financial leverage, a further element of risk is borne by its equity ownership.) The remaining horizontal enterprise groups are much smaller than the six under discussion here. The vertical enterprise groups are also smaller than the six largest horizontal groups and tend to be made up of a manufacturer together with its suppliers and distributors. As suggested above, a company that is a member of a horizontal enterprise group can also be in the core of an independent, vertical enterprise group. The advantage of membership in an enterprise group is that the member firms are obligated to render mutual assistance in many aspects of their operations. This is especially true during difficult economic periods. When sales are slow, manufacturers will help supplier affiliates by not reducing orders as much and as quickly as they might otherwise. (Nonaffiliate suppliers would lose a greater proportion of, or even all, orders, and lose them earlier.) Likewise, when shortages

90 The workings of the Japanese economy arise, suppliers would fill orders for manufacturing affiliates first. Banks and insurance companies would tend to save funds for, and be more generous in their loans to, members of the enterprise groups to which they belong. And so on. Thus the interdependencies among members of enterprise groups tend to moderate the variability of their economic performances. But, since showing favoritism in this way often results in reduced revenues or increased costs, members in enterprise groups tend to have smaller profits and grow at slower rates than they otherwise would. This, of course, is not a significant issue in Japan because of the culturally derived lack of emphasis on the pursuit of individual self-interest and profit maximization. Turning to the size distinction among Japanese firms, the large corporations differ qualitatively in both organization and operation from the small and mediumsize companies. And this difference goes beyond what one would expect from the above discussion of enterprise groups and what one might observe with respect to size disparities in the US. In Japan, large firms tend to hire male employees upon graduation from school and to keep them until their retirement. They rarely hire persons in mid-career and typically pay higher wages than those paid by the smaller firms. Although there are exceptions, employment in medium-size and small firms is less permanent and more often temporary and seasonal, and wages are usually lower. Some medium-size and small firms supply parts to large corporations and might have some enterprise-group affiliation with them. Others are in the service sector (e.g. retail shops and restaurants), and are family-owned and -operated. To the extent that these differences between large firms and medium-size and small firms exist, one of their consequences is a parallel division of the Japanese labor force into workers with “lifetime” employment in large companies at relatively high wages, and workers with less permanent employment in mediumsize and small companies at relatively low wages. Taking into account the obligations among members of enterprise groups described above, it is still generally true that when an economic slow-down occurs, large companies order less from their suppliers, though orders from suppliers outside their enterprise groups tend to be reduced more and affected first. In any case, with orders decreased, the suppliers lay off some of their temporary workers. Thus, in hard times, the work force of the small and medium-size firms acts as a buffer protecting the jobs of the workers in the large corporations. The employees in those large corporations may be re-assigned from, say, white- to blue-collar positions, but the pressure to terminate them is less than it would be if the large corporations produced their own supplies, and the threat to the lifetime employment system is thereby lessened. One more point, although a slight digression, deserves brief mention here. Boards of directors of American corporations generally include individuals appointed from communities outside of the firms they are directing. Even though those boards may be controlled by their chief executives, their purpose is, largely, to represent stockholders’ interests. This has tended to mean that corporate activity has focused more on the enhancement of short-term earnings, and that significant portions of corporate profits have been returned to stockholders as dividends

The workings of the Japanese economy 91 (although it remains true that substantial retention of earnings is a principal source of financing expansion, and short-run profits are sometimes sacrificed for long-run gains). It is primarily through boards of directors, then, that the American pursuit of self-interest is translated into profit maximization in US corporations. By contrast, the members of boards of directors of Japanese corporations are usually drawn from inside their corporate entities and, if they are members, from inside their enterprise groups. As a result, these boards frequently represent interests other than those of the nongroup stockholders and thus, in particular, Japanese corporations typically are managed more for their employees than for such stockholders. Dividends are relatively small and the outlook is decidedly longer run. In contrast to the appearance of American cultural values and norms such as self-interest in the running of US corporations, this is a reflection of the Japanese values of loyalty and service to the community, firm, and employee, and of Japan’s cultural indifference to self-interest and the maximization of profits. The labor market. Additional departures from the competitive ideal that are unique to Japanese experience also arise with respect to the labor market in Japan (e.g. Abegglen and Stalk [1, pp. 198–208] and Ito [10, Ch. 8]). Begin with the lifetime employment system which has been briefly mentioned earlier. Roughly 25% of the Japanese labor force (mostly males employed by the large corporations) have an unwritten lifetime contract with their employer. According to that contract, the company is to provide employment until retirement and the employee is to be totally loyal to the company. The latter means, in part, that the employee will do virtually any job he is asked to do. Employees are hired directly from school (as previously indicated) and, over the years, are rotated in and out of many different company jobs. Each starts at the bottom of the corporate hierarchy and progresses through higher and higher levels. If a particular skill is needed that is not available in the company, new people are not brought in. Rather, the appropriate portion of people already working for the company are either sent to school or otherwise trained. Upon retirement, an employee will often be found a position in one of the company’s subsidiaries or affiliates for a few more years. Even in smaller firms that do not provide “formal” lifetime employment and where workers are typically considered “temporary,” employees frequently exhibit the same kind of obligation and loyalty to the company. And the employers, equally obliged and loyal in return, are often very reluctant to let them go – sometimes to the point of threatening the firm with, and even resulting in its bankruptcy. Generally speaking, Japanese employees work longer hours than US employees (Horioka [9, p. 6]). Those covered by the lifetime employment system typically leave home early Monday through Friday mornings and do not return until 10:00 or 11:00 at night. Working Saturday mornings in addition is not uncommon. Sometimes work is combined with social activity at restaurants and bars during evening hours. Employees under lifetime contract often take only half of their paid vacations, mainly for family events like weddings and funerals. It is not that Americans do not work hard, but as a rule, they do not put in as many hours as the Japanese, and when they do work hard, they do so for different reasons. While the American pursuit of self-interest and material consumption frequently results in

92 The workings of the Japanese economy hard work, the Japanese work hard because, without any intervening motivations like self-interest, hard work is one of their chief cultural values. In large Japanese companies, especially those providing lifetime employment, there is a strong and pervasive feeling that everyone is part of the company “family.” Close emotional ties develop between supervisor and subordinate, between those who usually work together, and between within-company friends. Along with the long hours of work are the sharing of highly personal thoughts and experiences, socializing at bars and restaurants, company-sponsored social events, and company support at such private affairs as weddings and funerals. It is not surprising that in large companies the dividing line between employees’ private lives and work lives tends to blur. Unlike US firms, there are relatively few distinctions between white- and bluecollar workers in Japan. The gap in salaries is small and both belong to the same union and eat in the same cafeteria. Employees who are considered incompetent or who are no longer useful to the large companies are not given much responsibility. They are also not fired. Sometimes they are assigned a desk next to a window with a nice view and nothing to do. Such employees are called “window watchers” because they only sit and look out the window. However, the shame of being reduced to a window watcher is considerable. Many will quit the company within a few years of becoming one. Japanese companies of all sizes typically consider women workers as either temporary until marriage or after-having-had-and-raised-children returnees. In neither case are they given lifetime employment contracts, and they usually work as secretaries, sales persons, and production-line laborers.5 Women who are looking for more permanent employment with opportunities to advance and develop a career will frequently seek out jobs in foreign companies with offices in Japan. Such companies make no lifetime employment guarantees and tend to lie outside the social hierarchy described above, thereby leaving their employees without social rank. That would create major problems for Japanese men who, as a result, try to avoid jobs in foreign companies. But these “failures” are of little consequence for Japanese women who are unable to procure a lifetime employment contract anywhere anyway, and who do not need a social rank from their employment since, upon marriage, their social ranks become those of their husbands. Indeed, it is precisely because foreign companies are not constrained by the same cultural rules that restrict the operation of Japanese companies in general that they are able to offer opportunities to women that are not available elsewhere. Although there are exceptions, promotions in Japanese firms normally turn on seniority. Salaries and wages depend on seniority and need. The typical Japanese employee (including executives) receives, in a normal year, approximately onethird of his pay in the form of semi-annual bonuses. The smaller bonus comes in

5 Since 1985 it has been illegal in Japan to discriminate against women in employment, and small numbers of women have been hired by large firms on the same terms as elite-track males with permanent employment status.

The workings of the Japanese economy 93 June or July; the larger one in December. The amount of the bonus hangs on the profitability of the firm during the year. It tends to be larger for the employees of the larger firms and for people with more seniority. Under adverse economic conditions when it becomes necessary to cut costs, bonuses can be reduced or eliminated. This is a second mechanism, in addition to the buffer provided by the temporary employees of medium-size and small firms, that helps to lower the pressure to break lifetime employment contracts when the economy enters recession. It should also be noted that the salaries of executives of American firms are often far larger than those of comparable executives in Japanese companies – especially when all forms of compensation are taken into account. For example, the average compensation of the chief executive officers of 30 major corporations in the US were recently found to be 212 times that of the average American worker while, in Japan, the comparable number was 16 (Oshima [17, p. 198]). This discrepancy in US firms, though difficult to explain from a Japanese perspective, where the pursuit of self-interest and material consumption are decidedly less important in the general scheme of things, is easily understood in the context of American cultural norms and values. For, according to the vision of Essay 4 of the way in which the US economy works, individuals are paid according to their worth to the firm in relation to what the market will bear. Any interference with the operation of markets is, as a rule, not tolerated because it obstructs the preservation of equal opportunity for all and, consequently, offends the American sense of justice, and also because it destroys the unfettered ability of the individual to pursue his own self-interest as best he can. Another significant way in which the Japanese labor market diverges from that of the US is in reference to labor unions. Unions exist in Japan because, in spite of the obligations and loyalties between employer and employee, collective bargaining that engages labor and management is still a practical and effective way of reaching agreement on compensation, working conditions, and other matters of mutual concern. In contrast to America, national unions in Japan are usually based on industry type (such as metals) rather than on specific skills (for example, truck drivers). There are similarities too. National unions band together in large organizations like America’s AFL-CIO. Labor-management negotiations take place at both national and local levels. But in Japan, local agreements can vary widely across firms within an industry, often depending on the overall performance of the firm in question. Negotiated contracts are typically for a single year. Generally, negotiations begin in February and end in May. (Of course, in the US, contracts are usually for more than one year. Negotiations can start at any time, depending on the expiration date of the old contract, and frequently last longer than 3 or 4 months.) But the main difference between US and Japanese unions occurs with respect to the constitution of the local unions. A single local union in Japan consists of, as a rule, most of the employees of a single firm – that is, all those at the rank of subsection head and below – and no one else. It includes both white- and bluecollar workers: lower-level executives, secretaries, managers, sales personnel,

94 The workings of the Japanese economy production-line workers, etc. Perhaps because of the tendency of the Japanese to view their large corporations as families of employees, perhaps because of the loyalties between employer and employee, and perhaps because the latter derives his social status from being employed by his employer, local unions are more reluctant than their US counterparts to engage in activities that might damage the firm. It is a common perception within the local unions that hurting the firm hurts themselves. Hence negotiations are not usually as adversarial, or at least not in the same way, as in the US. The distribution system. The Japanese system for distributing produced goods to consumers through wholesalers and retailers has a number of distinctive features, some of which push the Japanese economy even further from the competitive ideal (e.g. Ito [10, Ch. 13]). Observe, first, that Japan has many more retailers per capita than the US. This may be due, in part, to government regulations that limit the size, number, and operation of large retail establishments.6 But it also has something to do with the fact that, since they have very limited storage space at home, the Japanese tend to shop much more frequently than Americans. With roads choked by automobiles and not much parking space available, small neighborhood stores to which one might walk are a great convenience. Even the use of buses and trains to travel to shopping areas has its drawbacks since it is more time-consuming than a walk to a neighborhood store, and a person can still carry only so much home. In any case, retail establishments tend to be small; on average, each is operated by only four persons. And over half function with only one or two persons. Many are family-owned and -operated businesses. Japanese wholesalers are a little larger. But with so many retailers, large numbers and many layers of wholesalers are needed. To illustrate, suppose, for the sake of argument, that one wholesaler can distribute goods to five retailers. If there are more than 25 retailers, more than five wholesalers are needed, and a wholesaler on a level above the first five wholesalers is required to distribute goods to them. As the number of retailers grows, the number of levels of wholesalers grows along with the number of wholesalers themselves. Thus the Japanese distribution system as a whole involves relatively more individuals than that in the US and is relatively more cumbersome. In addition to the larger proportion of resources devoted to distribution in Japan, many manufacturers have developed their own wholesaling and retailing systems to distribute their products to consumers. The wholesalers and retailers in these private systems carry only limited brands of goods as prescribed by the manufacturer. One variant of this arrangement is that large retail stores are frequently staffed with persons on loan from the manufacturers. These people sell and otherwise promote their own manufacturer’s products at the retail level. Although private distribution systems with characteristics similar to these exist in the US (e.g. in the American automobile industry and in supermarkets), they are

6 These regulations were relaxed somewhat in 1992 to make it easier for large-scale retail concerns to open.

The workings of the Japanese economy 95 much more common in Japan. But regardless of whether manufacturers have their own distribution system, it is generally the case that retail prices are maintained by implicit agreements between them and the retailers who handle their goods. Wholesalers are often the conduits for these agreements and matters relating to them. The agreements themselves necessarily include provisions for income for wholesalers at all wholesaler levels, and this suggests, due to the relatively large number of wholesaler levels in Japan, that the prices of these products may differ still further from those obtained under conditions of the competitive ideal referred to earlier. Moreover, retailers are permitted to return unsold merchandise even if those items were actually purchased from the manufacturers. All of this points towards rather close relationships between manufacturers, wholesalers, and retailers. Indeed, those relationships are essential to the Japanese distribution system. They foster long-term connections and deep obligations and loyalties that manufacturers enhance with personal socializing, reliable delivery records, and the provision of consistent quality and after-sale service. Such obligations and loyalties, recall, are important in Japan from a cultural standpoint. As has been mentioned earlier, consumers also like to establish long-term relationships with retailers, and will often continue to buy from those retailers even if prices are lower elsewhere. Do the characteristics of the Japanese distribution system described above effectively close off consumer markets in Japan to foreign competition? The answer is both yes and no. On the one hand, the obstacles standing in the way of any newcomer, and especially of foreigners with little understanding of Japan, trying to break into that system are clear. Part of the problem arises from government regulations that, although previously alluded to, have not been considered in detail here. But attempting to enter a system in which strong long-term relationships and loyalties have already been established is even more formidable. On the other hand, a number of American companies (e.g. Motorola and Microsoft) have been quite successful in entering Japanese markets. Regulations can be, and have been, reduced and eliminated through negotiation. But it is more important to have a product that the Japanese consumer wants and is willing to buy, and that the time and effort needed to establish a distribution arrangement with the appropriate relationships among the individuals involved are expended. Government participation. All governments engage in activities that have impacts on the economies in relation to which they function. They impose taxes and spend the funds taxes raise. They create armies. They build and enhance roads, ports, and dams. They regulate, police, and prosecute. Government in the US and Japan is no exception. In the US, one of the goals of government that emerges directly from the American cultural values and norms described earlier is to interfere as little as possible in private economic operations, and maintain and enhance the free and competitive functioning of markets. But in Japan, where cultural values and norms favor the collective above the individual, it is seen as the obligation of government to play an active part in private economic development. This means, among other things, that government activities will often lead the Japanese economy farther from the competitive ideal.

96 The workings of the Japanese economy To be specific, the Japanese government energetically helps industries and companies that have what it considers to be bright futures (e.g. Johnson [11] and Ito [10, Ch. 7]). Until they are thought to be able to stand on their own, it protects them from foreign competition with tariffs and quotas on competing goods imported from abroad. (This is also one source of the regulations mentioned above with respect to the Japanese distribution system.) The government further provides these companies with preferential (i.e. low cost) loans and access to other resources, such as foreign exchange, that they might require. During adverse economic times it helps industries it favors to reorganize to reduce competition within and among them and to lower costs and excess capacity. It has also subsidized firms in financial difficulty so that they do not have to release employees. (This has the effect of providing one more mechanism for the preservation of the lifetime employment system.) Even declining industries like mining and aluminum have been helped in order to slow their demise. All of this would be intolerable in the US, and some of it patently illegal. And by not permitting markets to operate independently, it forces the outcomes of economic activity to be more distant from those that would be delivered by the competitive ideal. In the past, the Japanese government has had considerable success in pursuing these policies. It was, to a large extent, responsible, after the devastation of World War II, for rebuilding the Japanese economy into the world’s second largest economic machine. But when it chooses the wrong projects to support (as in the case of the Japanese approach to high definitional TV), or makes other errors (as when it ignored the role of Japan’s banks in the speculative bubble in Japanese land prices in the late 1980s), the costs to Japan can be very high.

III The contrasts between the United States and Japanese economies that emerge from the descriptive presentation of the last section are rather striking. They become even more pronounced when certain specific characteristics of the American economic environment are brought clearly into view. First, as previously indicated, in the US the formation of enterprise groups is severely limited by anti-trust laws and the accompanying bureaucratic regulation. The activities of such groups would often be perceived as improperly, and sometimes illegally, reducing competition and restraining trade. Second, the qualitative distinction between large firms and medium-size and small firms in Japan is inapplicable in America. Of course size makes a difference in the United States, but not, as has been suggested in Essay 4, in the same kinds of ways. Third, although Americans can and do spend their entire careers in a single firm, there is no lifetime employment system in US companies with an unwritten commitment between employer and employee. Fourth, as a rule, American workers are not rotated around from job to job within firms. Rather than learning how to do many things as their Japanese counterparts do, they tend to become highly specialized in particular areas. Fifth, a US employee who has acquired marketable skills in one firm, will often leave (and is often expected to leave) to improve his status and salary. Loyalty to the firm that taught him his skills,

The workings of the Japanese economy 97 to the extent that it might exist, does not override the pursuit of his self-interest. Sixth, American women are able (and often encouraged) to have significant careers in firms of all sizes. Seventh, bonuses in the United States are usually limited to a company’s top executives. Eighth, there are significant economic and social differences between white- and blue-collar workers in America. Ninth, the American distribution system is more open and easier for foreigners to break into than that in Japan. In particular, the close relationships, obligations, and loyalties between manufacturers, wholesalers, and retailers in Japan generally do not exist in the US, and some activities that those relationships, obligations, and loyalties might produce are prevented by law. Tenth, and lastly, it should be emphasized that government in America, as has already been pointed out, is focused more on staying out of private economic affairs then on helping to manage them. Indeed, the responsibility of the American government is seen to be centered primarily on the creation of an environment that facilitates the successful pursuit of independent economic activity. In addition to these contrasts, previous argument has also suggested that the workings of the Japanese economy cannot really be understood in terms of the Western capitalist vision, as that was described at the outset, that is often invoked in thinking about the US economy. The reasons for this may be summarized as follows. The motivations indicated earlier that direct the economic behavior of consumers and firms in Japan are different from the self-interest that propels action in the Western capitalist vision. Price competition in Japanese markets is not the most important form of competition among Japanese firms, and it is insufficient to ensure that the prices established in those markets convey the information necessary to result in an outcome consistent with that of the vision of Western capitalism. Enterprise groups, the qualitative differences between large firms and medium-size and small firms, the lifetime employment system, the determination of pay and promotion on the basis of seniority, the large and multilayered distribution system with its close ties among manufacturers, wholesalers, and retailers, and the active participation of government in guiding the private allocation of resources in directions towards which markets do not necessarily point, all introduce factors and forces that are not present in the Western capitalist vision and that are sufficiently significant to require that they be addressed in explaining the workings of the Japanese economy. The Western capitalist vision is therefore not relevant as a standard for understanding and making judgments about that economy. It is true, of course, as was acknowledged at the beginning, that the very stringent conditions of the competitive ideal are not completely fulfilled in the US economy either. But previous discussion has shown that the actual economic circumstances in Japan are much farther from the conditions of the competitive ideal than those in the US. Another way of saying this is that market failure or divergence from the competitive ideal, as defined in the context of the vision of Western capitalism, is so endemic in Japan that it renders that vision useless as a basis for thinking about the Japanese economy. It follows that notions of efficiency characterized with respect to the Western capitalist vision have no applicability with respect to Japan. There are two such

98 The workings of the Japanese economy efficiency concepts that are worthy of mention here. (Both were described in greater detail earlier in this volume in Essay 4.) On the one hand, the outcomes of the ideal competitive system produce what people want most (consistent with the consumption satisfaction achieved by others) at minimum input cost. This is equivalent to saying that, in any one of those outcomes, it is not possible through reallocation of resources to make one person feel he is better off without making someone else feel that he is worse off. On the other hand, the outcomes of the ideal system lead to production everywhere at minimum unit cost. By either of these standards, the outcome of Japanese economic activity must necessarily appear, except coincidentally, to be inefficient. This is because many of the conditions that lead to efficiency in the Western capitalist vision are not satisfied in Japan, and many of the conditions that actually do prevail in Japan would, in that same vision, prevent efficiency from being achieved. But even so, there can be little doubt that the Japanese economy has achieved considerable success. It has already been pointed out that that economy rose from the ashes of World War II to become the second most powerful in the world. And it is both interesting and ironic that, in spite of the inefficiencies that are, and have been present by the standards of the vision of Western capitalism, during the 1980s many in America and the West regarded the Japanese economy as the most efficient on earth. To provide specific illustrations of how the inefficiencies that arise according to American or Western standards are “overcome” by the Japanese economy, consider the fact that by criteria derived from the vision of Western capitalism, Japanese companies, especially during recessionary times, would, due partly to lifetime employment contracts and loyalties between employer and employee, appear to have too many employees. This possibility, to the extent that it occurs during recessions, is mitigated in Japan, as previously suggested, through government subsidies, through the bonus system, and by the willingness of companies to accept less than maximum profit.7 With respect to the last mitigating effect, too many employees in the Western scheme of things would mean smaller profits. And, with respect to the second, if profits become too small, or the company were threatened by losses, one Japanese solution could be to reduce, if not eliminate bonuses. An offsetting effect to the general overhiring of workers arises from the long hours of effort that, as previously indicated, are contributed by a Japanese person to his company. Many of these hours are overtime hours for which the employee is not compensated. Indeed, it has been estimated by Higa and Lupardus [8, p. 92] that, in 1994, the typical Japanese male employee was paid for only 82% of the hours he actually worked. The willingness of individuals to give these extra hours to their employer could be attributed to the Japanese cultural value of hard work, to a sense of loyalty and obligation to their employers, or to both. But in any case, the underpaying of employees, or the unpaid overtime put in by them

7 The argument in relation to the latter two is more fully developed in Katzner [13] reprinted here as Essay 6.

The workings of the Japanese economy 99 clearly counterbalances, at least to some extent, the overhiring of labor by Japanese companies. Another counterbalancing effect is, of course, the readiness mentioned above of companies to be content with less than maximum profit. A further example relates to the possibly excessive (again, when compared to the US) quantities of resources used by Japan to distribute goods to consumers. Such practices would tend to increase costs and lower profits which, as in the previous examples, would not be considered to be a problem by the Japanese. (The need, in Japan, for large numbers of retailers and the wholesalers who support them has already been indicated.) Moreover, application of the vision of Western capitalism suggests that, if it were possible to reduce the resources devoted to distribution (perhaps by removing the regulations that restrict the development of large-scale retail outlets), then those saved resources could be used to produce more goods and services for everyone. But the Japanese do not necessarily desire, and are not even necessarily concerned about such increases in material consumption. It is the harmonious functioning of, and the benefits they perceive accruing to the collective or society at large that are most important to them. Regardless, the “inefficiencies” of the distribution system are overcome in some measure by higher prices charged for consumer goods produced and sold in Japan and, once again, the willingness of Japanese companies to accept less than maximum profit. Finally, it should be noted that policy prescriptions that might be relevant in America to mitigate the effects of such economic ills as unemployment and inflation, need not necessarily work in Japan. Since these prescriptions tend to spring from the vision of Western capitalism with respect to which the American economy is understood, their best chances of having the effects desired are in the United States (or any other Western) economy. But it has been argued here that the Japanese economy does not work in the same way as the American economy. Therefore, implementing the same policies to achieve the same purposes in Japan is quite problematic. In order to judge the efficacy of American policy prescriptions applied to the Japanese situation, it would be necessary to have a vision of the workings of the Japanese economy. Without that, there is no way of knowing if a cut in taxes in Japan, say, might stimulate spending, or raising interest rates there might lower investment. Clearly, the application of American policy prescriptions to resolve Japanese economic problems under those conditions is like shooting in the dark without a clear perception of the target.

References 1. Abegglen, J.C., and G. Stalk, jr., Kaisha, The Japanese Corporation (Tokyo: Tuttle, 1985). 2. Becker, G.S., The Economic Approach to Human Behavior (Chicago: University of Chicago Press, 1976). 3. Bellah, R.N., Tokugawa Religion (Glencoe: Free Press, 1957). 4. Bellah, R.N., R. Madsen, W.M. Sullivan, A. Swidler, and S.M. Tipton, Habits of the Heart (Berkeley: University of California Press, 1985). 5. Benedict, R., The Chrysanthemum and the Sword (Boston: Houghton Mifflin, 1946).

100 The workings of the Japanese economy 6. Fallows, J., Looking at the Sun (New York: Pantheon, 1994). 7. Greif, A., “Cultural Beliefs and the Organization of Society: A Historical and Theoretical Reflection on Collectivist and Individualist Societies,” Journal of Political Economy 102 (1994), pp. 912–950. 8. Higa, T. and K. Lupardus, “Why Don’t Workers Claim All Their Overtime?” Japan: Why it Works, Why It Doesn’t – Economics in Everyday Life, J. Mak, S. Sunder, S. Abe, and K. Igawa, eds. (Honolulu: University of Hawaii Press, 1998), pp. 91–97. 9. Horioka, C.Y., “Do the Japanese Live Better Than Americans?” Japan: Why It Works, Why It Doesn’t – Economics in Everyday Life, J. Mak, S. Sunder, S. Abe, and K. Igawa, eds. (Honolulu: University of Hawaii Press, 1998), pp. 3–10. 10. Ito, T., The Japanese Economy (Cambridge: MIT Press, 1992). 11. Johnson, C., MITI and the Japanese Miracle (Stanford: Stanford University Press, 1982). 12. Katzner, D.W., “Western Economics and the Economy of Japan,” Journal of Post Keynesian Economics 21 (1999), pp. 503–521. This volume: Essay 2. 13. ——, “Explaining the Japanese Economic Miracle,” Japan and the World Economy 13 (2001), pp. 303–319. This volume: Essay 6. 14. ——, “‘What Are the Questions?’” Journal of Post Keynesian Economics 25 (2002), pp. 51–68. This volume: Essay 3. 15. Koike, K., “Skill Formation Systems in the U.S. and Japan: A Comparative Study,” The Economic Analysis of the Japanese Firm, M. Aoki, ed. (Amsterdam: North-Holland, 1984), pp. 47–75. 16. Nakane, C., Japanese Society (Berkeley: University of California Press, 1970). 17. Oshima, H.T., “Is Japan an Egalitarian Society?” Japan: Why It Works, Why It Doesn’t – Economics in Everyday Life, J. Mak, S. Sunder, S. Abe, and K. Igawa, eds. (Honolulu: University of Hawaii Press, 1998), pp. 195–204. 18. Prestowitz, C.V., jr., Trading Places (New York: Basic Books, 1988).

6

Explaining the Japanese economic miracle1

The performance of the Japanese economy since the Meiji restoration in 1868 has been quite extraordinary. Japan emerged from the intense isolation imposed by the Tokugawa as an economic backwater, lagging far behind the West. But by the end of the twentieth century, her economy had become the second most powerful in the world. And this was accomplished in spite of the vast devastation that she suffered as a consequence of her defeat in World War II. Although the sustained growth of the Japanese economy from 1868 to 1989 was remarkable by itself, the 18-year period between 1955 and 1973 was especially noteworthy. For over the latter span the average annual growth rate of the Japanese economy measured in terms of real gross national product (GNP) rose to almost 10%. The singular nature of this achievement is captured in the phrase “Japanese economic miracle” which has come to be used to describe the phenomenon. As a consequence of the successful development of the Japanese economy in general, and the miracle in particular, Western economies began to take notice of Japan. The view that there was much to be learned by studying the Japanese approach became widespread. Western companies and governments began to emulate components (e.g. the kanban or just-in-time inventory system, worker participation in directing production, and, in some countries, governmental bureaucratic direction of the allocation of resources) that appeared to be significant in contributing to the Japanese accomplishment. By the 1980s, interest in Japan had intensified considerably and even turned, in some quarters, to worry. Japan, it seemed, had redefined the notions of competition, the organization of production, and economic progress and development. Moreover, “Japan Inc.” had become so efficient that it was taking over the world, and the only way to compete with the Japanese was to join the bandwagon and adopt their methods. Then, in 1990, the speculative bubble in Japanese land prices burst and the Japanese economy began its decade-long slide through lower growth rates into recession. Suddenly the tables were turned. Instead of being admired and emulated

1 Japan and the World Economy, 13 (2001), pp. 303–319. © Elsevier Science B.V. Reprinted in modified and corrected form with permission. Thanks are due to Charles Yuji Horioka and Douglas Vickers for their considerable help.

102 Explaining the Japanese economic miracle everywhere, the Japanese came to be seen as bumblers that threatened global prosperity. Companies formerly thought of as the most efficient in the world were now perceived of as having costs that were much too high. The government bureaucracy that previously had been viewed as one of the main engines driving Japanese economic success since the second world war was now regarded as inept and rudderless. The regulatory structures that earlier were given credit for fostering large-scale industry growth, such as the protection of “infant” industries from foreign competition and the provision of low-cost loans and foreign exchange to certain favored companies, became a tangle of barriers to competition and another of the primary reasons for Japan’s economic troubles. Now perceptions, of course, no matter how widespread, do not always reflect reality. Still, it is possible to claim that the popular perceptions of the Japanese economy before and after 1990 as previously described are both accurate, and that the change in perception around 1990 came about because the characteristics of the real economic world had modified. What worked before 1990 did not work afterwards. The Japanese economic system, while uniquely suited to spur rapid economic growth during the miracle period, did not mesh with the changed economic realities that had emerged and become established by the last decade of the twentieth century.2 It is also possible to assert, however, and this is the perspective taken in the present essay, that not only is the above perception of the Japanese economy since 1990 accurate for the decade of the 1990s, but that a similar perception actually reflects Japanese economic realities back through the miracle period and possibly even earlier. That is, the recent perception of Japanese inefficiency (as that concept is defined in Western economics) is closer to the mark for the entire post-war period than the previous perception that Japan (before 1990) was leading the way to a new kind of economic system that, by dint of its enhanced efficiency and hence competitive edge, would dominate the world during the twenty-first century. Moreover, the development of the Japanese economy since the second world war, and especially the Japanese economic miracle, are even more striking because they occurred in the face of rather significant inefficiencies that would ordinarily have caused Western economies to stumble badly and stagnate. If this view is correct, it raises many questions, including the following. If Japanese firms were so inefficient, how were they able to sell their products for almost 40 years at prices comparable to, and sometimes substantially lower than, those of considerably more efficient Western firms and not be driven out of world markets by the competition? And even if they survived, as they surely did quite well until 1990, how, in the light of such inefficiencies were they also able to prosper and lead the Japanese economy to and through the economic miracle? The purpose of this essay is to propose some preliminary answers to these questions. The basic argument that there were significant offsetting effects is given in the next

2 Katz [8, pp. 3,4,109–111], for example, has argued that this change in realities was a consequence of the economic growth of the Japanese economy itself.

Explaining the Japanese economic miracle 103 two sections; some suggestive empirical support for it is provided in Section III; and the essay ends with a few implications drawn from it in Section IV.

I There are a number of well-known reasons that have been proposed to explain the cause and maintenance for almost 20 years of the Japanese economic miracle. In summary form they include: (a) Low foreign prices of the raw materials and agricultural products that Japan needed to import. (b) Sound government monetary and fiscal policies. (c) A high savings rate that provided large quantities of funds for investment. (d) Government forecasts of business activity that, though optimistic, repeatedly turned out to be low, thereby making it possible for investors to realize returns on their investments that were over and over again greater than expected. This gave investors confidence throughout the period that future investments would be profitable. (e) Successful identification and targeting by the Ministry of International Trade and Industry (MITI) of “sunrise” industries that, once so designated, became eligible for favorable allocations of foreign currency and subsidized government loans. And (f) The technology needed for advancement was readily available from abroad and did not require large expenditures of resources to secure and adapt for Japanese purposes and to Japanese production conditions. By the end of the miracle period in 1973, foreign prices were no longer so low,3 forecasts were more accurate and did not portray as optimistic an outlook, MITI’s successes were mostly in the past, and Japan had caught up to Western technology to such an extent that there was little left to import. With respect to the last of these items, further Japanese technological advancement required the assignment of much more significant quantities of resources to research and development. Nevertheless, Japanese real GNP was still able to grow at a respectable rate (on average, about 4% per year) until 1991. Though certainly contributing factors, not one of reasons (a) through (f ) speaks directly to the issues raised above of the survival and prosperity of Japanese firms in the face of what would be called severe inefficiencies in Western economics textbooks. To address them requires identification of the source or sources of inefficiency, the manner in which the inefficiencies that emerged manifested themselves in relation to Japanese firms, and the compensating effects that mitigated their severity and even turned them into advantages. However, the notions of efficiency and inefficiency are abstract and, as such, can only be characterized and their manifestations described with respect to a suitable model of the firm. It is also necessary that appropriately related models, with variations that permit distinctions between the two cases, be employed as the basis for understanding and explaining the behavior of both Japanese and Western firms. Otherwise, efficiency and inefficiency concepts would have different meanings

3 The first wave of sharp increases in the prices of oil and petroleum products began in 1973 with the first Organization of the Petroleum Exporting Countries (OPEC) oil embargo.

104 Explaining the Japanese economic miracle for the two kinds of firms, and Japanese efficiency or the absence thereof (i.e. inefficiency) could not be compared to that in the West. Thus it is assumed below that the operation of Western firms is reasonably captured by the traditional, static model of “profit maximization” under perfectly competitive conditions.4 That model, of course, can appear in a variety of forms, short run or long run, involving alternative objective functions that are maximized or minimized, with or without constraint, and under conditions of certainty or uncertainty. Regardless, such a firm is henceforth referred to as a pure capitalist firm, and the perfectly competitive economy in which it is embedded is said to be a pure capitalist economy. Although this model, in all of its forms, is somewhat removed from the reality in which actual Western firms make decisions and take action, it is invoked in present argument not only because of its convenience and simplicity, but also because of the fact that it gives expression to the most common perceptions of Western efficiency both within the firm itself and with respect to the economy at large. The latter efficiency, which often goes under the name of Pareto optimality or allocative efficiency, needs no further attention at this point. But the former requires some clarification. There are three kinds of efficiency present in a pure capitalist firm that focuses on the maximization of profit in the form of the spread between revenue and costs: (i) technological efficiency in that the maximum output is obtained from each input combination; (ii) input-mix efficiency that results from the use of those input combinations that produce output quantities at minimum input cost; and (iii) output-level efficiency achieved by producing that level of output that maximizes profit.5 Typically, (ii) arises in the presence of (i), and (iii) arises in the presence of both (i) and (ii). It is, of course, possible for technological and input-mix efficiency to hold up in the face of a lack of output-level efficiency. All that is required is that the firm produce a non-profit-maximizing level of output. But if technological efficiency is in place and input-mix efficiency is not, then the firm cannot be maximizing its profit since an appropriate, alternative input mix applied to producing the same level of output would increase that profit. This does not mean, however, that the firm is necessarily producing an output level different from that which would maximize its profit with the efficient input mix. Similarly, without technological efficiency, the firm can be neither minimizing production cost nor maximizing profit. But it could still, by accident, be using the optimal input mix and producing the optimal output level that would obtain were all three types of efficiency present. Now it has been suggested elsewhere that, for cultural reasons, the model of the pure capitalist firm is not suitable for explaining the behavior of Japanese

4 This, clearly, is a general statement. It is not intended to imply that all Western firms always select the profit-maximizing course of action in all situations or that all Western markets are perfectly competitive. 5 The efficiency relating to the financing of the firm’s investment undertakings and, in the long run, that depending on minimum-per-unit-of-output cost are also relevant but, for present purposes, will be ignored.

Explaining the Japanese economic miracle 105 firms because, at a minimum, the assumption of profit maximization in any of its forms is not appropriate.6 Unfortunately, space will not permit repetition of the argument in detail here. Suffice it to say that, reflecting a cultural tradition quite different from that of the West, input quantities would not ordinarily be chosen to minimize cost at each level of output, output would not necessarily be produced to maximize profit, and investments would not generally be undertaken so as to maximize their expected rate of return for a given degree of risk. Such motivations as the fulfillment of social and economic obligations, the furtherance of intra-firm harmony and unity, loyalty to employees and others with whom the firm has business dealings, and the preservation of the firm’s social status are more important in the making of employment, production, investment, marketing, and financial decisions.7 Indeed, the primary purpose of the Japanese firm is to provide service to society and benefit to employees,8 and this is best achieved through steady, long-term growth. Thus input and output quantities may be determined on the basis of long-term targets emerging from, say, plans to expand into certain markets, where both the quantities and targets have little to do with short-run or long-run profitability considerations; parts may be purchased from suppliers with whom the firm has a long-term relationship rather than from sources offering the lowest price; investments might be made so as to maintain the firm’s social status by providing a product line similar to that of competitors; and so on. To introduce these alternative motivations by modifying the model of the pure capitalist firm, and then to analyze their effects in that context, are not straightforward tasks – though suggestions as to such modifications are presented in Essay 9 below. For, in addition to the absence of the input-mix and output-level efficiencies described above, loyalty to employees, say, might imply that the firm is prepared to “overhire” or hoard labor. Although, from the point of view of the pure capitalist firm, such excessive hiring of labor would imply the absence of technological efficiency, it would still be possible to define a production function for the Japanese firm with respect to which that same “excessive” quantity of labor would appear to be technologically efficient. But in that case, the model of the pure capitalist firm and that of the Japanese firm, each with its own distinct production function, would have become so disparate that it would not be possible to compare efficiency and inefficiency, in any of the above senses, across them. The approach below avoids this difficulty by ignoring those characteristics of the Japanese firm and its employees that are not describable with respect to a Western- or pure-capitalist-style production function. In particular, since it has been indicated above that profit maximization is generally not relevant to Japanese decision-making, all of the elements of the pure capitalist model will be retained

6 Katzner [9], appearing as Essay 2 in this volume. Like the previous assumption that Western firms are profit maximizers, this, too, is a generalization. It does not mean that no Japanese firm ever chooses a profit-maximizing option. Recall note 4. 7 Benedict [4, Chs. 5–8] and Nakane [11, pp. 82–83, 89–91]. 8 Clark [5, p. 137].

106 Explaining the Japanese economic miracle except that decisions will be made by some other means. Present purposes do not require that a precise alternative decision criterion be specified; it only matters that all forms of profit maximization, including cost minimization, be excluded. This will mean, in the argument that follows, that varying assumptions will be made regarding the absence of profit maximization and, hence, the extent of similarity between the pure capitalist and Japanese firms. Regardless, since the models of the Japanese and Western firm employed here are identical except for the respective lack and presence of profit maximization in decision-making, the notions of technological, input-mix, and output-level efficiency are comparable across cases, and only Western firms will, in general, be efficient with respect to all three. That is, Japanese firms, due to the lack of profit maximization in them, will typically be, by definition, less efficient than their Western counterparts. Such firms will be called quasi-capitalist, and the economy surrounding them, in which the behavior of all firms is not generally derived from profit maximization, a quasi-capitalist economy. It might be thought that, from a theoretical perspective, the Japanese focus on long-term growth may be interpreted as a way of maximizing the present value of expected future income streams. In this sense, then, the Japanese firm would be maximizing a form of profit just like any pure capitalist firm. But the act of merely pursuing growth carries, by itself, no logical implication either way of whether that growth is optimal, or whether the firm is maximizing anything at all. And, as suggested above, cultural traditions point to the use of criteria other than maximization in the making of decisions, even in the long run.

II Two manifestations of the absence of profit maximization in Japanese firms are explored here: the hiring and keeping of “too many” employees, and the “misallocation” of resources in the economy at large, and among investment opportunities in particular. Consider the hiring and keeping of employees first. Japanese firms, especially those who offer “lifetime” employment, have been known for their enthusiastic and even “excessive” hiring of new employees (especially university graduates) in good times, and their reluctance to fire them when business activity sours.9 This will be understood in the quasi-capitalist model to mean that Japanese firms employ more individuals than they would if they were making hiring decisions on the basis of profit maximization. And it signals the absence of one or more of technological efficiency, input-mix efficiency, and

9 Abegglen and Stalk [1, pp. 6, 199–203]. This statement applies mostly to the medium-size and large Japanese manufacturing firms (i.e. enterprises with 100 or more employees) which, between 1955 and 1990, employed roughly 45% of the manufacturing labor force and accounted for approximately 63% of value added in Japanese manufacturing (according to the Census of Manufacturers conducted by the Japanese Ministry of International Trade and Industry). These firms constituted the sector of the Japanese economy that, during the miracle period and up to 1990, was thought to be the most efficient in the world, and that was both emulated and feared almost everywhere.

Explaining the Japanese economic miracle 107 output-level efficiency. Now if such a firm were to try to sell its output in highly competitive (e.g. world) markets, it would find, according to the theory of the pure capitalist economy, that its costs were too high to survive, and it would be driven out of those markets by lower-cost firms that make decisions in compliance with profit maximization. But in the Japanese case there are at least two mitigating factors. On the one hand, it is well known that Japanese employees tend to work very long hours, more hours than they are compensated for by Western standards of work time in relation to compensation.10 This fact may be introduced into the quasicapitalist model of the Japanese firm in several ways. One method, consistent with the present requirement of maintaining identical production functions in the purecapitalist and quasi-capitalist firm, is as follows. In both pure- and quasi-capitalist cases, let labor be measured in hours, each unit of which carries the same quality. Let the argument in the production function representing labor denote only those labor hours for which the firm actually pays, and let the output of that production function be the output of the paid labor only. Given market conditions, taken to be the same in both cases, let the firm determine the quantities of inputs it hires and output it produces according to whatever decision criteria it uses. Assume that the quasi-capitalist Japanese firm is producing that level of output specified by the production function in relation to actually paid labor so that it is technologically efficient. But, as described in the preceding paragraph, that cannot mean that that firm is hiring labor and producing output in quantities that both minimize cost and maximize profit. For the Japanese firm is necessarily less efficient and has higher costs than the pure-capitalist firm. Now, with respect to the former, keep the quantities of nonlabor inputs fixed and add in the presence of unpaid labor together with the extra output derived from it. Then, although the quasi-capitalist firm’s total cost remains unchanged, its effective unit cost falls. Thus the increase in output provided by unpaid labor offsets at least some of the extra cost of hiring too much labor. On the other hand, Japanese firms pay their employees according to a bonus system11 which allows the firms to effectively lower wages and salaries generally, and reduce bonus payments during hard economic times. Whereas a pure capitalist, profit-maximizing firm would always pay an employee the value of his marginal product, since approximately one-third of the Japanese worker’s income from employment has, in the post-war period, come as a bonus, the Japanese firm could be thought of as paying the employee two-thirds of what it would “normally” pay to meet its labor requirements, were it operating as a pure capitalist firm.

10 The “salaryman” usually works into the evening and many of these hours would be considered unpaid or nonsalaried hours in the West. Even when evening hours are spent with colleagues at restaurants and Geisha houses, they are often thought by the Japanese to be work since work-related matters frequently are discussed and sometimes work-decisions are made. Higa and Lupardus [7, p. 92] estimate that, in 1994, the typical male worker in Japan was paid for only 82% of the hours he actually worked. 11 Abegglen and Stalk [1, pp. 196–197].

108 Explaining the Japanese economic miracle Additions to the latter employee’s income would come from profit, when the firm is profitable enough.12 Obviously this results in total labor costs that are one-third below what they would be if bonuses were not separated out from wages and salaries and paid out of profits, and if the same quantity of labor were hired in each situation. In the last assertion, “quantity of labor” is meant to refer to total labor hours, including, in the language of the previous paragraph, both “paid” and “unpaid” labor, and the statement itself is intended to carry no implication with respect to profit maximization or efficiency in any form. Should the firm not do well, it reduces or eliminates the bonus rather than letting employees go. Thus, when total labor hired (including the unpaid component) is considered, and when the total production cost associated with it is in view, the overhiring of employees by Japanese firms need not necessarily lead to cost disadvantages when competing in world markets against firms that, in theory, maximize profits. These general considerations suggest reasons why the labor-hiring and wagerenumeration practices and production decisions of the quasi-capitalist Japanese firm enable it to offset the lower profit level resulting from the inefficiencies that have been attributed to an absence of cost minimization or profit maximization. To examine the matter further, attention turns, for a moment, to the following comparative example. Focus on a production function of the form x=

√ y,

where x indicates quantities of output and y denotes quantities of total labor input. (The latter is comprised, in the case of the Japanese firm, of both paid and unpaid labor.) To keep matters simple, assume that no other inputs are employed. Suppose that this same production function arises in both the quasi-capitalist (Japanese) and pure-capitalist firm, that labor is homogeneous across these firms, and that the Japanese firm is technologically efficient. Since there is only one input, the issue of whether there is input-mix efficiency does not arise. Represent the wage per unit of labor (averaged over paid and unpaid components) by w. Taking the output price to be p, profit is given by √
= p y − wy.

(1)

With w and p fixed, choosing y so as to maximize
yields the input demand function y=

p2 . 4w2

(2)

Were the Japanese firm to determine its employment level like a pure capitalist firm, it would decide on the quantity of labor it would hire on the basis of profit

12 The fact that there may be a minimum bonus that is always paid regardless of economic conditions is ignored.

Explaining the Japanese economic miracle 109 maximization. Then, in that case, it would follow that if that firm’s employees were paid two-thirds of the full capitalist wage w, that is, if (2/3)w were substituted in place of w in (2), then the Japanese firm would hire nine-quarters or more than twice the number of employees than the pure capitalist firm would hire at wage w. In addition, substituting (2) into (1) and differentiating the result with respect to w leads to the inequality ∂
< 0, ∂w

(3)

for all w > 0. Thus, assuming p and w are given, if the Japanese firm were to pursue the objective of profit maximization, it would hire more labor at the wage (2/3)w than it would if it hired at w and, by (3), its profit would be larger. But recalling previous observations regarding overhiring by the Japanese firm, an alternative scenario might be envisaged. Consider, for that purpose, what has just been described as the firm’s profit-maximizing demand for labor at the wage (2/3)w. If the firm were to hire that same quantity of labor at wage w, it would clearly fail to realize maximum profit. In that case, however, profit maximization would, of course, be restored by paying labor (2/3)w, and the firm would not face what has been contemplated as a survival problem.13 But the implicit hypothetical comparison between the pure capitalist and Japanese firm in what has just been discussed does not fully describe the condition of the quasi-capitalist Japanese firm and its decision-making complex. For the quasi-capitalist firm, it has been assumed, does not appeal to profit maximization to determine its employment. Furthermore, overhiring by that firm means that, as previously indicated, it actually hires more labor than that which would equate the value of its marginal product to w. Thus, if VMP( y) represents the value of the marginal product of labor at y, if w is the given capitalist wage and y satisfies the equation VMP( y) = w from which (2) is derived when w is variable,14 and if y0 is the quantity of labor that is hired by the firm, then as shown in Figure 6.1, y0 > y  and VMP( y0 ) < VMP( y ). In addition, the actual wage paid by the Japanese firm is (2/3)w , not w . Thus, as specified above, if y is the value of y that solves VMP( y) = (2/3)w , then y < y (refer again to Figure 6.1), where y = (9/4)y  from (2). Hence, as long as the Japanese firm does not hire more than nine-quarters the labor it would hire if it were a pure capitalist firm, that is as long as y0 < y , the value of the employee’s marginal product is larger than the wage, (2/3)w , he is paid. Moreover, since VMP( y0 ) > (2/3)w for y ≤ y0 < y , inequality (3) implies that, with wage (2/3)w and a fixed value for p, before-bonus-payment profit at each such y0 is higher than the pure capitalist profit at y with wage w . And because, as is also implied by (3) when bonus payments are ignored, profit

13 Evidently, since the Japanese firm would then pay a bonus on top of (2/3)w, the allocation of resources that results would be unlikely to achieve the Paretian efficiency of pure capitalist systems. This latter manifestation of inefficiency is taken up shortly. √ 14 That is, y satisfies VMP( y) = p/(2 y) = w .

110 Explaining the Japanese economic miracle w VMP

A

w′

G B

w0 2 3

E

w″

F

D y′

C y0

y″

y

Figure 6.1 Profitability at various wages and employments in the Japanese firm.

at y is greater than that at y , that larger profit could be preserved to some degree were the firm to hire a little beyond y at a wage of (2/3)w . Once again, the loss from overhiring at wage w is more than offset (before bonus disbursement) by paying employees (2/3)w . But once the payment of the bonus is taken into account, the offsets to the consequences of overhiring described here are reduced by the total of all bonus disbursements. Returning to Figure 6.1, if, at wage w , the firm hires y0 instead of the profit maximizing y , its revenues rise by area ABCD. Since it actually pays a wage of only (2/3)w , its labor costs increase by EFCD rather than by AGCD for a pre-bonus offset savings of AGFE. Its pre-bonus profit rises by ABFE. That profit increase, however, will be eaten away depending on the size of the bonus paid. For example, suppose that, while the Japanese firm pays a nominal wage of (2/3)w , the effective wage payment is actually somewhere between w and (2/3)w by virtue of the additional bonus to be subsequently provided. That would mean that the payment of the bonus would absorb some, all, or possibly even more than all, of the additional profit achievable by employing y0 labor at nominal wage (2/3)w . In general, the degree to which that absorption takes place depends on the elasticity of production with respect to labor input implicit in the production function. The latter point, however, need not be pursued further here. Regardless, to the extent that Japanese firms are in fact prepared to accept a smaller profit than that which would accrue to the comparable pure capitalist firm (a willingness that, as argued below, is quite common in Japan), they could tolerate bonus payments associated with effective wages above, say, w0 in Figure 6.1. In that illustration, were the firm to return the full capitalist wage w to each employee,

Explaining the Japanese economic miracle 111 its after-bonus-payment profit would be reduced from the capitalist profit at A by area AGB. The second manifestation of inefficiency in Japanese firms has to do with the implications for the misallocation of resources in the economy at large and among investment opportunities in particular, and the resulting higher costs that that entails. The misallocation comes about from such practices as overhiring and the use of criteria (like, for example, the preservation of the firm’s social status) other than expected rate-of-return maximization (for a given risk) in making investment decisions. The compensating effect that offsets the inefficiencies and higher costs here is the willing acceptance of less-than-maximum profit by the firms’ owners, managers, and bankers. That willingness comes about for several reasons.15 First, the members of the board of directors of a firm largely come from inside that firm. Any outsiders are usually representatives of other companies that have very close business relations with the firm. Thus, as a consequence, firms tend to be managed more for their employees than for their owners. This means that longterm growth, which often implies enlargement of market share by expansion in foreign markets, and which generally provides greater job-security for employees, is, as suggested earlier, a much higher priority than profit maximization in any of its forms. Second, the owners of the firm, i.e. those who hold its equity, have little choice since, although they are entitled to dividends by dint of their ownership position, once those dividends are paid, they have little say in the management of the company’s affairs. Moreover, a sizable portion of the firm’s equity is held by those same firms that have close business relations with it and who, as indicated above, are more interested in long-term growth than either short- or long-run profit maximization. Third, the Japanese banks are managed in much the same way with emphasis on long-term growth. A firm and bank that are tied together, in that the former is heavily in debt to the latter, will hold each other’s equity. A bank representative may sit on the firm’s board of directors. The bank feels obligated to underwrite portions of the firm’s investments with low-cost loans, thereby decreasing the spread between its lending and borrowing rates (and hence its profit), and to help in rescuing and reconstructing the firm when it runs into financial difficulty.16 But, it should be emphasized, the lack of concern within the Japanese firm about its failure to maximize profit is not due only to the fact that it can count on banks’ help during times of severe financial distress. For, as has been repeatedly noted earlier, firms do not consider profit maximization to be a goal. Indeed, Japanese firms have generally done quite well without maximizing profit and there is therefore no reason to be concerned about it. Of course, Japanese firms

15 Abegglen and Stalk [1, pp. 166–168, 174–175, 183–186]. A rigorous theoretical analysis of the mechanism through which the willingness of firms’ owners, managers, and bankers to accept less-than-maximum profits will lead to a misallocation of resources is not developed here. 16 In this regard it might also be noted that downward pressure on lending rates was exerted by the monetary policy pursued by the Bank of Japan throughout the miracle period.

112 Explaining the Japanese economic miracle do care about the possibility of bankruptcy because it implies a failure to achieve their primary goals of service to society and benefit to employees. But any worry about bankruptcy still does not focus their attention on profit maximization. In any case, previous argument indicates that the net effect of these practices and policies by Japanese firms is to reduce their profitability to levels lower than those that would be secured if they behaved like pure capitalist firms and maximized profits. And being content with less-than-maximum profit tends to mitigate the effects of taking obligation, internal harmony, and loyalty as the primary motivations behind business decisions in general, and of using, say, the preservation of the firm’s social status as a criterion for investment decisions in particular. That is, although the allocation of resources that results is not efficient with respect to either input mix or output level or, in other words, although the Japanese economy is “wasting” resources by producing beneath what, in terms of the pure capitalist model assumed here for comparative purposes, is designated as its production possibility surface,17 the inefficiencies that arise are offset on the plane of competition in domestic markets by the willingness to accept smaller profits and lower rates of return on invested capital. The same conclusion applies with respect to world markets too. For pure capitalist economies, the explanation of why countries trade and what they gain from doing so is provided by the theory of comparative advantage. In that theory, comparative advantages, as traditionally defined under Western suppositions, are based on marginal rates of transformation along an economy’s production possibility surface and, to be on that surface, requires, in part, that all firms produce outputs with cost-minimizing quantities of inputs, which the firms in a pure capitalist economy are assumed to do automatically. Countries specialize in the production of goods in which they have comparative advantages and trade part of their output to other countries in exchange for goods in which the original countries have comparative disadvantages. At equilibrium exchange rates after trade has taken place, and ignoring international capital flows, a country that has a comparative advantage in a specific good produces that good at a smaller dollar cost after exchange rates are taken into account. Thus it sells that good in world markets at lower prices than other countries are able to offer. Those other countries buy the good rather than produce it at home because it costs them less to do so. But none of this applies to a country like Japan which cannot generally be represented as producing on its production possibility surface. And the Japanese cost disadvantage in world markets that results, assuming the theory of comparative advantage is used as the point of reference, is offset by pricing goods at levels at which they will be sold and accepting smaller profits and, sometimes, even losses. This is entirely consistent with the goals of Japanese firms described above to grow in the long term and expand world market share.

17 If, by chance, it actually wound up on that surface, the absence of output-level efficiency is likely to result in the Japanese economy producing at the “wrong” i.e. allocatively inefficient) surface point.

Explaining the Japanese economic miracle 113

III As described in the previous section, two inefficiencies that, from a Western vantage point, might prevent the Japanese economy from growing as fast as it could are the hiring of too many employees and the using of criteria other than expected rate-of-return maximization in investment decision-making. In sustaining the Japanese economic miracle, however, it has been argued that these inefficiencies were partially offset by compensating labor with less than what a pure capitalist firm would normally pay to meet its labor requirements (for example, w in Figure 6.1). An additional offset was due to a willingness, on the part of the owners of capital, to accept profit or a rate of return on capital that was lower than what could otherwise be obtained. To test empirically these latter propositions would, ideally, require comparing estimates of actual wages and pure capitalist wage requirements on the one hand, and estimates of actual and maximal profit rates on the other. Such comparisons are far too complex to undertake here. Rather, it will suffice instead to present some data that are not only consistent with the propositions developed here, but also suggest that they might hold up under more rigorous empirical scrutiny. Data are provided not only for portions of the miracle period (1955–1973), but also for certain time intervals up to 17 years beyond in order to cast doubt on the claim, held in many quarters, that the Japanese economy was highly efficient until the early 1990s. The basic idea is to take the US economy as representative of an “efficient” economy and use it as a foil against which the Japanese inefficiencies and their offsets mentioned above can be empirically placed in sharp relief. This is quite parallel to the approach outlined in Section I and applied in Section II of thinking of the Japanese firm in terms of a pure capitalist model in which the assumption of profit maximization has been dropped. Thus, under the appropriate conditions, lower wages in Japan as compared to the US would point to the possibility that Japanese firms were overhiring as described earlier and paying labor less than what pure capitalist requirements would demand, and lower profit rates in Japan would suggest the possibility that Japanese investors willingly accept smaller rates of return on their invested capital. Consider wages first. Actually, it will be convenient to focus on productivity data which may be assumed to reflect wage values.18 Pilat [12] compares Japanese–US labor productivity both for the economy as a whole and in seven of its sectors (agriculture-forestry-fisheries, mining, manufacturing, construction, transportation-communication, electricity-gas-water, and government-other services) for selected years between 1885 and 1990. The specific measure with respect to which his comparisons are made is estimated sectoral contributions to

18 For example, in terms of the marginal products implicit in Figure 6.1 (which are assumed here to approximate average products), an American firm would be thought of as being at point A where it employs y and pays a wage of w . The comparable Japanese firm, however, would hire, say, y0 , where marginal productivity is lower than in the American firm, and pay the wage (2/3)w < w . Thus lower productivity in the Japanese firm would correspond to a lower wage.

114 Explaining the Japanese economic miracle Table 6.1 Comparative labor productivity levels by sector: Japan as a percentage of the US 1953

1965

1975

1990

Agriculture, forestry, and fisheries Mining Manufacturing Construction Transportation and communication Electricity, gas, and water Government and other services

5.8 4.5 22.5 19.8 15.6 15.5 42.6

8.3 11.2 36.3 19.7 25.5 22.1 48.0

9.4 31.2 69.7 49.0 39.2 29.2 67.9

9.1 68.7 100.6 75.5 39.1 43.3 91.2

Whole economy

19.8

30.7

52.5

77.7

Source: Pilat [12, p. 368]

gross domestic product (GDP) per number of persons employed in those sectors. In the years in and around the miracle period that he covered (1953, 1965, and 1975), Pilat found that labor productivity, compared to the US, was consistently and significantly lower overall and in all sectors in Japan (Table 6.1). For example, in 1975, manufacturing GDP per person in Japan was 69.7% of that in the US. With one exception, the same was true in all other years included in his study (six years running from 1885 to 1939 – not shown in Table 6.1 – and 1990). The one exception was in manufacturing in 1990 where GDP per number of persons employed was essentially the same in both countries. However, it does not follow that the Japanese were not overhiring and paying labor less than pure capitalist wage requirements in manufacturing in 1990 for at least two reasons. First, Pilat [12, p. 366] also indicates that manufacturing GDP per hour worked was still significantly lower in Japan as compared to the US in 1990. (As a rule, there was a greater sectoral discrepancy than that indicated above between Japan and the US in terms of GDP per hour worked.19 ) Second, in 1990 the US economy was close to the trough of its business cycle, whereas the Japanese economy was near its peak. Under such conditions, aggregate productivity data at the sectoral level would tend to blur the comparison between wages paid in manufacturing by individual Japanese and US firms. Movshuk [10, p. 61] provides an alternative approach to productivity comparisons that is broader in concept but relates only to manufacturing. Using the Tornquist-Theil index, and accounting for labor, capital, and intermediate inputs, he estimates total factor productivity in the manufacturing sector and 23 of its subsectors, and compares the results for Japan (among other countries) and the US in 1975 and 1990. Movshuk concludes that, in both years, total factor productivity in manufacturing as a whole was significantly lower in Japan (Table 6.2): 72% of

19 To the extent that the Japanese work longer hours than actually reported (it has been suggested earlier that they work more hours than those for which they are paid), this discrepancy would be increased. See note 10 above.

Explaining the Japanese economic miracle 115 Table 6.2 Total factor productivity in Japan relative to the US (US total factor productivity = 100)

Food products Beverages Textiles Clothing Leather products Footwear Wood Furniture Paper products Printing, publishing Industrial chemicals Chemical products Rubber products Plastic products Pottery, china, etc. Glass products Non-metallic products Iron and steel Non-ferrous metals Metal products General machinery Electrical machinery Transportation equipment Total manufacturing

1975

1990

86 107 79 123 133 112 28 47 77 51 68 97 150 75 60 88 68 100 68 87 77 60 96

66 84 66 95 119 109 31 42 84 66 66 81 136 64 73 84 79 115 78 102 92 118 100

72

84

Source: Movshuk [10, p. 61]

that of the US in 1975, and 84% of that of the US in 1990. Moreover, it was higher in Japan in only five of the 23 subsectors in 1975 (beverages, clothing, leather products, footwear, and rubber products), and in only six, three of which were the same, of the 23 subsectors in 1990 (leather products, footwear, rubber products, iron and steel, metal products, and electrical machinery). Thus, under the assumptions made here, Movshuk’s results are suggestive of the same phenomenon of wages paid being less than pure capitalist wage requirements in Japan.20

20 It should be noted, however, that there are other possible explanations of why labor productivity and returns to capital (to be discussed next) have been lower in Japan. For example, labor productivity might have been lower due to technological backwardness. And returns to capital might have been lower because, in the catching-up-to-the-West process, Japan was more dependent on physical capital accumulation than was the United States (Hayami and Ogasawara [6]) implying, thereby, a lower marginal productivity of capital. Thus, although the data presented here are unable to establish the proposition that overhiring and a willingness to accept lower rates of return on invested capital were present in Japan between 1955 and 1990, they are still consistent with it. A rigorous test of the proposition is beyond the scope of the present paper.

116 Explaining the Japanese economic miracle Turn now to the comparison of profit rates between the US and Japan. Ando and Auerbach [2, p. 327] provide two estimates of the returns to capital in Japan and the US in each of the 22 years from 1967 to 1988. One of these estimated series is derived from accounting earnings (adjusted for such things as the inflation-induced understatement of depreciation and inventory costs, capital gains on net financial liabilities and, in the case of Japan, reserves) and is presented in Table 6.3. The table shows that the average rate of return to capital over the entire period was considerably higher in the US (0.060 in Japan as opposed to 0.119 in the US). Furthermore, in only three of the 22 years (1968–1970) was the rate of return on capital higher in Japan. The second series estimated by Ando and Auerbach is based on market equity yields and includes capital gains on land. Since the conclusions drawn from the second, although not as strong, largely mirror those obtained from the first, the second series is not reproduced here. It has been argued that the data presented above are suggestive of the possibilities that, as a rule, the Japanese pay labor less than pure capitalist wage requirements, and accept less than the maximal rate of return on their invested capital. To the extent that these

Table 6.3 Returns to capital in Japan and the US 1967–1988 Japan

US

1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988

0.070 0.088 0.083 0.081 0.082 0.065 0.030 0.010 0.045 0.042 0.048 0.047 0.058 0.067 0.082 0.076 0.076 0.068 0.069 0.054 0.045 0.046

0.090 0.075 0.069 0.077 0.085 0.087 0.097 0.140 0.140 0.135 0.140 0.170 0.193 0.173 0.138 0.117 0.122 0.124 0.121 0.093 0.103 0.127

Average

0.060

0.119

Source: Ando and Auerbach [2, p. 327]

Explaining the Japanese economic miracle 117 characteristics hold in fact, they may be viewed as compensating offsets to the inefficiencies created by the practices of Japanese firms of (i) hiring too much labor, and (ii) using criteria other than maximizing the expected rate of return in making their investment decisions.21

IV The argument of this essay has been that Japanese firms led the Japanese economy to and through the economic miracle period in part by successfully compensating for what in Western economics would be regarded as significant inefficiencies. These inefficiencies are sufficient in pure capitalist economies to result in large numbers of bankruptcies, unemployment, and possibly even depression. They arise from the pursuit by Japanese firms of goals other than profit maximization, and they manifest themselves at least in overhiring by firms, and a misallocation of resources in general and among investment opportunities in particular. There are several implications of this argument that can be applied on a much broader scale rather than simply to the particular case of Japan. First, efficiency in the Western sense (i.e. technological, input-mix, and outputlevel efficiencies, and optimal resource allocation) may be partly sufficient, but it certainly is not necessary for strong economic performance and development. As long as there are appropriate offsetting features that compensate for the inefficiencies, significant economic progress can still occur as it did for almost 40 years in post-World-War-II Japan. In such economies, which have been referred to above as quasi-capitalist, the price mechanism operates differently and does not function to allocate resources in the same way as in a purely capitalist system. Regardless, it is clear that the normative idea that Western efficiency is “good” and ought to be pursued is culturally based. Such efficiency may be relevant and appropriate for Western-type economies or purely capitalist systems, but its pursuit in the quasi-capitalist economies of the non-Western world may, at best, not have the desired effects and, at worst, could cause substantial harm. Second, such quasi-capitalist economies, that is, their institutions and the behavior of economic agents in them, can reflect cultural and social norms and traditions quite different from those of the West. In the argument of the present essay, this appears in the motivations other than profit maximization that drive economic action.22 The cultural and social roots of such alternative motivations in the case of Japan have been described by Bellah [3], Benedict [4], and Nakane [11].23 And the occurrence of the Japanese economic miracle is prima facie evidence that strong economic growth can take place in a cultural and social

21 The deterioration of the Japanese economy after 1990 does not necessarily mean that these offsetting effects disappeared or even were reduced in size. Rather, the offsetting effects could be seen as still present, though no longer sufficient to maintain prosperity and stave off decline. 22 Japanese institutions such as the lifetime employment system can also be thought to emerge from cultural characteristics that do not exist or are of relatively minor significance in the West. 23 See also Katzner [9], this volume, Essay 2.

118 Explaining the Japanese economic miracle environment that eschews Western norms and values. Different (i.e. non-Western) cultural and social norms and traditions, then, need not necessarily relegate an economy to a permanently undeveloped state. Finally, reflecting on the entire post-World-War-II Japanese economic experience, it is clear that quasi-capitalist economies are subject to boom and bust just like capitalist economies. The Western world is not the only one vulnerable to inflation, speculative price bubbles, miscalculations by individuals and policymakers, and so on. Sustained prosperity is possible as is recession and, although not in recent Japanese and Western history, depression. Of course, since the internal structures and dynamics of these economies are different, the causes of downturns and upswings, and the economic reactions and adjustments to them, will not be the same. And different policies are likely to be required to deal with problems like inflation and unemployment that arise over the business cycle.

References 1. Abegglen, J.C. and G. Stalk, jr., Kaisha, The Japanese Corporation (Tokyo: Tuttle, 1985). 2. Ando, A. and A.J. Auerbach, “The Cost of Capital in Japan: Recent Evidence and Further Results,” Journal of the Japanese and International Economies 4 (1990), pp. 323–350. 3. Bellah, R.N., Tokugawa Religion (Glencoe: Free Press, 1957). 4. Benedict, R., The Chrysanthemum and the Sword (Boston: Houghton Mifflin, 1946). 5. Clark, R., The Japanese Company (New Haven: Yale University Press, 1979). 6. Hayami, Y. and J. Ogasawara, “Changes in the Sources of Modern Economic Growth: Japan Compared with the United States,” Journal of the Japanese and International Economies 13 (1999), pp. 1–21. 7. Higa, T. and K. Lupardus, “Why Don’t Workers Claim All Their Overtime?” Japan: Why It Works, Why It Doesn’t – Economics in Everyday Life, J. Mak, S. Sunder, S. Abe, and K. Igawa, eds. (Honolulu: University of Hawaii Press, 1998), pp. 91–97. 8. Katz, R., Japan: The System that Soured (New York: Sharpe, 1998). 9. Katzner, D.W., “Western Economics and the Economy of Japan,” Journal of Post Keynesian Economics 21 (1999), pp. 503–521. This volume: Essay 2. 10. Movshuk, R., “International Comparison of Productivity Levels and Consumer Preferences in Industrial Countries,” Ph.D. thesis, Osaka University, 1999. 11. Nakane, C., Japanese Society (Berkeley: University of California Press, 1970). 12. Pilat, D., “The Sectoral Productivity Performance of Japan and the U.S., 1885–1990,” Review of Income and Wealth 39 (1993), pp. 357–375.

7

The role of optimization in economics1

As a discipline, economics is primarily concerned with the allocation to alternative employments of scarce resources that have various possible uses. How, it asks, are the finite amounts of land, labor, and capital that are available to a society apportioned to individual and government activities, and to the production of the various products that benefit individuals and society at large? How do those allocations modify over time? And how might allocations that are thought to be inappropriate or unsatisfactory be changed? Because the quantities of resources are limited to the extent that not all needs and desires can be met, it is clear that choices have to be made. And therefore, much of economics is concerned with the analysis and consequences of the making of choices or, in the jargon of the discipline, the making of decisions. Of the many analytical tools that could be brought to bear on the decision problem, the most prominent has been optimization. For almost 200 years, economists have been concerned to find, in a substantial number of contexts, the optimum solution values of vectors of decision variables, thereby discovering the attainable extrema of certain appropriately defined objective functions. The logic behind that approach is derived from the idea that the “best” alternative from any collection of available options ought to be pursued, and hence the necessity arises of making the optimal choice decision under the circumstances presented. In all of this, economists are conscious of the fact that most economic choice decisions are inevitably constrained decisions. The finiteness of resources imposes constraints at all analytical levels, from individual consumers and firms, to industries and markets, to the economy as a whole. Such constraints may be either implicit or explicit, thus affecting the nature of the mathematics required. Indeed, both unconstrained or, respectively, constrained optimization techniques in relation to first- and second-order maximization or minimization conditions have been employed in the past. Although the appearance of optimization in economics has

1 Optimization and Optimal Control, P.M. Pardalos, I. Tseveendorj, and R. Enkhbat, eds. (Singapore: World Scientific, 2003), pp. 141–153. © World Scientific Publishing Co. Reprinted with permission. Thanks are due to Rob Kusner and Douglas Vickers for their help in the preparation of this essay.

120 The role of optimization in economics not, as will be indicated below, gone unchallenged, it nevertheless retains a central place in the discipline to this day. The exact moment at which economists became aware, either explicitly or implicitly, of the idea of optimization and its relevance to economics is not clear. But it is possible to pinpoint what may be the first appearance of optimization as a mathematical issue in the economics literature. That initial taste, and it was, indeed, a very small sip, was a fleeting one. For in this instance, mathematics relating only peripherally to the optimization problem seems to have crept into economics surreptitiously through the back door and was thereafter quickly expelled. The story, according to Groenewegen [6, pp. vii–viii, xxvi–xxvii] and Robertson [16, pp. 526–527], is as follows. In 1771, an Italian economist by the name of Pietro Verri published a book whose title translated into English is Reflections on Political Economy. The book immediately met with some success, and its initial edition was reprinted several times with various and differing annotations into 1772. One of the reprinted versions (the sixth) contained some mathematical commentary attributed to Verri’s friend and mathematician, Paolo Frisi. But a subsequent, and what was considered to be the definitive edition of the book that was published in 1781 deleted, among other things, Frisi’s commentary. Two parts of Frisi’s commentary are of interest here. First, early in his book, Verri [21, p. 19] argued,2 without using mathematical symbolism, that “… the price of things will be in direct proportion to the number of buyers and in inverse proportion to the number of sellers.” Verri then went on to suggest that the ends of the state could best be served by keeping the price as low as possible [21, p. 21]. In his commentary, Frisi approximated Verri’s proportionality statement by the equation P = C/V , where P represented price, C the number of buyers, and V the number of sellers. To examine the properties of Verri’s “minimum” price value, should the state be able to maintain it, Frisi then expressed that equation in differential form and equated the price differential dP to zero. This led him to the conclusion that the minimum (or, for that matter, any value of P) could be maintained by offsetting the changes in one of C or V by changes in the other so as to keep the ratio required for achieving the minimum (or, respectively, another value of P) intact.3 Later on in his book, Verri discusses the problem of obtaining maximum production with minimum labor [21, p. 68]. Frisi’s commentary here asserts that this problem, when reduced to analytical terms, is an isoperimetric problem – that is, it is mathematically equivalent to the problem of finding the maximum area that can be enclosed within a perimeter of fixed length. Although Frisi provides

2 It should be noted that the following references to and quotation from Verri’s work are based on the translation of the 1781 edition, which is not the same as the sixth version on which Frisi based his commentary. However, it is safe to assume that the material in question is similar in spirit, if not identical in wording, to that to which Frisi was responding. 3 That is, dP = [VdC − CdV ]/V 2 = 0, from which it follows that dC/dV = C/V . See Robertson [16, p. 526].

The role of optimization in economics 121 some mathematical discussion relating to the problem as set in the context posed by Verri, he does not solve it. Nevertheless, this commentary, together with that described above, seem to be the earliest mathematical references to optimization in economics. The broader historical context concerning the development of the mathematics of optimization at the time Verri’s book and Frisi’s commentary were published is worth noting. First, the use of the calculus to determine the values at which functions have maxima and minima had been worked out by Fermat 100 years earlier in the middle of the seventeenth century.4 Second, the isoperimetric problem, which was related to the Principle of Least Action (the doctrine dating to antiquity that nature acts in the shortest, most efficient way) and Fermat’s Principle of Least Time (that light only travels along paths that take the least time), had recently been solved by Lagrange in the 1750s. In so doing, Lagrange created a general technique for dealing with such matters to which Euler subsequently gave the name “calculus of variations.” 5 Because of the likelihood that there was much discussion of Lagrange’s “new” method among mathematicians at the time, it is not surprising that Frisi chose to point out in his commentary the implicit presence of the isoperimetric problem in Verri’s book. Third, around or shortly after the time that the first edition of Verri’s book was being reprinted with annotations, Lagrange was developing what later came to be known as his multiplier method for determining the critical values of functions that have maximum and minimum values under constraints.6 But, as will be pointed out momentarily, the mathematics of constrained optimization did not enter the economics literature for another 100 years. It is possible that Frisi’s use of mathematics to express indirectly the idea of optimization – short-lived in print though it was – inspired other, more direct, mathematical discussions of optimization in economics.7 Be that as it may, in 1815, Georg von Buquoy, a German economist, gave an analysis of profit maximization, with respect to the depth to which, when planting crops, fields are plowed, that was remarkably prescient of what was to come. Robertson [16, p. 527] provided a translation:8 Let the law according to which the value of the crop [R] depends upon the depth of plowing [x] be given by the equation R = f (x), and let the law according to which costs [C] depend upon the depth of plowing be expressed by the equation C = g(x). The net revenue can generally be expressed at any depth by [R − C =] f (x) − g(x), from which that value [of x] can be found at which the net revenue is a maximum. And this maximum will be at that

4 5 6 7 8

See, for example, Kline [14, pp. 347–348]. Ibid., pp. 580–583. Grattan-Guinness [5, pp. 326–327]. See Groenewegen [6, p. xxvii]. The mathematical notation employed below has been modified slightly from the original.

122 The role of optimization in economics value of x at which the first derived function of f (x) − g(x) disappears and at which simultaneously the second derived function becomes negative. Although it is expressed in terms of plowing depth, this, of course, is the marginalcost-equals-marginal-revenue statement of the necessary first-order condition for “profit” maximization, along with a statement of the second-order condition sufficient to ensure a maximum at the critical point. It was the French economist Agustin Cournot [3] who, in 1838, placed the mathematics of optimization in many of the economic contexts in which it appears today. Profit was the difference between revenue and costs, the latter both expressed as functions of output, and upon equating its derivative to zero, Cournot derived the marginal-cost-equals-marginal-revenue standard that attends virtually all of today’s introductory and intermediate microeconomics text books. He analyzed profit maximization under perfect competition (Ch. 8) and monopoly (Ch. 5). He developed the analysis of a market with two competing firms (duopoly) who are simultaneously maximizing their own profit under the assumption that the competing firm’s output remains fixed (Ch. 7). And he even explored the meaning of revenue maximization along a demand curve in terms of first- and second-order maximization conditions (Ch. 4). In all of this Cournot gives extensive accounts of the derivations and implications of first- and second-order conditions. It is with Cournot that optimization and the mathematics associated with optimization became an integral part of economic analysis. However, up to the middle of the nineteenth century, the constraints imposed by the finiteness of resources remained implicit in mathematical discussions of optimization in economics. Cournot’s presentations in particular did not require explicit account be taken of them. The first appearance of explicit constraints in relation to optimization occurred in 1854 with the verbal statement by Hermann H. Gossen [4, pp. 108–109], another German economist, of the first-order condition resulting from constrained maximization of utility (pleasure) subject to the (linear) budget constraint: Man obtains the maximum of life pleasure if he allocates all his earned money … between the various pleasures and determines the … [proportion of income for a specific pleasure] in such a manner that the last atom of money spent for each pleasure offers the same amount of pleasure. Then, in 1871, the English economist W. Stanley Jevons [8, pp. 99–100] wrote down the first-order mathematical equations for a slightly different maximization problem and, in Switzerland in 1874, Léon Walras [22, p. 127] spelled out both first- and second-order conditions for, among other things, that same problem.9

9 The page references in this Jevons citation are to the fifth (1957) edition of his The Theory of Political Economy; those to Walras’ Elements of Pure Economics are to the English translation of the 1926 edition.

The role of optimization in economics 123 Thus 100 years after Frisi, constrained optimization finally entered the economics literature in a formal, mathematical way. (In what follows, the tern “optimization” will refer to optimization in both constrained and unconstrained formats.) Regardless, during the more than 160 years since Cournot, the use of optimization and the mathematics of optimization in economics exploded. Today optimization and the mathematics of optimization appear almost everywhere across the discipline. At least three broad categories of applications may be identified. First, optimization has become the cornerstone of models that purport to explain the behavior of individual consumers and firms, and sometimes even collections of consumers and collections of firms. In this category fall such applications (some with explicit constraints and some without) as utility maximization, expenditure minimization, cost minimization, profit maximization, revenue maximization, maximization of the rate of return on invested capital, and the determination of various solutions, characterized in terms of optimization criteria, of a wide variety of games. Second, optimization has been employed to define an ideal that a collective or a society might want to attain.10 Now societies are distinct from individuals. They are made up of individuals and have an aggregative quality that individuals lack. Societies have priorities and motivations that are generally different from those of individuals, and their behaviors fall in different spheres of activity. Although there may be analogies between them, explanations of the behaviors of individuals and societies must, of necessity, deviate from each other, at least in detail. The optimization ideals that a society might pursue include Pareto optimality (a circumstance in which resources are fully employed and in which it is impossible through resource reallocation to make one person better off without making someone else worse off), welfare maximization, maximization of the value of national output, optimal taxation, and optimal economic growth. Third, optimization is used as a convenience. For example, it may be possible to frame propositions more economically by drawing on a “maximum principle.”11 Thus a representative, maximizing consumer might serve as the basis for formulating market demand or aggregate consumption functions. And it is often advantageous to establish the existence of “equilibrium” in various forms – such as, say, the existence of equilibrium in a perfectly competitive economy – by placing that equilibrium in the context of an optimization problem. In all three categories the assumptions and conditions under which optimization is pursued vary enormously. Thus optimization can be applied in a static (timeless) or dynamic framework; it can determine the choice of a vector, a function, or a path; it may involve a differentiable or only a continuous objective function; and it may result in interior or boundary solutions.

10 Of course, models intended to explain behavior may also be taken to define ideal behavior. But their primary purpose remains explanation. 11 Samuelson [17, p. 251].

124 The role of optimization in economics The extensive appearance of optimization in economics raises two central questions that will be addressed in the remainder of this essay. On the one hand, why has optimization become so central to economic theorizing? On the other, is optimization likely to remain so in the future? To answer these questions requires separate consideration of each of the three categories of optimization applications described above. The former question will be discussed first; the latter will be taken up at the end of this essay. In reference to the question of why optimization has become fundamental in economic analysis, the last of the above categories of applications of optimization may be disposed of rather easily. For the applications that fall in this category are all applications of economical or mathematical expediency. That is, optimization is applied in such cases because it provides a simple, direct, or otherwise appropriate method of demonstrating a desired result. In terms of one of the examples cited above, to prove the existence of equilibrium in a competitive economy generally requires the use of a fixed point theorem. But under certain conditions, that proof can be reduced to the application of a maximization problem, namely, the maximization of a “community utility” function subject to the economy-wide “transformation” function,12 and the existence of the equilibrium falls out of that problem upon invoking the Weierstrass Theorem that a continuous function defined on a compact set always achieves a maximum (and minimum) value on that set. Not only is the proof of existence greatly simplified by going this route, but the proof itself can be illustrated geometrically in an intuitive and suggestive way. A more general proof that equates the competitive equilibrium to a Nash equilibrium in a noncooperative game has been given by Arrow and Debreu [1]. The latter, of course, is defined in relation to a specific optimization problem [1, p. 273]. Consider, now, the first of the above categories in which optimization figures centrally in models that are constructed to explain the behavior of individual economic agents (consumers or firms). These models all attempt to clarify what is happening in the real world. That is, they purport to give relatively accurate and seemingly satisfactory reasons why certain behavior has been or might be observed.13 This, in turn, implies that there must be acceptable accordances between any particular model itself and the reality that is to be explained. For it is the presence of such accordances that permits the investigator to conclude that the

12 See Katzner [13, Sect. 8.4]. The transformation function indicates the maximum amount of a consumption or investment commodity that can be produced by the economy, given the levels of production of all other consumer and investment goods and the available factor supplies. It is implicit in this definition that all production takes place as efficiently as possible. 13 But it does not follow that any such model is capable of immediate concrete application to specific real-world situations. To confuse a model of a person or a firm with, respectively, an actual person or an actual firm, is to commit what Machlup called the “fallacy of misplaced concreteness.” According to Machlup [15, p. 9], the fallacy occurs when theoretical symbols are used “… as though they had a direct, observable, concrete meaning.” What a model does do, of course, is to establish principles and modes of causation that apply only in very broad and general ways.

The role of optimization in economics 125 model is a credible and appropriate reflection of what is actually going on. One way to establish the presence of acceptable accordances is by subjecting the observable parts of the model to empirical tests. Not all parts of all models in economics relate to variables that are observable. Some things cannot be seen. Thus economists expend considerable energy deriving properties in relation to observable variables that follow from assumptions made with respect to unobservable variables. It is the former properties that can be subjected to tests against data that have been recorded in reality. Optimization fits into explanatory models in that it is usually reflective of what is thought to be the driving force behind the behavior of agents in them. One common motive that is often seen as important in directing behavior in reality is self-interest. Self-interest focuses an agent’s attention on the achievement of a goal that, in the eyes of the agent, lies in the interest of that agent. Rational behavior is the manifestation of general goal-oriented (not necessarily self-interested) behavior in a model. In the case of the economist’s theory of consumer behavior, for example, it is assumed that the consumer has fixed preferences among vectors of quantities of commodities, and that these preferences are represented by a utility function (which is, in this case, the objective function) with appropriate properties defined over the space of all nonnegative commodity vectors. (Utility functions and the preferences underlying them are generally not observable.) The consumer is then assumed to be rational in the particular sense that, in making purchases, he always chooses that basket that maximizes utility subject to the limitations imposed by the prices of commodities (assumed to be given) and the amount of money he has to spend (also assumed to be given). It is the optimization, then, that translates the consumer’s unobservable preferences into buying action. Once translated, the characteristics of the consumer’s buying behavior, which are observable and which emerge from the properties of his fixed preferences or utility function through the maximization, can be determined and subjected to empirical tests. Clearly, optimization lies at the very core of this explanation of consumer behavior. It is no less than the representation in the model of what is thought to be the fundamental propellent that determines the consumer’s choice decisions and brings about his purchases. Optimization plays a similar role in most explanatory models in economics. But there are problems relating to the use of optimization in these models. To illustrate them, attention will focus on the model of consumer behavior described above. Also, although they are distinct concepts and may be interpreted in different ways, it will be convenient in what follows to conflate the meaning of the terms “self-interest,” “rationality,” and “optimization” so that the latter two both refer to constrained utility maximization and are taken to be the representation of self-interest in the model. The first difficulty arises with respect to the establishment of acceptable accordances between the model and reality through empirical testing of the observable parts of the model. To conduct such tests, it is necessary to observe the buying behavior of the individual at different times when the prices he faces and/or the amount of money he has to spend have changed. The rate of response of observed quantities purchased to changes in any price or alterations of

126 The role of optimization in economics funds available may be characterized in terms of appropriate partial derivatives, for example, ∂q/∂p or ∂q/∂m, where q, p, and m, refer, respectively, to the quantity of the commodity purchased, its price, and the money available to spend. Or alternatively, the proportional change in quantity purchased, ∂q/q, may be expressed as a ratio of the proportional change in price, ∂p/p, to provide what is referred to as the “price-elasticity of demand,” (∂q/∂p)(p/q). A corresponding expression may be derived for the “funds-available” or “income elasticity of demand.” For any test to be successful, the behavior so observed, described in part, in the language of partial derivatives or elasticities, has to be consistent with the corresponding characteristics predicted by the model, given the fixed preferences of the consumer in it. But the fixed preferences of the model are a depiction of predilections in the mind of the consumer, and the economist who is testing the model has no control over them. Thus any collection of purchase data that contradicts those predicted characteristics, and which, therefore, would cast doubt on the relevance of the model, can always be explained away as a consequence of changing preferences in the consumer’s mind that occurred in the interval between the observation of one data point and another, and that could not be controlled. It follows that the model cannot be falsified and remains viable as a potential explanation of consumer behavior regardless of whether the observations are consistent with it or not, and hence the ability to establish acceptable accordances through empirical testing breaks down. Where then, do the acceptable accordances come from? Elsewhere [10] it has been suggested that these are derived in part from a judgment by the investigator that the assumption content of the model is consistent with the dominant features of the cultural heritage of the particular consumer in question. The argument may be briefly summarized as follows: Mental processes are heavily influenced by the cultural milieu in which they take place [12]. In any culture, some cultural traits are dominant over others. Since the purchases of a consumer are determined through mental processes, it is both reasonable and realistic to abstract from what are thought to be less important cultural characteristics and focus on the dominant traits of the cultural heritage of that individual in making assumptions about what motivates those purchases. In the case of the Western world, the relevant dominant trait is self-interest. That trait, according to Weber [23], was institutionalized as a central tenet of Western culture during the seventeenth century through the dogmas of certain religious texts. Self-interest as a motive remains strong and dominant today – especially in America. It is therefore thought to be appropriate and accurate to explain consumer purchasing behavior in Western societies in terms of constrained utility maximization that reflects self-interest. It follows from this argument, of course, that in societies where optimization is irrelevant or inappropriate as a representation of the existing dominant cultural traits, it should not figure into explanations of consumer buying behavior. This, it has been suggested [11], is the case with Japan. According to Bellah [2], the dominant traits of Japanese culture (these were institutionalized during the Tokugawa era from 1600 to 1868) include loyalty, the necessity to fulfill obligation, and frugality. But self-interest and other motives that might be represented in

The role of optimization in economics 127 terms of an optimization problem are not among them. By previous argument, then, the model of constrained utility maximization cannot be used successfully to explain consumer behavior in Japan. The same is true of most explanatory models that employ optimization techniques when attempting to explain the economic behavior of Japanese agents in other contexts and in general. Additional examples of societies in which deviations from self-interested behavior seem to occur have been found by Henrich et al. [7]. For the most part, modern explanatory economics was developed by Western economists to explain Western economic behavior. It is not surprising, then, that there are not many explanatory models in the economics literature that represent the driving force behind human behavior by something other than an optimization.14 One could still argue, however, that all economic behavior (both Western and non-Western) is explainable in terms of some form of optimization as long as the objective function is appropriately modified from its Western specification to fit the situation at hand. Indeed, a mathematical theorem of this sort has been proved by Katzner [11] within the framework of the theory of choice. But applying optimization techniques as representative of the motivating propellant of human behavior when optimization cannot represent the actual force that brings about that behavior is to produce inaccurate and incorrect explanation. And this can be dangerous not only because it leads to misunderstanding, but also because economists often use their explanatory models as the basis for policy recommendations. Hence the use of optimization models where unwarranted can actually lead to considerable harm. Returning to the original question, then, one may conclude that optimization has become so important in economic explanations of individual behavior not because it has been empirically established that optimization is universally the engine that drives that behavior but, rather, because it is a consequence of the peculiarities of Western culture and the fact that models using optimization were created primarily to explain behavior in that environment. The second difficulty, actually an entire class of difficulties, relating to the application of optimization in explanatory economic models has more to do with the form in which optimization is applied than with the application of optimization itself. For, in the absence of the ability to verify empirically that any particular assumption of optimization is appropriate, and the consequent reliance on the judgment of the investigator as to whether there are the requisite accordances between a model and the reality it is intended to explain, certain aspects of some applications of optimization in economics have come under heavy criticism. These criticisms frequently have to do with whether or not the individual or individuals whose behavior is in question are actually able to know or obtain enough information to be able to optimize in the manner spelled out in the model. Two examples of such criticisms will be detailed here, each drawing attention to a perceived weakness in the utility maximization model of consumer behavior.

14 A few examples have been provided by Katzner [11] appearing as Essay 10 below.

128 The role of optimization in economics Simon [20] has suggested that to be able to maximize utility as described above, the consumer has to have complete knowledge of the commodities that might make up each purchase and complete knowledge of his preferences for all possible purchases. For only then can the true maximizing choice be determined. But in most situations, the number of options is so great and the amount of information so vast, that such knowledge is beyond the capacity of the consumer to secure and manage. Faced with this difficulty, consumers do the best that they can by choosing a vector that “satisfices” or that seems to be “good enough.” Thus, to restore its optimizing character, the model of consumer behavior needs to be modified by introducing certain limits on the numbers of goods and the numbers of different possible quantities of goods that can be purchased, and that have heretofore been absent in typical economic argument. Shackle [18], [19] extends this lack of information to cover much greater territory. Among other things, and remaining in the context of the utilitymaximization model, he is concerned about the time that elapses between the moment a decision to purchase is made and the consumption of that purchase which necessarily occurs later. Since the future is unknown and unknowable, the individual cannot know what his circumstances will be and what he himself will be like at the time of consumption, and hence he cannot know what his preferences will be at that moment. Something bought today may turn out to be despised or useless when the time comes to consume it. And restricting the numbers of goods that can be purchased and their possible quantities is not enough to provide an adequate explanation of purchasing behavior. Nor is the introduction of probabilities in the context of mathematical expectation, the theory of games, or some other formal structure, very satisfactory since the individual does not have enough information at his disposal to determine those probabilities in a meaningful way either. Rather, a complete restructuring of the decision model is required in which nonprobabilistic uncertainty and the consequent ignorance in which the individual decides plays a major role. Shackle [18] proposed such a model in which a concept called “potential surprise” replaces probability and in which optimization occurs in reference to the maximization of what are referred to as “attractiveness” and “decision index” functions.15 Turning to the second category of optimizations, where every element of that category characterizes an ideal that might in some way be considered desirable by a collective or a society, it is clear that the optimizations involved do not themselves need to represent something present in reality. Each model, of course, has to have appropriate associations with the real world with respect to part or all of its structural components. But the optimization itself appears only as the representation of an end towards which policy-makers, say, are striving or that investigators are studying. Because societies are comprised of individuals, such ends or ideals are necessarily amalgams of individual thought forms. Those amalgams are constrained by institutions and, in the process of formation, are

15 An extensive discussion of Shackle’s approach and his decision model appears in Katzner [9].

The role of optimization in economics 129 buffeted by pressures from various sources. The latter include, in part, cultural habits and political forces. Cultural imperatives are clearly significant because, in addition to what has been said of them earlier, they largely determine what is considered to be important by the individuals in the society whose economy is at issue, and are therefore crucial inputs to the crucible out of which society’s ends flow. In American society, for example, where self-interest dominates, and where success as reflected in accumulated wealth and consumption is highly respected, the emergence of the maximization of national income, economic growth, and wealth, as desirable ends is quite natural. With respect to the study of ends by investigators, any end, related to society’s ideals or not, could be employed as the basis for a standard of comparison against which an analyst might contrast actual or potential situations. But, of course, the choice of that end, if it is to reflect the values of the society under investigation, has to be sensitive to that society’s cultural heritage. Regardless, in light of the role that optimization plays in explanatory models, and taking into account both the prevalence in Western societies of cultural predilections that give rise to the employment of optimization, and the fact that much of current economics was developed in the context of and to apply to Western societies, the widespread use of optimization to define desirable ends is not surprising. Thus, in addition to having been the main impetus for the extensive utilization of optimization in explanations of individual economic behavior, Western cultural forces have also been responsible to a considerable degree for the frequent appearance of optimization in the economic characterization of collective or societal ideals. With regard to the future of optimization in economics, it is clear that as long as optimization remains a convenient mathematical tool in, to cite a previously discussed example, establishing the existence of equilibrium in economic models, optimization will continue to be used in such contexts. Optimization is also likely to remain important in explaining behavior where self-interest continues as the primary force generating actual economic behavior. Under such conditions, one would further expect the use of optimization in characterizing ideal situations to flourish as well. Likewise, with respect to cultures in which self-interest or other motives that could be represented by optimization continue to have little or no influence in determining behavior in general, optimization ought not and probably will not figure significantly in explanatory models and in analyzing ideal circumstances. But, of course, cultural shifts over time that cause the institutionalization or deinstitutionalization of motives that can be represented by optimization will alter the prevalence and usefulness of the application of optimization in economics.

References 1. Arrow, K.J., and G. Debreu, “Existence of an Equilibrium for a Competitive Economy,” Econometrica 22 (1954), pp. 265–290. 2. Bellah, R.N., Tokugawa Religion (Glencoe: Free Press, 1957).

130 The role of optimization in economics 3. Cournot, A., Researches into the Mathematical Principles of the Theory of Wealth, N.T. Bacon, trans. (New York: Kelley, 1960). 4. Gossen, H.H., The Laws of Human Relations and the Rules of Human Action Derived Therefrom, R.C. Blitz, trans. (Cambridge: MIT Press, 1983). 5. Grattan-Guinness, I., The Norton History of the Mathematical Sciences (New York: Norton, 1997). 6. Groenewegen, P.D., “Introduction,” Reflections on Political Economy by P. Verri, trans. B. McGilvray and P.D. Groenewegen (Fairfield: Kelley, 1993), pp. vii–xxix. 7. Henrich, J., R. Boyd, S. Bowles, C. Camerer, E. Fehr, H. Gintis, and R. McElreath, “In Search of Homo Economicus: Behavioral Experiments in 15 Small-Scale Societies,” American Economic Review 91, no. 2 (May, 2002), pp. 73–78. 8. Jevons, W.S., The Theory of Political Economy (New York: Kelley, 1965). 9. Katzner, D.W., Time, Ignorance, and Uncertainty in Economic Models (Ann Arbor: University of Michigan Press, 1998). 10. ——, “Western Economics and the Economy of Japan,” Journal of Post Keynesian Economics 21 (1999), pp. 503–521. This volume: Essay 2. 11. ——, “Culture and the Explanation of Choice Behavior,” Theory and Decision 48 (2000), pp. 241–262. This volume: Essay 10. 12. ——, “‘What Are the Questions?’” Journal of Post Keynesian Economics 25 (2002), pp. 51–68. This volume: Essay 3. 13. ——, An Introduction to the Economic Theory of Market Behavior: Microeconomics from a Walrasian Perspective (Cheltenham: Elgar, 2006). 14. Kline, M., Mathematical Thought from Ancient to Modern Times (New York: Oxford University Press, 1972). 15. Machlup, F., “Theories of the Firm: Marginalist, Behavioral, Managerial,” American Economic Review 57 (1967), pp. 1–33. 16. Robertson, R.M., “Mathematical Economics before Cournot,” Journal of Political Economy 57 (1949), pp. 523–536. 17. Samuelson, P.A., “Maximum Principles in Analytical Economics,” American Economic Review 67 (1972), pp. 249–262. 18. Shackle, G.L.S., Decision, Order and Time in Human Affairs, 2nd ed. (Cambridge: Cambridge University Press, 1969). 19. ——, Epistemics and Economics (Cambridge: Cambridge University Press, 1972). 20. Simon, H.A., Administrative Behavior, 2nd ed. (New York: Macmillan, 1957). 21. Verri, P., Reflections on Political Economy, trans. B. McGilvray and P.D. Groenewegen (Fairfield: Kelley, 1993). 22. Walras, L., Elements of Pure Economics, W. Jaffé, trans. (Homewood: Irwin, 1954). 23. Weber, M., The Protestant Ethic and the Spirit of Capitalism, trans. T. Parsons (New York: Scribner, 1958).

8

Economics and the principle of uniformity1

There is a long tradition in economics of explaining economic phenomena through the importation of concepts, structures, methodologies, and philosophies from the physical sciences. Mirowski [10, p. 146] traces the practice as far back as Aristotle who, for example, applied the metaphor of physical motion to understand trade. Subsequent early economic discourse and debate were additionally concerned [10, pp. 9, 186] with the search for a unique substance which could be seen to have properties analogous to corresponding characteristics in physical science. Those analogical properties associated with Euclidean distance were among the most important since, once a substance with such properties was found, it could be measured and taken to represent value. This search culminated in the labor theory of value that emerged early in the nineteenth century during the classical period. Still later, neoclassical economists introduced ideas and relations that were obtained, respectively, by relabeling notions taken from, and changing the names of variables in certain equations of, nineteenth-century physics. Thus, for example, Fisher [4, p. 85] associates, in part, particles with individuals, and energy with utility. And, in the hands of Jevons [6, pp. 95–106], the physics equation of the lever where, at equilibrium, the weights of objects at each end are inversely proportional to their respective distances to the fulcrum, became the economic equation of exchange in which (at equilibrium) the marginal utilities of two goods are inversely proportional to their respective quantities traded (or, equivalently, directly proportional to their trading prices). However, the success of these appropriations from the physical sciences in making sense of economic phenomena is open to debate. According to Mirowski [10, pp. 185–192], the substance representation of value foundered with the realization that the geometry of the real physical world was not exactly “selfevident” Euclidean geometry, and hence that the analogy of value to distance, and the measurement of value based on that analogy, were murkier than had been previously thought. Mirowski [10, p. 9] further suggests that, since it fails

1 Economia Politica 17 (2000), pp. 237–252. Reprinted with permission. The author would like to thank Douglas Vickers for suggesting this essay and helping with its execution. Thanks are also due to Jonathan Barrett for a number of useful comments.

132 Economics and the principle of uniformity to contain the analogue of a crucial part of the physics counterpart from which it is drawn, neoclassical economics, that is the dominant economic paradigm of the twentieth century, is wanting in terms of the structural standards to which the physics counterpart adheres.2 In particular, it does not achieve a coherence similar to that counterpart because the analogical neoclassical construction ignores the conservation of energy principle lying at the counterpart’s heart. The purpose of this essay is to focus on another doctrine in the physical sciences, namely, uniformitarianism or the principle of uniformity, and to show how its emergence in economics has also left economic analysis wanting. In this case, however, it has led to the exclusion of certain facts of economic reality that are worthy of consideration by economists.

I Beginning in the late eighteenth century, uniformitarianism arose and developed in geological thought in opposition to the doctrine of catastrophism. The latter explained the discontinuities observed in the record of ancient life as a consequence of recurrent catastrophes that were brief, convulsive, and lethal. Sudden upheavals, floods, and climate changes not only altered the physical surface of the earth in cataclysmic ways, but transformed its flora and fauna as well. Between catastrophic episodes, the normal order of nature was restored, albeit with modified collections of species and populations of plants and animals. Moreover, these catastrophic interruptions of the normal course of nature were frequently interpreted as divine intrusions in which a supernatural being directed, shaped, and generally maintained control of his creations. Specifically, according to one particular approach, after each catastrophe, new and higher forms of life replaced some of the older and lower forms that had been destroyed, culminating, according to divine plan, in the eventual emergence of man. Against this view, it was argued that the so-called catastrophic episodes were not abrupt, miraculous interventions, but actually occurred quite naturally and gradually at constant rates over large spans of time. This led to the postulation of the principle of uniformity (uniformitarianism) that (i) natural processes and known causes have always operated at the rates at which they currently function and that (ii) the laws of nature or the properties of matter and energy are invariant through time and across space. Thus, as a partial consequence, knowledge of the present provides knowledge of the past, and vice versa. Knowledge of the future is also implied. It was in this context that Darwin developed his mechanism of natural selection explaining the evolution of the species and the origins of mankind as the result of a long, natural, and gradual occurrence. Now it is evident that uniformitarianism is neither necessary nor sufficient for natural selection.

2 Mirowski [10, pp. 397–398] believes that, to avoid arbitrariness, all economic science must, of necessity, be firmly grounded in analogies to physical science. But, of course, such analogies could also come from generally accepted cultural norms and mores, or religious tenets.

Economics and the principle of uniformity 133 For although that mechanism is itself taken to be independent of time, the rates of change it identifies in species could vary with time. But in any case, uniformitarianism reduced what was considered under catastrophism to be divine intervention to statements that were seen to be subject to scientific investigation. Actually, the two parts of the principle of uniformity indicated above can be separated into distinct, and (at least in the case of one of them) more widely applicable categories that extend beyond the geological framework in which they were introduced (Gould [5]). And each part or category by itself can be (and has been) identified as a separate formulation of the principle of uniformity. Thus, on the one hand, substantive uniformitarianism, or the general, historically descriptive statement that rates of change are invariant over time, is the explicit application of the first half of the uniformity principle set out above as a scientifically testable proposition. On the other, the general assertion that natural laws are constant across time and space may be called methodological uniformitarianism. That is, methodological uniformitarianism employs the second half of the originally presented principle of uniformity as an element of the methodology by which history is investigated, and it directs, in part, that causes that are not in operation at present should be avoided in scientific inquiry. Clearly, substantive uniformitarianism is not necessarily implied when scholarly investigation proceeds in accordance with methodological uniformitarianism. While the former is mostly thought to be outmoded insofar as geological science is concerned, having more recently been refuted both in theory and by newer scientific data,3 it still has uses in deriving first approximations of the times required for the creation of geological formations. And Gould considers the latter as having been absorbed into the fundamental principles that underlie all modern empirical science. Both forms of the uniformity principle have appeared in economics. First, substantive uniformitarianism is one of the main stays of that common version of long-run analysis in which constant rates of change over time are assumed. Models falling in this category often reflect the idea that the time path along which the object of inquiry, e.g. the economy, is thought to be moving, approximates or converges to a stationary or a steadily growing state. Of course, at stationarity, all rates of change vanish identically for all time; in a steady state, all rates of change in the form of growth rates are positive constants for all time. Such models usually specify a differential equation or a system of differential equations that characterize all motion through time. The differential equation or system itself is fixed and taken to be time-invariant, and the rates of change in the model are the same regardless of whether the model is applied to a time frame that is past, present, future, or some combination thereof. Second, methodological uniformitarianism arises in the frequent acceptance of the idea that economic assumptions and laws, as they reflect the properties of human and institutional behavior, are invariant through time and across space. Thus, for example, it is often thought that people living in

3 For example, Gould [5, p. 226].

134 Economics and the principle of uniformity all ages (past, present, and future) and in every culture always make decisions on the basis of utility maximization, constrained or otherwise. The utility functions that are maximized may vary across persons and even, in some respects, over time, and the observable properties of action derived from such maximizations that are independent of the particular utility functions employed are taken to be scientific laws or descriptions of visible behavior in every case. The primary issues to be raised in the remainder of this essay concern the appropriateness of these kinds of assumptions and the laws derived from them for explanations of economic reality, and the elements of that reality that are lost by invoking such assumptions and laws in the construction of economic models.

II Start with the observation that the activities relating to science in general may be seen as falling into at least four categories: explanation, prediction, general policy formulation, and specific policy decision-making and implementation. Explanation involves the development of such things as concepts, assumptions, reasoning, laws, models, and theories. It employs a specific methodology and rests on an underlying epistemology. Explanation is always explanation of something. The latter something has either been observed in the real world or is presumed to be there. And explanation is often tested to determine the extent to which it actually explains the something that it is intended to explain. Prediction is an assertion put forward that is derived from things that are already known. The things already known are often identified in terms of what are referred to as tendencies or trends, and prediction is frequently obtained by projecting those tendencies or trends into the future. General policy formulation involves the setting of broad goals that scientific activity should pursue. Depending on the context in question, these goals may be set by an individual scholar (to direct his own work), a group of scholars working together, a firm (to provide direction for its research unit), or society at large (e.g. its government). Often considerable thought and discussion is required and numerous persons are enmeshed in their identification. Finally, specific policy decision-making and implementation refers to the decisions by appropriate individuals or groups that have to be made to achieve the goals set, and the carrying out of those decisions to implement the general policy. Examples of explanation and prediction are well known and need no further illustration here. To illustrate the latter two categories with respect to the physical sciences, consider the scientific activity during the 1960s that resulted in man walking on the moon. After the general goal of a moon landing was set by the US government, many decisions had to be made and many activities had to be undertaken to achieve it. The US government agency given the responsibility for doing this was the National Aeronautics and Space Administration (NASA). In pursuing the moon landing, not only were previously developed explanations of the relevant physical phenomena employed, but new explanations of previously unexplained phenomena and new predictions of certain other phenomena had to be devised. General policy formulation and specific policy decision-making and

Economics and the principle of uniformity 135 implementation outside of the physical sciences follow a similar pattern. Examples are well known and not discussed. One of the fundamental elements of explanation mentioned above, which is relevant for all four categories of activity relating to science, is that of laws. A law is a generalization about cause and effect that results from successful inquiry, that is, an inquiry that has been empirically confirmed or, alternatively, has not been empirically falsified. Laws define the tendencies and trends upon which predictions are often based and are a major portion of the analytical building blocks required for policy implementation. Laws can also provide the background within which general policy is formulated. With respect to the physical sciences, the notion of law usually carries additional properties. In particular, such laws are understood to be universal statements that are independent of any specific object to which they apply, and are also independent of time and place. It is clear, then, that the physical-science concept of law is grounded in the thought forms of methodological uniformitarianism. Indeed, it is the basis for Gould’s assertion cited earlier that methodological uniformitarianism has been absorbed in the principles on which modern science rests. And since many of these principles of modern science have been imported into economics, much of current economic analysis (as indicated earlier) is also expressed in the thought forms of methodological uniformitarianism. It has already been pointed out that in addition to methodological uniformitarianism, substantive uniformitarianism, or the assumption of constant rates of change, has also found a place in economics. However, it is possible to argue that reliance on the principle of uniformity in economics, in either its substantive or methodological aspects, and either in part or in its entirety, is too restrictive for explanations of human behavior. In dealing with inanimate objects that neither think nor feel, as is the case in the physical sciences, an approach derived from uniformitarianism, to the degree that it produces results that are consistent with observation, leads to the capturing of physical properties with sufficient accuracy to yield important and meaningful explanation. But when attempting to understand the behavior of sentient beings, the same approach prevents account being taken of prominent facets of existence that relate, in considerable part, precisely to thoughts and feelings (which also include, of course, such things as reactions and motivations). And this occurs to such an extent that explanation based on the principle-of-uniformity approach frequently departs from reality sufficiently to lose both its relevance and significance. The argument is made in detail in the next three sections.

III Consider methodological uniformitarianism and the assumed constancy of laws across space first. Before taking up the matter in general, focus for a moment on the possibility of deriving laws of economic behavior from utility maximization, constrained or otherwise. There are two kinds of laws that are relevant here. On the one hand are the laws relating to individual action that are generated

136 Economics and the principle of uniformity directly by utility maximization; on the other are the laws of aggregate activity that are obtained by combining the laws of individual action across individuals. The particular application of methodological uniformitarianism that, in this case will be challenged initially, is the statement that all observable behavior of all individuals everywhere is governed by laws of individual action derived from utility maximization because all behavior of all individuals everywhere can be meaningfully understood as if each person were a utility maximizer.4 (Questions concerning the laws of aggregate activity that emerge from such laws of individual action will be raised later on.) Of course, those that hold this view would probably not assert that all people actually make all decisions on the basis of utility maximization. But they are likely to claim that, as a rule, the utility maximization assumption is, for theoretical purposes, a reasonable approximation of the actual motivation or motivations underlying all individual economic, and sometimes even noneconomic, behavior.5 Now the justification often given for this claim, or for assuming that utility maximization is universally applicable for explaining all individual action, is that, since the behavior at issue is observable and since the patterns of individual behavior drawn from (constrained) utility maximization can also be seen, the assumption of utility maximization itself can be subjected to indirect empirical tests. If a given observed pattern of behavior is inconsistent with that derived from theoretical utility maximization then, in that case, the assumption of utility maximization would be falsified or refuted, and would have to be discarded. And one refutation anywhere would mean that the universal application of utility maximization would have to be rejected. Thus, until utility maximization is convincingly disproved somewhere by empirical tests in this way, it stands as a permissible and plausible assumption in the elucidation of individual economic behavior everywhere. But as Caldwell [2, pp. 156–157] has pointed out, such tests have little practical merit. A person’s preferences, after all, are a product of his mind, and no economics laboratory is ever able to control them from one decision or behavior to the next, preventing them from conscious or subconscious transformation. Thus, since observations of an individual’s behavior necessarily take place sequentially through time, any visible deviations from behavioral patterns derived from utility maximization can always be explained away as a consequence of changing preferences from one observation to the next. What might have been thought to have been a falsifiable assumption, then, is actually not, and the grounds described above for accepting utility maximization universally as the driving force behind all individual economic action fall flat. This leaves open the question of how the assumption of utility maximization is to be justified anywhere as a central tenet in explanations of individual behavior. In another essay [9] it has been suggested that a credible justification for

4 A more complete statement of the argument may be found in Katzner [9] which appears in this volume as Essay 2. 5 For example, Becker [1, p. 14].

Economics and the principle of uniformity 137 at least Western economic behavior lies in the culture of Western society. Utility maximization is a reasonable assumption because it is an expression of what Westerners see as the main or most significant force in making decisions in their private, everyday lives. That force, which, for present purposes, may be taken as a manifestation of self-interest, is thought generally to govern their own behavior, and it is invoked in understanding the behavior of others. Of course, the pursuit of self-interest has become institutionalized and is now deeply ingrained in Western culture, going as far back as the end of the seventeenth century.6 And the fact that it figures so prominently there as an imperative is what leads people to interpret what they see in others and what they do themselves in its terms. It is therefore quite natural for a Western economist to be comfortable with the assumption that, in theory, utility maximization is the fundamental determinant of individual behavior. But, by the same token, it follows that the assumption of utility maximization may not be appropriate in explaining behaviors that arise in non-Western cultures. The most significant force driving human action in those cultures may be something other than self-interest. In that case, different models of behavior that do not employ the assumption of utility maximization are required. Such an argument has been made, for example, in the case of Japan.7 Of course, it is always possible for the same laws of individual action to be derived from different assumption sets. But, in the context of self-interest, since the laws in question usually both imply the existence of models of utility maximization that “explain” them, and are, themselves, mainly derived from utility-maximization-type models, it cannot, as a matter of course, be assumed that these laws are relevant in countries like Japan. Thus applications of methodological uniformitarianism in the guise of invariance of the properties of economic behavior derived from utility maximization across space would likely mislead economists, in at least some important situations, to spurious representations and explanations of economic reality. The second application of methodological uniformitarianism in relation to utility maximization referred to above is the assertion that the laws of aggregate activity derived from laws of individual action are invariant across space. This, too, is open to serious doubt. For in spite of the fact that aggregate laws may be more stable over space because they tend to average out individual variation, they still emerge from, and depend on, laws of individual action that, as has just been argued, tend to modify with changes in culture. Hence, to the extent that cultural divergence across economies is significant enough to produce deviations in primary motivating forces, the laws of aggregate economic activity would be expected to modulate, and the application of methodological uniformitarianism as a universal principal is therefore unlikely to hold up. Returning to the issue of whether it is possible in general to have laws describing economic behavior that are invariant over space, the preceding argument indicates that the motivations behind economic decisions, and hence economic behavior

6 See Weber [14]. 7 Katzner [9] (this volume, Essay 2).

138 Economics and the principle of uniformity itself, are dependent on the cultural environment in which they are set. Now culture has to do with (i) meanings and orientations out of which perceptions of reality emerge, and (ii) standards for interpretations, judgments, and action. And it is well known that with different cultural backgrounds, different people can have different perceptions of the same thing along with different interpretations of it. They can also make different judgments relating to it, and take different actions in response to it. Thus, not only can the motivations that drive economic activity vary from culture to culture, but so do the knowledge and nature of the expectations that inform the decisions out of which that behavior arises. Even perceptions relating to the characteristics of individual decision options, and to the number of possible options available for choice, can diverge. It follows that if relevance to economic reality is to be maintained, culture must necessarily influence the assumptions that are invoked by economists to explain economic behavior. To the degree that these assumptions are the principal determinants of laws of economic behavior, the latter cannot be assumed to be constant from one culture to the next. As a rule, then, the economist cannot presume that the laws of economic behavior, such as they are, are constant across space.8 Invoking this aspect of methodological uniformitarianism is both unwarranted and likely to result, in certain cases, in error.

IV Attention now turns to the remaining forms, and their possible application in economics, of the principle of uniformity. The contrast between the arguments which, in one instance, purport to justify, and in the other attempt to establish the impropriety of, the use in economics of both methodological uniformitarianism with respect to the constancy of laws through time and substantive uniformitarianism is closely connected to the distinction between logical and historical time. That distinction, in turn, may be related to the approach taken towards the ontological status of mankind. Generally speaking, there are two different interpretations of human existence. On the one hand, from the vantage point that may be referred to as philosophical materialism, individuals may be perceived as being totally constituted by the material or physical pieces, such as atoms and cells, of which they are made up. If enough were known about the properties of these material pieces (including the ways in which they interact with other material pieces both within and outside of the individual), then, from that point of view, a person’s behavior in any situation could be determined with a high degree of accuracy from a careful examination of his material characteristics and his environment. On the other hand, it may be thought that human beings have cognitive capacities that go beyond, and are not solely determined by, their material parts. This implies

8 Although it might be possible to develop laws that take cultural variation into account, such laws, because they encompass more than just economic phenomena, would extend beyond the conventional limits of economic inquiry. Laws of this sort are therefore irrelevant to the present argument.

Economics and the principle of uniformity 139 that, in relation to the world in which human beings live and act, it is only the experiencing of things and events that can interact with an individual’s extramaterial cognitive capacities, since the material pieces cannot go that far. Of course, due to the individual’s continually changing knowledge and circumstances, the interactions with respect to things and events can happen in any particular way only once, only when the individual makes intellectual contact with them, and only in the order in which those contacts actually occur. All mental processes, regardless of whether they relate to material or nonmaterial elements, together with whatever knowledge has been accumulated, necessarily contribute to the making of an individual’s decisions. It follows that experience and its orderly interaction with the extra-material capacities become of major significance in the generation of human behavior. From the perspective of philosophical materialism, since experience in relation to extra-material cognitive capacities is irrelevant to the determination of human action, time may be conceived as that which merely sequences events. Analytically, then, time is simply a serially dated variable. Events are contemplated in the order of their intertemporal occurrence without regard to the manner in which those events are actually and extra-materially experienced as past events, present events, and future events. All that matters is that past events come before present events, and present events come before future events. The notion of time emerging is often referred to as logical time. On such a view, individual behavior is reducible to explanation by laws that govern the behavior of matter. And methodological uniformitarianism implies that these latter laws, and hence those depicting human behavior at both individual and aggregative levels, are constant through time. Adopting philosophical materialism, then, this aspect of methodological uniformitarianism enters economics attached to the philosophical-materialism relation between human behavior and physical laws. Alternatively, methodological uniformitarianism as the constancy of laws over time can simply be assumed to apply to the economic world without any reference to philosophical materialism at all. In both instances, since it will subsequently be argued that uniformitarianism (in either of its forms) is inconsistent with what will be called “historical” time, methodological uniformitarianism will henceforth, for reasons of convenience, be identified with logical time. Similarly, substantive uniformitarianism, although generally discredited in the physical sciences as suggested earlier, is relevant in economics as it applies in the context of long-run analysis. Convenience will dictate that it, too, be linked with logical time. Thus, with respect to either methodological or substantive uniformitarianism, the one or more differential equations that might be employed to describe dynamic economic behavior can be stopped at any point and restarted at any initial condition as the investigator moves his analytical focus back and forth through the logical time that engages his subject. But with respect to the alternative, previously described, non-philosophicalmaterialism approach to the onotolgical status of man, where the impact of experience on extra-material cognitive capacities is necessarily an important ingredient in generating individual behavior, the relevant concept of time is historical time. By that is meant time that is actually experienced by human

140 Economics and the principle of uniformity beings, in which each moment in history is unique (due to the continually changing knowledge and circumstances of every individual noted above), and in which time reversals (such as the restarting of a differential equation to analyze the recurrence of the same event) are not possible. Such reversals cannot be undertaken because the restarting of an explanatory system would imply that the individuals under investigation were able to relive moments of historical time that have already passed, without their having undergone any changes as a consequence of the experiences gained during the initial running of the system. The changes in individuals that must be taken into account as time passes include not only alterations in knowledge or epistemic statuses, but also modifications in economic endowments and transformations in perceptions of external environments and the possibilities of action taken by others. (It follows that an investigator who wished, for any reason, to re-analyze a previous event occurring in its original circumstances may find that his intervening personal experience has caused him to fundamentally revise his analysis.) Because experience substantively changes such things beyond what can be accounted for by pre-existing knowledge alone, it is clear that the laws directing human behavior, both individual and aggregative, cannot generally be regarded as constant through time. As a consequence, methodological uniformitarianism with its assumption of the constancy of laws over time, is, from this vantage point, irrelevant and inappropriate as an analytical import into economics. A similar conclusion applies to substantive uniformitarianism since its assumption would prevent the effects of experience from modifying rates of change over the long run. It should be emphasized that philosophical materialism is neither necessary nor sufficient for uniformitarianism in either or both of its forms. Moreover, a philosophical materialist can still recognize the importance of the passage of time through history by acknowledging the presence and significance of unforseen, stochastic events that occur from time to time. But this is not the same thing as accepting historical time as defined here because the philosophical materialist would continue to reject the role of experience in relation to extra-material cognitive capacities that lies at its heart. Of course, someone who does not admit to philosophical materialism may still embrace logical time. Finally, although it has been argued that uniformitarianism (regardless of its form) and historical time cannot be upheld simultaneously, rejection of uniformitarianism does not preclude subscribing to either logical or historical time. Keeping these relationships in mind, previous discussion has set out two general approaches in economics that relate, in part, to the conceptualization of time employed – one grounded in logical time and the principal of uniformity (in either or both forms), and the other based on historical time in the absence of that principle. The remainder of this essay focuses on the contrasts between them, with particular emphasis on the inadequacy of the former. A number of economists, including Shackle [11], [12, Ch. 7], and Vickers [13] adhere to the non-uniformitarianism, historical-time view. In addition, they argue further that the historical time environment in which human beings (both the objects of study as well as the investigator himself ) live rules out the possibility

Economics and the principle of uniformity 141 of their ever having knowledge of future human behavior. Only activities that have occurred in the past or are currently unfolding can (however imperfectly) be known. The actions of individuals that are yet to come cannot. And not only is the future unknown to humans; it is also unknowable. This absence of knowledge of the future precludes the specification by anyone of all probabilities of future events, subjective or otherwise, that relate to human behavior. Even a complete demarcation of the collection of possible events itself, which would be necessary for any application of the probability calculus, cannot be accomplished. (The unavailability of the probability calculus will be taken up in greater detail in the next section.) An individual, for example, being a unique person with unique knowledge at each moment of time, cannot say today what action he will take tomorrow because he does not know (even probabilistically) the conditions that will exist when tomorrow arrives, nor does he know how he himself will change in the interim. More generally, it is not possible to know in advance if any structure in place one day will still exist in the same form, or even exist at all, the next. For decisions and activities that are taken today create the structures, unknown today, that will be in place tomorrow. Hence the individual does not stand independently of markets and market forces and activities. Rather, he is continually formed and changed by them. (This, of course, provides another reason why, from the historical-time vantage point, long-run analysis that reflects substantive uniformitarianism has no place in economics.) Thus, for example, before the collapse of the Soviet Union (which, from the perspective of catastrophism, could be viewed as a catastrophe), an astute observer might have been able to determine that that country was in serious trouble. He might even have been able to spin various scenarios depicting possible paths that the Soviet Union might have followed. But he would not have been able to predict, in the usual scientific meaning of the term, its demise. In a very real sense, this was a novel or unanticipatable event. Generally, then, with such pervasive uncertainty about the future, prediction is neither tenable nor possible. Its untenability and impossibility arise from the general instability of structures over time and from the fact that actions in the present necessarily create an unknowable future.9 Indeed, structures that might form the basis of such predictions have been likened by Shackle [12, p. 42] to the images in a kaleidoscope in their ability to dissolve “at a touch.” Clearly, “kaleidic change” and novelty would generate the catastrophes upon which attention would focus were catastrophism to be taken as one of the basic organizing principles of economic inquiry. But, as will be hinted at below (and as has been discussed at length elsewhere10 ), that does not mean that choice in economics is not meaningful and significant, or that nothing useful and enlightening can be said about it.

9 In spite of the impossibility of prediction, economic behavior is frequently based on the assumption that certain established conventions will continue to impart a quasi-stability to the economic environment. See, for example, Vickers [13, pp. 157–160]. 10 For example, Shackle [11], and Katzner [7], [8, Ch. 4].

142 Economics and the principle of uniformity Not surprisingly, however, the implications for the analysis of the problem of choice in this context are rather striking. Choice can no longer involve picking (as would be permitted under methodological uniformitarianism) from a complete collection of outcomes that are, and whose consequences are, fully or probabilistically known. For in the non-uniformitarianism, historical-time setting discussed here, such a choice situation cannot exist. No one ever has sufficient knowledge for it to arise. Instead, and reflecting more accurately actual human experience, choice now requires selecting from a possibly incomplete list of imagined outcomes that are genuinely (nonprobabilistically) uncertain and whose consequences are unknowable.11 Under such conditions, moreover, where personal perceptions, interpretations, and judgments become of paramount importance, the role of culture in determining choice behavior clearly achieves a new level of significance, well beyond that which might possibly have at first been inferred from previous discussion. Such matters are brushed aside and neglected by adopting logical time and the principle of uniformity as the basis for economic analysis. In that case, knowledge of the future is assumed to be knowable, at least probabilistically, as an extension of the past and present through the precepts of methodological uniformitarianism. Individuals make decisions and take action in the light of that knowledge, and their behavior is thus predictable. But since prediction is often based on the extrapolation of trends and their projection into the future, and since there are “gaps” in knowledge making it hard to foresee changes in trends, the application of uniformitarianism in either form to economic behavior does not produce results as reliable as in the physical sciences. In any event, it is clear that the two approaches to economics discussed here require the adoption of distinct assumptions and structures in efforts to explain economic reality. Indeed, in each case, the questions that are asked and the issues that are raised differ markedly from each other as well. In particular, and as has just been suggested, there are aspects of reality relevant to economic analysis that tend to be avoided and ignored along the logical-time, principle-ofuniformity route. Many of these omissions fall into one or more of the following three overlapping categories. First, economic decisions are often made in the midst of such ignorance of past and present circumstances and such ignorance about the likelihood of possible decision outcomes that meaningful estimates of probabilities are not available. Second, this is especially true with respect to future possible outcomes and events that are unknowable because they are of a kind that have never been observed before, or because they are subject to unanticipatable kaleidic change. And third, the evolutionary development of economic history is replete with newness that emerges unpredictably from current events and structures. Rather than following prescribed patterns, history is, as a rule, created period by period. The logical-time, principle-of-uniformity approach hides human ignorance with a frequently unwarranted assumption of

11 For example, Shackle [12, pp. 354–366].

Economics and the principle of uniformity 143 the knowledge of probability and, in addition, altogether disregards the possibility of newness in outcomes and events. The latter newness is then acknowledged somewhat lamely after the fact by shifting parameters and modifying structures in the models employed for purposes of explanation. In this way, significant aspects of reality are removed from consideration. While the alternative, nonuniformitarianism, historical-time perspective can, of course, do no better in knowing what cannot be known, its analytical structure is still able to account for the possible emergence of the unknown. And in such a manner it incorporates and deals directly with the realities of ignorance of the future and the uncertainty arising from it. To summarize thus far, the argument against the use of uniformitarianism as the basis for economic analysis has two parts. First, economic laws relating to both individual and aggregate action describe the outcomes of economic behavior. Those outcomes derive, in part, from perceptions, knowledge, expectations, and motivations. But these latter elements, in turn, depend on culture. Since culture varies from country to country and region to region, economic laws at both levels cannot be expected to be invariant across space. Second, because the future is unknowable, and because future happenings fundamentally alter reality in ways that cannot be anticipated, even probabilistically, from what has gone before, economic laws and rates of change applicable in the past need not be relevant to future economic behavior. Thus, in general, both individual and aggregate economic laws are not invariant, and economic rates of change are not constant, over time.

V Previous discussion has suggested that in adopting the non-uniformitarianism, historical-time framework described in Section IV, the notion of probability as it relates to future events has to be discarded because it requires knowledge that is not available. It is worth pausing for a moment to expand on this point and to link probability directly to methodological uniformitarianism and logical time. The unavailability of knowledge, of course, applies to both frequency probability and subjective probability. Now frequency probability emerges out of repeated trials while subjective probability derives from judgments based on knowledge. It is the frequency interpretation of probability that is of interest here. A precise definition would be something like the following. The (frequency) probability of an event (a set of outcomes) is the relative frequency that an outcome in that event is observed in an arbitrarily large number of trials, under constant conditions across both space and time. Therefore to be able to obtain frequency probabilities necessitates replication of the same circumstances over and over again. Invoking logical time and methodological uniformitarianism allows this requirement to be met at least on a conceptual level. For if time is logical, and laws are constant through space and time, then it is possible to conceive of repeated replication of the same circumstance as time passes and from place to place. Thus hypothetical frequencies can be postulated. To the extent that, through an

144 Economics and the principle of uniformity appropriate process of abstraction, such replication is taken to have been observed in reality, empirical frequencies can also be calculated. In particular, and under logical time and methodological uniformitarianism assumptions, if one replication immediately follows the next across time, a statistical average of the future values of a probability frequency variable or, for that matter, any variable observable through time, cannot systematically differ from averages of values of the same variable reckoned over past dates. Davidson [3, p. 132] has referred to this kind of environment as ergodic. Regardless, it is legitimate and proper, under the stringent conditions postulated, to apply such probability frequencies to future events. But clearly, under conditions of historical time, replication of the same circumstance even once is not possible even conceptually. Each moment or period of time, as has been suggested above, arrives with its own unique history, circumstances, and structures, including institutions and individual perceptions and interpretations of what has gone before and what might yet come. There has been no other moment like this in the past, and no moment like it can ever occur again.12 Of course, nothing prevents an individual from thinking, under certain conditions, that he finds himself in the same circumstance over and over again. But this is a highly specialized case that can only be “confirmed” with hindsight. Indeed, if the individual who thinks circumstances have repeated themselves were willing and able to examine the situation closely enough, he would be forced to change his mind. In addition, with sufficient cultural variation, a parallel argument would imply a similar impossibility of replication across space. Such situations would not arise, of course, if the investigator were to rely completely on the assumptions of uniformitarianism and logical time, and on the legitimacy of replication over space and time to which that may give rise. The foregoing discussion of this and the two preceding sections has important implications for both the formulation of general policy and specific policy decision-making and implementation. For in the absence of methodological uniformitarianism, in which case laws are generally space-dependent, a policy that has been successful in achieving a particular goal in one cultural environment might not work in another. In addition, laws that are independent of time and that hence can be predictively extrapolated into the future, are unavailable. That is, unlike the situation with respect to the physical sciences, there are no time-independent laws on the basis of which policy can be formulated and policy decisions to implement it can be made. It is therefore impossible to tell if a potential general policy would, at the time at which it might be implemented, be consistent with the economic forces then at work, and, were it to be implemented, if the specific decisions that might be made to carry it out could actually be capable of doing their job. Moreover, there

12 A similar argument applies to frequency interpretations of such concepts as “capacities” that are derived from probability by weakening its properties. Non-frequency interpretations break down for other reasons. See Katzner [8, pp. 71–74].

Economics and the principle of uniformity 145 is no equilibrium position towards which the economy might be headed that could serve as a guide. As pointed out earlier, economic decisions are history-creating, determining in a beforehand-unknowable way the institutions or structures that lie ahead. It is within these unknowable structures that present decisions resolve themselves into unknowable-ahead-of-time future economic positions. And, of course, the decisions made and institutions determined have effects at both the micro and macro levels. All that policy-makers (and consumers and firms too!) can do under such conditions is to work out scenarios of possible future happenings and make judgments of the proper or best course to follow. Vickers [13, pp. 203–204] has referred to that course in the complex of historical time as the decision-maker’s “best next move.” Finally, it should also be pointed out that in a non-uniformitarianism, historicaltime world, the absence of replication and the impossibility of knowing or predicting the future does not reduce the importance of gathering historical data. For historical data still provides the only basis for imagining the possibilities that the future might hold. And it is the evaluation of such imagined possibilities that leads to choice or the making of decisions.13 This is clearly so for both the individual attempting to determine his best next move as well as the policy-maker in the process of implementing policy at either the microeconomic or macroeconomic level. Although the future cannot be known in advance, this does not preclude the undertaking of carefully thought out action.

References 1. Becker, G.S., The Economic Approach to Human Behavior (Chicago: University of Chicago Press, 1976). 2. Caldwell, B.J., Beyond Positivism: Economic Methodology in the Twentieth Century (London: Allen & Unwin, 1982). 3. Davidson, P., “Is Probability Theory Relevant for Uncertainty? A Post Keynesian Perspective,” Journal of Economic Perspectives 5 (1991), pp. 129–143. 4. Fisher, I., Mathematical Investigations in the Theory of Value and Prices (New York: Kelley, 1965). 5. Gould, S.J., “Is Uniformitarianism Necessary?” American Journal of Science 263 (1965), pp. 223–228. 6. Jevons, W.S., The Theory of Political Economy, 5th ed. (New York: Kelley, 1965). 7. Katzner, D.W., “The Shackle-Vickers Approach to Decision-Making in Ignorance,” Journal of Post Keynesian Economics 12 (1989–90), pp. 237–259. 8. ——, Time, Ignorance, and Uncertainty in Economic Models (Ann Arbor: University of Michigan Press, 1998). 9. ——, “Western Economics and the Economy of Japan,” Journal of Post Keynesian Economics, 21 (1999), pp. 503–521. This volume: Essay 2. 10. Mirowski, P., More Heat than Light (Cambridge: Cambridge University Press, 1989).

13 For example, Shackle [11], and Katzner [7], [8, Ch. 4].

146 Economics and the principle of uniformity 11. Shackle, G.L.S., Decision, Order and Time in Human Affairs, 2nd ed. (Cambridge: Cambridge University Press, 1969). 12. ——, Epistemics and Economics (Cambridge: Cambridge University Press, 1972). 13. Vickers, D., Economics and the Antagonism of Time (Ann Arbor: University of Michigan Press, 1994). 14. Weber, M., The Protestant Ethic and the Spirit of Capitalism, T. Parsons, trans. (New York: Scribner, 1958).

9

Cultural variation in the theory of the firm1

When explaining firm behavior, economists naturally focus on what seems to be the most important features of economic reality to secure the building blocks for their models. In the case of Western firms, this has led to the postulation of such elements as production functions with their attendant properties, and the pursuit of maximization by firms as the motivating force that drives the determination of their behaviors. (The use of the term “maximization” here is intended to be quite general, covering both the long and short runs, and including the maximization or minimization of an appropriate objective function, with or without constraint, under conditions of either certainty or uncertainty.) Since it is not possible, given the limits of the experimental laboratory that economists have at their disposal, to confirm the veracity of such assumptions or even their implications (Caldwell [6, pp. 156–157], and Katzner [16, pp. 8–9]), they must, of necessity, be accepted on faith. That faith, however, is still justifiable in terms of the cultural heritage of the particular subjects under investigation. The reason why culture plays a role here is that the forces that motivate real people are determined, to a considerable extent, by their cultural backgrounds (Katzner [17]). And it is real people who make the decisions that guide the behavior of real firms. Returning to the case of the West, one of the most important cultural components, at least insofar as economic analysis is concerned, is self-interest. That component was institutionalized at the center of certain Western cultures in the seventeenth century and spread to all of them, to one degree or another, shortly thereafter (Weber [20]). Indeed, it has become so fundamental in Western life and thought that self-interest, in the form of postulates of rationality or maximization,2 is, today, quite naturally and readily accepted by many economists in explanations

1 Journal of Post Keynesian Economics 28, no. 2 (Winter, 2005–2006), pp. 277–293. © M.E. Sharpe, Inc. Reprinted with permission. The author gratefully acknowledges the help of Peter Skott and Douglas Vickers in the preparation of this essay. 2 In general, of course, rationality in decision-making can be said to exist in the absence of selfinterest and maximization. But for purposes of present discussion, it will be convenient to refer to action as derived from self-interest, and only that action, as being both rational and satisfying postulates of rationality or maximization.

148 Cultural variation in the theory of the firm of Western economic reality as the most important single force driving individual behavior (Katzner [14]).3 It follows from this that in places where self-interest is not a significant component of cultural backgrounds, postulates of rationality have no place in explanations of firm behavior because the real people who make the decisions do not think in those terms. Such is the case, for example, with Japan (Katzner [14]). Rather than pursue the maximization of profit (including, under conditions of perfect competition, the selection of a cost-minimizing input mix, the hiring of inputs to the point at which they are paid the values of their marginal products, and the production of output to equate marginal cost to output price), it has been claimed, among other things, that Japanese firms (i) hold service to the general community and the provision of social and economic benefits to their employees as their primary objectives (Abegglen and Stalk [1, pp. 199–203]); (ii) maintain close ties with certain suppliers, in part, by not seeking out supplies from other firms that may be cheaper (Gerlach [8], and Ito [11, pp. 177–196, 214–226]); (iii) frequently make investment decisions on the basis of keeping up with the competition so as to maintain their social status (Nakane [19, p. 90]); and (iv) pay employees according to seniority (not productivity) and tend not to dismiss them during hard economic times when they would be dismissed were they working for Western firms (Nakane [19, pp. 15, 82–83]). All of these characteristics are confluent with the well known Japanese cultural values of loyalty, the honoring of the myriad of obligations that all (Japanese) people and institutions bear, and the maintenance of harmony within society (e.g. Benedict [3] and Nakane [19]). These are not the only differences between Western and Japanese firms that are attributable to cultural dissimilarities. For, in addition to the owners and managers of Western firms, Western employees, too, are propelled in their behavior by selfinterest. This means that to direct its employees’ acitivites, a Western firm cannot only offer them financial and other incentives that appeal to persons driven by self-interest, but also expect that those incentives will elicit, to some significant degree, the desired outcome. In Japan, however, where self-interest is not the main spring of action, such incentive schemes are unlikely to be as successful. Rather, to draw forth desirable conduct, Japanese firms tend to rely more on a process of socialization of employees that appeals to the Japanese cultural values of loyalty, honoring obligations, and harmony, and which leaves employees wanting to do the “right” thing on their own. There have been a few efforts to incorporate some of these anti-profit- antiutility-maximization characteristics in the theory of the firm. For example, Blinder [4] presents a model in which the firm attempts to maximize both profit and employee welfare combined and which, as a consequence, leads it actually to maximize revenue as opposed to profit. And Hay and Morris [10, Ch. 10] describe a model in which firm behavior is generated by aligning the growth of its inputs

3 The growing literature on social preferences that challenges the idea that individuals always behave selfishly (e.g. Fehr and Gächter [7] and Bowles [5, pp. 109ff ]) is ignored.

Cultural variation in the theory of the firm 149 with the growth of the demand for its output. In the latter case, although the model includes a particular notion of growth maximization, the focus on growth itself, with its obvious benefits to both society and employees, nevertheless satisfies, to a considerable extent, characteristic (i) attributed to Japanese firms above. But although profit and utility maximization are avoided as such, since both of these models still contain other forms of maximization by the firm, they do not provide a clean break with postulates of rationality. And, as has been indicated above, those postulates are necessarily irrelevant in models that purport to explain the behavior of Japanese firms. The purpose of this essay is to provide a general model of the firm that, in certain specific instances, allows for movement away from postulates of rationality or maximization and that, therefore, would possibly be applicable in a country like Japan. In particular, the model developed encompasses, as special cases, both firms that do rely on maximization in decision-making, as well as those that do not. For convenience, a specific subset of the firms in the model that rely on postulates of rationality for decision-making are referred to as “Western firms,” while a subset of those whose decisions are based on some other criterion are called “Japanese firms.” Formal definitions of these types of firms, along with a third type that combines elements of each, are provided in Section II. The inclusion of three distinct kinds of firms in the same model in this way permits comparisons between them that would otherwise not be possible. The particular contrasts that emerge relate to both organizational structure and efficiency. With regard to the special case of the Japanese firm, it should be noted that Aoki’s construction [2], although emphasizing different organizational elements, is quite consistent with that set out below. Whereas Aoki describes an organizational structure having both a lack of employee specialization and considerable flexibility in planning (so that the operating units within the firm are easily able to coordinate responses to unforeseen circumstances), subsequent argument is concerned with the organizational coordination of employee production efforts in general. These two approaches, of course, address similar issues from different angles. Aoki also indicates, in confluence with Abegglen and Stalk [1] cited above, how bank and employer interests rather than those of the owners of the Japanese firm dominate and control management decision-making. And this may be understood as a manifestation of the absence of profit maximization – that absence being an integral part of the view of the Japanese firm presented here.

I To account for an expanded range of cultural effects on the firm, attention will focus, in addition to outputs and inputs, on social interactions among its employees. Although inclusion of the latter will require the introduction of variables that appear to be incapable of measurement, those variables will be handled according to techniques set forth by Katzner [12], [16]. The model itself is a variant of that originally proposed by Gintis and Katzner [9] and subsequently invoked by Katzner [13] to study the efficiency of organizational forms.

150 Cultural variation in the theory of the firm Assume lines of authority in the firm are pyramidal so that every employee k (where k = 1, . . . , K) has exactly one immediate superior or supervisor. All employees except those at the bottom end of a line of authority in the pyramid have at least one subordinate. Two or more employees with the same supervisor are co-workers. Situated at the top of the pyramid is the director who is designated by the symbol k = 0. The director may also be thought of as a chief executive, board of directors, or as a combination of both. Each employee k = 0 supplies labor time zk to the firm. During time supplied the employee performs activities ak which are assumed to be independent of zk and fall within the terms of his work arrangement.4 These activities involve social interactions among employees and are limited by the technological imperatives of production. It is not required that the ak be quantifiable. The only restrictions are that each “value” of ak be capable of distinct, discrete, verbal description, and that the set of activities, Ak , over which ak may range also be subject to similarly precise definition. In addition to performing activities, each employee provides rules which constrain the activities (i.e. limit the activity sets) of all workers (not only immediate subordinates) below him in authority. These rules are characterized in terms of the constraints they impose. Thus, if k is above k  in authority and if Rkk  is a set of rules prescribed by k for k  , then Rkk  is defined as a subset of Ak  , or Rkk  ⊆ Ak  . The furnishing of rules may be thought of as an activity which is independent of all other activities and singled out for special attention. Rules may be so restrictive as to dictate exactly what must be done (i.e. select a single element from the employee’s activity set) or they may allow for considerable flexibility and choice.5 Rules also have to be adapted to the technology of production and the capabilities of the individuals for whom they are intended. Moreover, the collection of rules imposed on any employee k  by all persons above him have to be consistent, that is, ∩k Rkk  must be nonempty, where the intersection is taken over all k above k  in authority. Let rk denote the collection of sets of rules – one set Rkk  for each person k  below him – issued by person k. When k has no subordinates, the symbol rk has no meaning. Write a = (a1 , . . . , aK ) and r = (r1 , . . . , rK ). The director’s rules, r0 , are presumed given. The output of the firm, x, depends on quantities of labor time and nonlabor inputs purchased by the firm, as well as on rules for subordinates and activities (consistent with rules received from above) supplied by employees. This production function may be written mathematically as x = f ( y, z, a, r),

(1)

4 The definition of an activity would normally include a statement or an implication regarding the length of time necessary to perform that activity. And this, in turn, suggests a relation between zk and ak . The latter relation, however, lies beyond the scope of present discussion, and will be ignored. 5 From the firm’s perspective, highly restrictive rule sets may be undesirable because under them employees may have more to gain (in relation to their own utility) by rules violations. Thus a costly monitoring mechanism might be necessary to ensure that the rules are carried out.

Cultural variation in the theory of the firm 151 where y = ( y1 , . . . , yI ) is an I -vector of quantities of nonlabor inputs and z = (z1 , . . . , zK ). Actually, since it only serves to limit the values that a can take on and does not affect output directly, r need not be listed as an argument of f . But retaining r in the production function introduces no difficulties and at the same time provides convenience that is useful below. In the currently standard textbook treatment, y and z are picked by the firm and a and r are ignored. The present approach continues the selection of y and z by the firm. But now a separate choice ( y, z) is made for each value of the vector (a, r). And, in addition, a and r themselves are determined through an independent decision process that engages only the firm’s employees.6 Of course, the choice of a and r reflects and describes the relevant social interactions among the employees of the firm. Regardless, it will simplify matters to assume that the markets in which the firm buys its nonlabor inputs and sells its output are all perfectly competitive. Ignoring efficiency wage concerns, let the firm pay each employee k a wage wk according to the exogenous “incentive” function7 wk = W k (ak , rk ),

(2)

where k = 1, . . . , K. Nonmonetary incentives, such as the possibility of promotion and the guaranteed long-term employment granted by many Japanese firms are not considered. Moreover, it is possible for W k to be a constant function associating the same wage to all (ak , rk ) in its domain. Constant incentive functions, of course, are consistent with the practice of paying the employee according to his seniority. The firm’s profit function is
( y, z, a, r, ) = px f ( y, z, a, r) −

I
i=1

pi yi −

K


wk zk ,

(3)

k=1

where px denotes output price, and pi the price of nonlabor input i. To keep matters simple, the director is assumed to be paid out of profit rather than provided a wage. In Western firms, the functions W k frequently have the property, implicit in previous discussion, that those values of (a, r) = (a1 , . . . , aK , r1 , . . . , rK ) that enhance the productivity or profitability of the firm are assigned a higher wage. In that circumstance, however, suppose that for fixed values of px , p1 , . . . , pI , y, and z, output is higher at (a , r  ) than at (a , r  ). If profit is also to be higher at (a , r  ), then it is necessary that the W k be set so that the cost of the additional wages paid employees at (a , r  ) over those paid at (a , r  ) is less than the additional revenue

6 The presence of real capital assets among the components of y could be taken to imply that the model, in making y a decision variable, is determining an instantaneous optimal structure for the firm. Analytically, that raises the question as to whether employees’ choices of activities and rules remain constant throughout the lives of the assets. But this problem is set aside and not considered in subsequent discussion. 7 It is implicit here that the values that the (ak , rk ) can take on are known to the firm. The determination of W k in (2) and its relation to the labor market is disregarded.

152 Cultural variation in the theory of the firm received at (a , r  ). Were this not the case, then the most productive (a, r) need not be the most profitable, and hence moving to increase productivity might reduce profitability. Formally, the collection of incentive functions {W k }, one for each employee k = 1, . . . , K, is said to be profit efficient (Gintis and Katzner [9, p. 282] and Katzner [13, pp. 548–549]) whenever f ( y, z, a , r  ) ≥ f ( y, z, a , r  ) if and only if
( y, z, a , r  ) ≥
( y, z, a , r  ), for all y, z, a , a , r  and r  . Profit-efficient incentives, then, have the property that productivity increases due to changes in (a, r) always are translated into larger profit. Note that, in addition to incentives which provide wage raises that, in the aggregate, are smaller than resulting revenue increments, all collections of constant incentive functions are also profit efficient. According to the cultural distinctions drawn at the outset, there are two general possibilities for the selection of y and z by the firm. On the one hand, Western firms choose y and z on the basis of profit maximization in relation to (3), given the W k and values for px , p1 , . . . , pI , a, and r. For fixed (a, r), the consequences of these choices under standard assumptions imposed on the production function f in (1) are well known and not repeated here. On the other hand, non-Western firms determine y and z given a and r according to some generally non-profit-maximizing criterion that, for present purposes, is not necessary to specify. For convenience it is assumed that those firms also take prices px and p1 , . . . , pI as parameters. With these latter parameters fixed, the relationship between selections of y and z and values of a and r in non-Western firms may be summarized by the firm selection function s f as follows: ( y, z) = s f (a, r).

(4)

Of course, the possibility that s f might reflect profit maximization as a special case has not been ruled out. It is worth providing an example to illustrate how the function s f might be determined. Consider the Japanese firm. With (a, r) set, specific social interactions among employees obtain. Given those interactions, the hours each employee supplies and the quantity of nonlabor inputs with which he works impinge on his ability to fulfill his social and productive obligations and on the nature of the harmony present within the firm. This, in turn, affects overall cooperation and productivity. Historically the firm has established and become comfortable with a certain realization of harmony and a particular ability among employees to fulfill obligations. Suppose, then, that given any (a, r), the firm chooses y and z to maintain as much as possible that harmony and ability. In this way, a ( y, z) is determined for each (a, r).

Cultural variation in the theory of the firm 153 There are also two general possibilities for the selection of a = (a1 , . . . , aK ) and r = (r1 , . . . , rK ) by the firm’s employees. For both instances, assume employees choose values of the ak and rk (subject to the constraints indicated below) independently of their selection of leisure time and consumption so that these latter selections may be ignored. With respect to Western firms, let the utility function, uk , (assumed to exist) of employee k be written as µk = uk (ak , rk , wk ).

(5)

Note that the appearance of wk as an argument of uk makes the role of incentives in employee decision-making explicit. However, substitution of (2) into (5) eliminates wk and reduces uk to a function U k of ak and rk alone: µk = U k (ak , rk ).

(6)

Suppose that the chosen value of (a, r) = (a1 , . . . , aK , rk , . . . , rK ) emerges uniquely from the simultaneous maximization of U k (ak , rk ) by each employee k = 0, subject to the constraints that, as k varies, the ak are consistent with all rules imposed on k from above or, in other words, that ak is in ∩k  Rk  k for every k, where the intersection is taken over all k  above k in authority. Such an (a, r) is called an internal equilibrium for the Western firm. Thus, in the Western case, and with the director’s rules r0 given, (a, r) is obtained from utility maximization, ( y, z) is secured from profit maximization as described earlier, and the firm’s output and profit are determined from (1) and (3). The collection of all internal equilibria for the Western firm generated as the director’s rules vary is denoted by E. Note that E contains the internal equilibria arising when the director’s rules are such that they impose no effective constraints on the activities of any worker. For non-Western firms, let each (ak , rk ) be chosen by k in light of the given directoral rules r0 and some generally non-utility-maximizing principle that is (in parallel to the selection of y and z by the non-Western firm) left unspecified and is subject to the constraint that ak is in ∩k  Rk  k for each k, where, once again, the intersection is taken over all k  above k in authority.8 As before, the possibility that (ak , rk ) is determined from the maximization of a utility function like (6) is not precluded, and the selected vector (a, r) = (a1 , . . . , aK , r1 , . . . , rK ) is an internal equilibrium for the non-Western firm. The set of all such internal equilibria as the director’s rules vary is identified by the same symbol E used earlier.

II The model set out above encompasses, among other possibilities, two general kinds of firms: the profit maximizing firm with utility-maximizing employees and the

8 Although it could happen by coincidence, in neither the Western nor the non-Western firm are employees necessarily choosing the (ak , rk ) on the basis of wage or income maximization. In other words, the general approach here includes the latter as a special case.

154 Cultural variation in the theory of the firm non-profit-maximizing, non-utility-maximizing alternative. In each case, the actual behavior of the firm in question rests, in addition to the input quantities selected by the firm, on the internal equilibrium chosen by its employees. Also in each case, it is clear that the director (or board of directors) is unlikely to be indifferent among the options from which his respective employees choose. Suppose the preferences of the director are represented by the utility function U 0 (a, r). Of course, the director of the Western firm would prefer (a , r  ) to (a , r  ) whenever there exists a ( yˆ , zˆ ) such that
( yˆ , zˆ , a , r  ) −
( y, z, a , r  ) > 0 for all values of ( y, z), and this would be reflected in the nature of his U 0 . The director of the non-Western firm, however, would probably have different priorities. For example, it has been pointed out above in reference to Abegglen and Stalk [1] and Aoki [2] that Japanese firms take their primary objectives to be service to the community and the provision of social and economic benefits to their employees. The Japanese director’s utility function might also be influenced by the socialization and consensus-building that normally takes place within the Japanese firm. In either the Western or non-Western case, denote the firm’s target values of output and input as determined by the director by xT , yT , and z T . Assume the Western director obtains these targets as follows. Find the profit-maximizing vector ( y, z) corresponding to each internal equilibrium. Then choose for yT and z T those respective values of y and z that yield the greatest profit when accompanied by their associated internal equilibria. The output target xT is now secured upon substitution in (1). Obviously, the (a, r) identified with ( yT , z T ) maximizes U 0 . For the nonWestern firm, the targets would be determined with respect to some generally non-profit-maximizing criterion. In the Japanese firm, say, that criterion might have to do with the maintenance of both the firm’s social status and its past rate of growth. Regardless, such targets would relate back to the firm’s internal equilibria through (1) and (4), and may interact with the director’s preferences among those equilibria as expressed in U 0 . To analyze the role of the director’s preferences and targets in relation to firm behavior, additional concepts are needed. These concepts all relate to the means of coordination of employee productive activities within the firm. An employee k = 0, whose behavior is derived from utility maximization, is called incentive motivated (Gintis and Katzner [9, p. 282] and Katzner [13, p. 556]) if for all pairs of values (ak , rk ) and (ak , rk ) the numerical inequality W k (ak , rk ) ≥ W k (ak , rk ) implies U k (ak , rk ) ≥ U k (ak , rk ). Incentive motivation means that individual preferences are such that, upon the maximization of their utility functions, employees respond in the “same direction”

Cultural variation in the theory of the firm 155 as that of incentives.9 This is the theoretical representation of self-interest on the part of employees in Western and other firms. That self-interest, together with the proper structuring of incentive functions (not yet completely specified), is what permits those firms to harness, should they wish to do so, their employees to the engine of profit maximization. Now recall production function (1). For employee k = 0 consider any two vectors (ak , rk ) and (ak , rk ) such that (i) output is at least as high using ak in (1) instead of ak ; and (ii) rk restricts all subordinates of k to subsets of their respective activity sets obtained under rk on which output is at least as large as it is anywhere else in the rk sets. Other things being equal, then, the output associated with (ak , rk ) cannot be less than that associated with (ak , rk ). Assume that all pairs of vectors (ak , rk ) can be compared in this way, and that (i) and (ii) hold regardless of the activities performed by the remaining employees and regardless of the quantities of labor time and nonlabor inputs hired by the firm. Under these conditions, if W k (ak , rk ) ≥ W k (ak , rk ), then W k is referred to as nondecreasing (Katzner [13, p. 557]). In other words, incentive functions are nondecreasing provided they are set so as to reward employees toward activities which do not decrease output. Similarly, an employee k, whose behavior is derived from utility maximization, is said to internalize the worth of productivity 10 when, under the same conditions, U k (ak , rk ) ≥ U k (ak , rk ). Thus, the employee’s preferences among pairs of vectors of activities and rules favor those that do not reduce output. The reason for the introduction of this concept will become clear shortly. Finally, employee k = 0 internalizes the values of the firm provided that, with (ak , rk ) a component of (a , r  ) = (a1 , . . . , aK , r1 , . . . , rK ) and (ak , rk ) a component of (a , r  ) = (a1 , . . . , aK , r1 , . . . , rK ), he chooses (ak , rk ) over (ak , rk ) if and only if U 0 (a , r  ) ≥ U 0 (a , r  ). The presence of this property among all of the firm’s employees may be interpreted to mean that a socialization process takes place within the firm that leaves employees always prepared to select those activities and rules that are consistent with the director’s preferences. (The latter socialization process is the mechanism

9 Employees may be receiving job rents. The possibility that those rents could be eroded by modification of the W k is not considered. 10 This concept was referred to in Gintis and Katzner [9, p. 284] and Katzner [13, p. 557] as “internalizing the values of the firm.” As noted momentarily, it is a special case of a more general notion that is given the latter name below.

156 Cultural variation in the theory of the firm of organizational coordination in the Japanese firm alluded to earlier.) In general, such choices do not reflect postulates of rationality since they involve neither the employees’ own utility functions (preferences) nor their maximization. However, internalizing the worth of productivity may still be viewed as the special case of internalizing the values of the firm in which (i) the director has preferences that always lean towards increased output; (ii) the director and employee utility functions are consistent in the sense that, for all (a , r  ) and (a , r  ), U 0 (a , r  ) ≥ U 0 (a , r  ) if and only if U k (ak , rk ) ≥ U k (ak , rk ) for every k; and (iii) the U k are maximized to determine the employees’ choices of (ak , rk ). It is now possible to give full definitions of the notions of Western and Japanese firms. Like the standard characterization of perfect competition, these may be thought of as “ideal types” which do not usually appear in reality exactly as pictured, but which are useful for study nonetheless. A firm is called Western provided that both it and its employees make decisions (including the setting of targets) on the basis of, respectively, profit and utility maximization, all incentive functions W k (where k  = 0) are nondecreasing and (as a group) profit efficient, and all employees (k = 0) are incentive motivated.11 It is referred to as Japanese whenever profit and utility maximization do not guide (respectively) the firm and its employees in their decision-making, all incentive functions are constant functions (and hence profit efficient), and all employees (k  = 0) internalize the values of the firm. These definitions capture and formalize the essence of the concepts of Western and Japanese firms, along with the cultural differences they reflect, suggested in earlier discussion. A third type of firm is useful for comparative purposes. A clan (Ouchi [18, p. 132]) is a firm in which all decisions (including the setting of targets) of employees and firm are, respectively, utility- and profit maximizing, all incentive functions are constant functions (and hence profit-efficient), and all employees (k  = 0) internalize the worth of productivity.12 Evidently, the clan is a hybrid containing elements of both Western and Japanese firms in which utility and profit maximization are taken from the former, and the special instance of internalizing the values of the firm, namely, internalizing the worth of productivity, comes from the latter. That is, the clan is a Japanese firm in which utility and profit maximization on the part of, respectively, employees and firm have been added.

11 The Western firm was called a “bureaucracy” by Ouchi [18, p. 134] and Katzner [13, p. 558]. 12 The fact that it is technically possible for a firm to be, at the same time, both Western and a clan is irrelevant for present argument. See Katzner [13, p. 558n].

Cultural variation in the theory of the firm 157 Of course, the targets xT , yT , and z T are set by both Western firms and clans as those which maximize profits over all vectors ( y, z, a, r) according to the method described previously. The latter maximizing vectors in the Western firm and the clan are assumed to exist uniquely in each case. Consider now a distinct internal equilibrium (a, r) in different sets E for each of the Western firm and the clan described above. Such equilibria have been defined in general in Section I and hence apply directly to these two cases. The following proposition follows immediately from previous definitions and assumptions. Theorem 1. Let fixed values of y and z be given and suppose the director’s rules impose no effective constraints on the selection of activities and rules by any worker. Then in both Western firms and clans, internal equilibrium always maximizes profit over the set of all consistent vectors (a, r), that is, over the relevant E. Thus, with profit-maximizing targets xT , yT , and z T , if the firm (director) sets y = yT and z = z T , then employee selections of (a, r) ensure that x = xT and hence that all targets are met. Theorem 1 has several implications. First, every (a, r) in the relevant E that is not profit maximizing must arise because the director’s rules exclude the profitmaximizing (a, r) from the options opened to employees. Although logically possible, one would not expect the director’s rules in Western firms and clans to be so constituted. Second, in each case, individual action results in profit maximization automatically without needing the director to dictate rules. Even if all targets were met, Japanese firms do not, except by coincidence, operate at profit- and utilitymaximizing values of y, z, a, and r. Hence, in this sense they are not generally as efficient as their Western counterparts.13 But there is still another sense in which efficiency escapes the Japanese firm. This second kind of efficiency arises in relation to the notion of internal Pareto optimality (Gintis and Katzner [9, p. 280], and Katzner [13, p. 560]). And inefficiency is present because, under the conditions that generally apply in the Japanese situation, it is possible, without any change in input quantities and without making anyone worse off, to expand output by reallocating activities and rules inside the firm, that is, by rearranging the social interaction among the firm’s employees. The argument is as follows. Let E be given along with y and z in the production function (1). A vector (a0 , r 0 ) = (a01 , . . . , a0K , r10 , . . . , rK0 ) in E is called internally Pareto optimal in production as long as there is no other (a, r) in E such that U k (ak , rk ) ≥ U k (a0k , rk0 ),

13 For further discussion of this and related kinds of inefficiency in Japanese firms, see Katzner [15] which appears as Essay 6 above.

158 Cultural variation in the theory of the firm for all k = 1, . . . , K, U k (ak , rk ) > U k (a0k , rk0 ), for at least one k = 1, . . . , K, and f ( y, z, a, r) ≥ f ( y, z, a0 , r 0 ). Thus at an internal Pareto optimum, no reorganization of production in terms of activities and rules can make one employee better off, and no one else worse off, without lowering output. If, for example, through the imposition of rules, workers are not permitted to perform their jobs in ways in which they prefer, even though their preferred ways leads to the same output, then the (a, r) obtained is not internally Pareto optimal. Note that the director is not included in this notion of Pareto optimality and that the functions U k and f are all taken to be given. Propositions asserting the internal Pareto optimality of internal equilibria in Western firms and clans are presented below. In all cases, y and z are assumed fixed. Only the proof of Theorem 3 is outlined here; that of Theorem 2 is similar.14 Theorem 2. Let the firm described in Section I be Western and suppose the director’s rules impose no effective constraints on the selection of activities and rules by any worker. If (¯a, r¯ ) is an internal equilibrium for the firm, then it is also internally Pareto optimal in production. Theorem 3. Let the firm described in Section I be a clan and suppose the director’s rules impose no effective constraints on the selection of activities and rules by any worker. If (¯a, r¯ ) is an internal equilibrium for the firm, then it is also internally Pareto optimal in production. Proof : Suppose (¯a, r¯ ) = (¯a1 , . . . , a¯ K , r¯1 , . . . , r¯K ) is an internal equilibrium in the appropriate E (given y and z). Invoking Theorem 1 and the profit efficiency property that characterizes (in part) clans, (¯a, r¯ ) also maximizes both profit and output. If (¯a, r¯ ) were not internally Pareto optimal, then there would exist a reorganization of production (a, r) = (a1 , . . . , aK , r1 , . . . , rK ) in E such that U k (ak , rk ) > U k (¯ak , r¯k ),

(7)

for some employee k, and the levels of output and of utility of all other employees are no lower at (a, r) than at (¯a, r¯ ). Now because the firm is a clan, all employees internalize the worth of productivity. But applying the definition of internalizing the worth of productivity by employee k to (7) implies15 f ( y, z, a, r) > f ( y, z, a¯ , r¯ ),

14 A proof of Theorem 2 may be found in Katzner [13, pp. 561–562]. 15 Recall that underlying the notion of internalizing the worth of productivity is the assumption that all vectors (ak , rk ) can be compared, as previously indicated, in relation to the firm’s output.

Cultural variation in the theory of the firm 159 and this contradicts the maximality of output at (¯a, r¯ ). Therefore (¯a, r¯ ) is internally Pareto optimal in production. Q.E.D. Thus the clan, which, recall, is a Japanese firm to which maximization has been added, is efficient. But without that latter addition, because neither the employees of the firm, nor the firm itself maximizes in their respective decision-making, the Japanese firm cannot, in general, be expected to achieve internal Pareto optimality. It should be pointed out, however, that the absence of efficiency in this sense, even when combined with that in the preceding sense derived from Theorem 1, does not necessarily threaten the survival of the firm. This is due to the fact that mitigating circumstances and offsetting effects might be present. Indeed, it has been argued elsewhere (Katzner [15]) that, in spite of inefficiencies like those elucidated here, offsets such as the presence of large quantities of unpaid labor, the bonus system with respect to which employees are remunerated, and a willingness to accept low rates of return on invested capital, permitted actual Japanese firms, which have many of the characteristics of the theoretical Japanese firm defined above, to lead the Japanese economy to almost 20 years of unprecedented growth during the decades following the Korean War. Finally, it is clear, and has been amply demonstrated above, that the theory of the firm is perfectly capable of expansion to include the possibility of variation across different cultures. Such expansions, moreover, can be made sufficiently rich to permit the comparison of properties of firms that operate on the basis of distinct cultural backgrounds. And adding these cultural dimensions to the theory of the firm provides, in many cases, more realistic and significant explanations of firm behavior than those obtained by the forced imposition of irrelevant postulates of rationality as representative of the primary motivating impulse that drives individual and firm action.

References 1. Abegglen, J.C. and G. Stalk, Jr., Kaisha, The Japanese Corporation (New York: Basic Books, 1985). 2. Aoki, M., “Toward an Economic Model of the Japanese Firm,” Journal of Economic Literature 28 (1990), pp. 1–27. 3. Benedict, R., The Chrysanthemum and the Sword (Boston: Houghton Mifflin, 1946). 4. Blinder, A.S., “A Simple Note on the Japanese Firm”, Journal of the Japanese and International Economies 7 (1993), pp. 238–255. 5. Bowles, S., Microeconomics: Behavior, Institutions and Evolution (Princeton: Princeton University Press, 2004). 6. Caldwell, B.J., Beyond Positivism: Economic Methodology in the Twentieth Century (London: Allen & Unwin, 1982). 7. Fehr, E. and S. Gächter, “Fairness and Retaliation: The Economics of Reciprocity,” Journal of Economic Perspectives 14, no. 3 (Summer, 2000), pp. 159–181. 8. Gerlach, M.L., Alliance Capitalism: The Social Organization of Japanese Business (Berkeley: University of California Press, 1992).

160 Cultural variation in the theory of the firm 9. Gintis, H. and D.W. Katzner, “Profits, Optimality and the Social Division of Labor in the Firm,” Sociological Economics, L. Lévy-Garboua, ed. (London: SAGE, 1979), pp. 269–297. Reprinted as Ch. 11 in D.W. Katzner, Analysis without Measurement (London: Cambridge University Press, 1983). 10. Hay, D.A. and D.J. Morris, Industrial Economics and Organization (Oxford: Oxford University Press, 1991). 11. Ito, T., The Japanese Economy (Cambridge: MIT Press, 1992). 12. Katzner, D.W., Analysis without Measurement (Cambridge: Cambridge University Press, 1983). 13. ——, “The Efficiency of Organizational Forms,” Économie Appliquée 40 (1987), pp. 539–564. Reprinted in D.W. Katzner, Unmeasured Information and the Methodology of Social Scientific Inquiry (Boston: Kluwer, 2001), pp. 231–252. 14. ——, “Western Economics and the Economy of Japan,” Journal of Post Keynesian Economics 21 (1999), pp. 503–521. This volume: Essay 2. 15. ——, “Explaining the Japanese Economic Miracle,” Japan and the World Economy 13 (2001), pp. 303–319. This volume: Essay 6. 16. ——, Unmeasured Information and the Methodology of Social Scientific Inquiry (Boston: Kluwer, 2001). 17. ——, “‘What Are the Questions?’” Journal of Post Keynesian Economics 25 (2002), pp. 51–68. This volume: Essay 3. 18. Ouchi, W.G., “Markets, Bureaucracies, and Clans,” Administrative Science Quarterly 25 (1980), pp. 129–141. 19. Nakane, C., Japanese Society (Berkeley: University of California Press, 1970). 20. Weber, M., The Protestant Ethic and the Spirit of Capitalism, T. Parsons, trans. (New York: Scribner, 1958).

10 Culture and the explanation of choice behavior1

To explain is to clarify, to render intelligible, and to provide reason for. In economics, the object of explanation has always been observable economic behavior, where the behavior in question is that of human beings as it arises either in isolation or in the context of the activity of economic institutions. And in recent times, the means of explanation in economics has largely come to consist of the construction of (usually mathematical) models in which the observed behavior under scrutiny is derived, at least in principle, from initial assumptions. It is an obvious consequence of the combination of these latter facts concerning practice in economics with the meaning of explanation, that for models to have relevance for economic explanation they must have a basis in human experience. That is, they have to reflect or be consistent with the standards that real people employ in arriving at their own personal perceptions, judgments, and actions. Now, since it is well known that individuals coming from different cultural backgrounds have different perceptions and different standards for interpretation, judgment, and action, it would seem imperative that explanations of economic behavior arising in different cultural environments would necessarily require the construction of different models. Such differences, of course, would appear first in the assumptions upon which the models are founded and second, and as a consequence, in the properties of the observable behavior that the model is designed to explain. Not only would variation in assumption content echo differences of setting or context within which behavior takes place, but it would also mirror cultural differences in values, norms, and rules of action. This includes the possibility, in some cases, of claims of such cultural imperatives as courtesy, modesty, conformity, loyalty, or custom overriding self-interest in the form of pursuit of the most preferred as the driving force in the making of economic decisions.

1 Theory and Decision 48 (2000), pp. 241–262. © Kluwer Academic Publishers. Reprinted in slightly modified form with permission. This essay is an outgrowth of a workshop presentation at the University of Massachusetts / Amherst by Wulf Gaertner. The author would like to thank Professor Gaertner for that presentation and for his subsequent help.

162 Culture and the explanation of choice behavior The role of cultural factors in influencing the construction of explanatory economic models was implicitly recognized by Sen in his recent assertion [10, p. 499] that the notion of consistency of choice behavior is only meaningful with respect to some “external referent” that indicates what the individual in question is trying to do. Now that referent clearly has to be, at least in significant part, culturally determined. In the “traditional” case relating to Western culture and society, the external referent is preference optimization (i.e. the selection of the most preferred), and choice consistency is characterized in terms of the satisfaction of certain properties of choice behavior that are derived from the imposition of preference optimization and other assumptions. That the hypothesis of preference optimization has emerged as the external referent in Western culture is not because it has withstood a barrage of empirical tests hurled against it through the study of observable choice behavior. Indeed, since any disconfirming instance can always be attributed to, say, preference variation from one observation to the next, and since confirming instances only indicate consistency with the hypothesis and not that it is actually in play, complete and convincing tests of the hypothesis are not possible. Rather, the hypothesis has reached its lofty heights because preference optimization is a reflection of self-interest, a motivating force that has come to lie at the heart of Western culture. Indeed, the explanation of choice behavior derived from the hypothesis is meaningful precisely because of the link between preference optimization and the central pillars of Western culture (Katzner [5]). It follows that for cultures in which self-interest plays a less important role, explanations of choice behavior based on preference optimization are less relevant. Such is the case, for example, with Japan. For, in contrast with the West, self-interest has not been institutionalized into Japanese culture, is not a fundamental component of it, and is not a major theme interwoven into the fabric of Japanese life. Thus, if explanations of Japanese choice behavior are to be made more meaningful and relevant to Japanese experience, it might be better to discard preference optimization as the motivating force in models that purport to explain that behavior (Katzner [5]). Nor should preference optimization appear as an external referent. As a result, choice consistency would mean something different with respect to Japanese society from what it means in relation to the West. Additional arguments supporting the view that culture matters in determining economic behavior, and that it needs to be accounted for in explaining that behavior, have been given by Greif [3], Hampden-Turner and Trompenaars [4], North [7], and Roth et al. [8].2

2 Althouth Roth et al. assume, somewhat contrary to the above argument, that all players of their ultimatum games (including the Japanese) are attempting to maximize earnings, they do wind up questioning that assumption [8, p. 1093]. It is also possible that their experimental design was such as to surpress the primary motivations of the Japanese (if, in fact, that actually hapened), and to permit the residual emergence of only earnings maximization or some other secondary force. Regardless, Roth et al. do conclude that “… differences in bargaining behavior among countries … may tentatively be attributed to cultural differences” [8, p. 1068].

Culture and the explanation of choice behavior 163 The purposes of this essay are (i) to suggest that, although all choice behavior may be understood in terms of preference optimization, in certain cultural circumstances better explanation derives from the recognition and incorporation of more significant cultural imperatives, and (ii) to show how alternative cultural imperatives can be substituted in place of self-interest as the primary force that propels action in the theory of choice. In particular, while the conventional assumption of preferences among choice options is retained, forces different from optimization that drive decision-making activity are introduced. The latter forces include selection of the “second best” and selection of the “median” or “central” option.3 A version of the standard choice model is examined first to set the stage for subsequent discussion. In all circumstances considered, the choice problem is limited to a finite universe of choice options.

I Let X = {x1 , . . . , xI } be a finite set of I objects and S(X ) the collection of all nonempty subsets of X . A (strict) preference relation is a binary relation defined on X which is irreflexive, transitive, and total.4 Take c to be a (single-valued) choice function on S(X ), that is, c is a single-valued function, with domain S(X ), such that c(S) is in S, for every S in S(X ). Although current fashion in the theory of choice is to focus on generalized choice functions that may be multi-valued, and out of which single-valued choice functions emerge as a special case, there is little point in doing so here. For observable choice behavior generally takes the form of selecting a single object from a collection of many objects. And since multi-valued choice functions do not describe such phenomena, they are unable to provide full explanations of observable choice behavior. In any case, the preference relation is said to rationalize c if and only if for every S in S(X ), where S contains more than one object, c(S) x,

(1)

3 Biagent and Gaertner [1] and Gaertner and Xu [2] discuss formally similar issues in the context of an extended notion of rationality or self-interest. For them, behavior that is nonoptimizing in the traditional sense is still regarded as optimizing, but in a more general sense. In the present approach, however, such activity is interpreted as nonrational or nonoptimizing behavior that may emerge in alternative (non-Western) cultural environments. 4 The relation is irreflexive whenever, for every x in X , the statement x x is not valid. It is transitive if, for all x , x , and x in X , x x and x x imply x x . And it is total provided that, for every distinct pair x and x in X , either x x or x x . Observe that irreflexivity and transitivity imply asymmetry, that is, for any x and x in X , if x x , then x x cannot hold. For, were both x x and x x to be in force together, then by transitivity, x x , contrary to irreflexivity. Typically, a nontotal could be derived from a preference-indifference relation that is reflexive, transitive, and total. But since the choice function defined next is single-valued, the possibility of indifference between distinct pairs of objects is ruled out, and such a preference-indifference relation would reduce to the preference relation described in the text.

164 Culture and the explanation of choice behavior for all x  = c(S) in S. When c is rationalizable, the behavior exhibited in c is called rational, and the motivation behind it is referred to as self-interest or rationality.5 Choice functions that are rationalizable exhibit a characteristic known as Property α: For all S and T in S(X ), if S ⊆ T and c(T ) is in S, then c(S) = c(T ). In other words, if the object selected by c from T is contained in a subset S of T , then the same object is selected from S. Two propositions (Sen [9]) describe the relationship between rationalizability of (single-valued) choice functions and Property α. Theorem 1. Let X be given and be defined on X . If rationalizes c, then c satisfies Property α. Proof : Choose S and T in S(X ) with S ⊆ T and c(T ) in S. If S = T , there is nothing to prove. Otherwise, suppose rationalizes c. Then from (1), c(T ) x, for all x = c(T ) in T . In particular, c(T ) x, for all x  = c(T ) in S. Applying (1) again, c(S) = c(T ). Q.E.D. Theorem 2. Let X be given and c be a choice function defined on S(X ). If c satisfies Property α, then there exists a preference relation rationalizing c. Proof : Let c exhibit Property α. For all x and x in X , set x x if and only if = x and there exists an S in S(X ) such that x = c(S) and x is in S. Since S(X ) contains all sets of distinct pairs x and x , where x and x are in X , and since the domain of c is all of S(X ), one may conclude that is a total relation defined on X . It is now demonstrated that is irreflexive and transitive. That is irreflexive results from the fact that x  = x in the definition of . To show transitivity, suppose x x and x x , for some x , x , and x in X . Then there exist S and T in S(X ) such that x = c(S) and x is in S, and x = c(T ) and x is in T . Since {x , x } ⊆ S, it follows from Property α that c({x , x }) = x . Similarly, c({x , x }) = x . Consider the set {x , x , x } in S(X ). Were c({x , x , x }) = x , then since {x , x } ⊆ {x , x , x }, Property α would imply c({x , x }) = x , contrary to the above. Likewise, c({x , x , x }) = x leads to a similar contradiction. Thus c({x , x , x }) = x and the set {x , x , x } has the requisite property of S in the definition of ensuring that x x . Hence is transitive. Clearly, also satisfies (1). Therefore rationalizes c. Q.E.D. x

5 Other notions of rational, self-interest, and rationality are, of course, possible. See, for example, Biagent and Gaertner [1] and Gaertner and Xu [2].

Culture and the explanation of choice behavior 165 Thus choice behavior satisfying Property α is rational in that it can be explained as selecting the “best” or most preferred object under from each S in S(X ). But the same choice behavior satisfying the same Property α can also be explained in terms of a decision rule that requires the choosing of the “worst ” object in each case. For reversing the inequality in (1) would have no effect on Theorems 1 and 2. Indeed, the proofs of these theorems would need no modification except for the definition of in the proof of Theorem 2, which would now require that x x mean x = c(S) and x is in S, instead of x = c(S) and x is in S. In this sense, then, the selection of the worst object from each S in S(X ) could also be thought of as a form of rationality. From here on, however, the terms “rational,” “rationalize,” and “rationality” will be used only in reference to (1). The possibility of reversing the inequality in (1) will be ignored. There are, of course, many choice functions that violate Property α and are not, as a result, rationalizable. One possibility is related to an example given by Sen [10, p. 501]. Imagine an individual whose preferences among the objects of X are x1 x2 · · · xI . In Sen’s case the xi are pieces of cake ordered by size and not by preference. Here, however, assume that the individual prefers more to less so that the ordering of pieces of cake by size becomes identical to that by preference. Now, in spite of his preference for larger size, suppose also that the individual lives in a cultural environment in which it is not proper to take the largest piece off of any tray presented to him for choice. Thus he always selects the next-to-largest piece of cake from any subset containing two or more elements of X . (For present purposes, though this will be changed later, it is not necessary to consider choices, if any, from single-element subsets.) In particular, if T = {x1 , x2 , x3 } and S = {x2 , x3 }, then c(T ) = x2 and c(S) = x3 . But even though S ⊆ T and x2 is in S, c(S)  = c(T ). Therefore, c cannot satisfy Property α, and the behavior characterized by c cannot be rational. It is possible to argue, however, that the nonrationality exhibited in the cake example arises because the choice problem has not been properly formulated. For each piece of cake may be viewed as carrying, in the mind of the individual, some cultural baggage along with it. To select a piece of cake is also to select, at the same time, a social element that is measured against certain cultural standards to determine its propriety. Thus the individual is simultaneously choosing a piece of cake and a “level” of propriety. And permitting the full expression of that person’s concern for propriety will lead to a preference relation that rationalizes c. This suggests that, in general, whenever an investigator is confronted with a choice function that appears to be nonrationalizable, it is always possible, by taking into account extra dimensions or characteristics of the objects in X , to embed it in a “larger” choice function that is rationalizable. To see what is involved, observe first that Property α is, without exception, in force for all choice functions obtained with respect to an X that contains but one (I = 1) or two (I = 2) objects. It is only when I ≥ 3 that the possibilities of violations of Property α and nonrationalizability arise. Consider, then, the case in which I = 3. Here X = {x1 , x2 , x3 } and S(X ) = {{x1 }, {x2 }, {x3 }, {x1 , x2 }, {x1 , x3 }, {x2 , x3 }, {x1 , x2 , x3 }}. Assume x1 x2 x3 ,

166 Culture and the explanation of choice behavior where is irreflexive, transitive, and total. Suppose the choice function is given by 6 c({xi }) = xi ,

i = 1, 2, 3,

c({xi , xj }) = xj ,

i = 1, 2, j = 2, 3,

and i < j,

c({x1 , x2 , x3 }) = x1 . It is clear that c is not rationalizable and that the only violations of the condition of Property α occur with respect to the two pairs of sets (i) {x1 , x2 } and {x1 , x2 , x3 }, and (ii) {x1 , x3 } and {x1 , x2 , x3 }. Now consider a social characteristic of the objects of X , such as, to use illustrations somewhat different from that of the cake example, the extent to which the selection of particular objects mirrors loyalty to a person or group, or reflects the fulfilling of one’s social obligations. Clearly, since manifestations of characteristics of this sort may have different significances in different cultural environments, and may differ in their impact on decision-making in different ways at different times, choice behavior as subsequently described becomes, in part, culturally determined, and any notion that the choosing agent or individual is autonomous or independent of cultural variation is compromised. Regardless, let ki , where i = 1, 2, 3, represent three possible manifestations of this characteristic, verbally or numerically described. Write K = {k1 , k2 , k3 }. Think of the objects now in view as consisting of the original object together with a manifestation of the characteristic, that is, as vectors (xi , kj ). Thus the three-element collection X is replaced by the nine-element collection X × K = {(xi , kj ) : i, j = 1, 2, 3}, and S(X ), now written S(X × K), is revised accordingly. Maintaining consistency with the original ordering, extend to X × K so that7 (x1 , kj ) (x2 , kj ) (x3 , kj ),

j = 1, 2, 3.

Further require the extension to be irreflexive, transitive, and total on X × K, and to include (x3 , k2 ) (x1 , k1 ),

and

(x3 , k3 ) (x1 , k2 ).

Then it follows, in particular, that (x2 , k2 ) (x1 , k1 ),

(2)

(x3 , k3 ) (x1 , k1 ),

(3)

6 This is similar, but not identical, to the cake example described above. To have identity would require at least the modification c({x1 , x2 , x3 }) = x2 . Recall also that, in the cake case, choices from the single-object sets {xi } were not specified. 7 In his analysis of demand behavior, Lancaster [6] employs characteristics somewhat differently. For Lancaster, objects are formally related to the vectors of characteristic quantities they possess, and preferences are defined only among the latter.

Culture and the explanation of choice behavior 167 and (x3 , k3 ) (x2 , k2 ). The choice function c* rationalized by this ordering consists, in part, of the following: c*({(xi , ki )}) = (xi , ki ),

i = 1, 2, 3,

c*({(x1 , k1 ), (x2 , k2 )}) = (x2 , k2 ),

(4)

c*({(x1 , k1 ), (x3 , k3 )}) = (x3 , k3 ),

(5)

c*({(x2 , k2 ), (x3 , k3 )}) = (x3 , k3 ), c*({(x1 , k1 ), (x2 , k1 ), (x3 , k1 )}) = (x1 , k1 ). Observe that (2) and (3) are responsible for implying, respectively, (4) and (5), and hence, ignoring the added dimension k, the selection of x2 over x1 , and x3 over x1 . But substituting k1 , say, for k2 and k3 in the left-hand sides of (4) and (5), x1 is chosen over both x2 and x3 . Thus the presence of (2) and (3) eliminates the violation of Property α with respect to c in terms of the original pairs of sets (i) {x1 , x2 } and {x1 , x2 , x3 }, and (ii) {x1 , x3 } and {x1 , x2 , x3 } because, with the specific values of k added to the elements of {x1 , x2 } and {x1 , x3 } in (4) and (5), {x1 , x2 } and {x1 , x3 } are no longer subsets of {x1 , x2 , x3 }.8 Moreover, if the components kj were deleted from the part of c* displayed above, then that part of c* would be identical to c. But c* is rationalizable and c is not. It is in this sense that the nonrationalizable choice function c can be imbedded in the rationalizable choice function c*, and “irrational” behavior can be explained as if it were rational. As demonstrated by the next proposition, this result is perfectly generalizable to all finite X and all nonrationalizable c, with I ≥ 3. Theorem 3. Let X be given with I ≥ 3, and let c be a nonrationalizable choice function defined on S(X ). Then there exists a finite set K and a rationalizable choice function c* defined on S(X × K) in which c is embedded. Proof : Let c be nonrationalizable. Then there exist at least one pair S and T in S(X ) with respect to which c does not satisfy Property α. Attach a new variable component k to the elements of X so that each x in X becomes (x, k) , where specific values of k are to be indicated below. Let k range over some finite set of objects K exactly large enough (and no larger) to meet the requirements of the subsequent argument.

8 Of course, {(x1 , k1 ), (x2 , k1 )}, say, remains a subset of {(x1 , k1 ), (x2 , k1 ), (x3 , k1 )}. But the values of c* assigned to these sets are now consistent with . Thus, by Theorem 1, Property α is still satisfied.

168 Culture and the explanation of choice behavior 0

Choose a k 0 in K and let X k = {(x, k 0 ) : x is in X }. Then there is a 1–1 0 0 correspondence between subsets S of X and subsets S k of X k such that 0

S k = {(x, k 0 ) : x is in S}. 0

(6) 0

Letting S(X k ) be the collection of nonempty subsets of X k , define c* for all S k 0 in S(X k ) by

0

0

c*(S k ) = c(S), 0

0

where S k and S are related by (6). Then c* does not satisfy Property α for S k and 0 0 T k in S(X k ) if and only if c does not satisfy Property α for the corresponding S and T in X . It will now be shown that by appropriately modifying the values of k in (x, k), the violations of Property α by c* can be eliminated. 0 0 Let S  (X k ) be the subcollection of S(X k ) on which c* satisfies Property α. 0 Then, as indicated earlier, S  (X k ) contains at least all one- and two-element 0 0 0 0 subsets of X k . Let S¯ k and T¯ k be a pair of nonempty subsets of X k with respect 0 to which Property α is violated (i.e. that are not in S  (X k )). Without loss of 0 0 0 0 0 0 generality suppose S¯ k ⊆ T¯ k , c*(T¯ k ) is in S¯ k , and c*(S¯ k )  = c*(T¯ k ). Evidently, 0 0 x, k 0 ) for some  x in X . Now change ( x, k 0 ) in S¯ k to ( x, k  ), keeping c*(S¯ k ) = ( 0 0 0 k   0 k ¯ ¯ x, k ), for some k  = k in K. Then S can no longer be a subset of T¯ k , c*(S ) = ( 0 0 and c* no longer violates Property α with respect to S¯ k and T¯ k . Moreover, since 0 0 0 x, k  ) cannot be an element of any S k in S  (X k ), Property α cannot c*(S¯ k ) = ( 0 0 0 be violated by c* with respect to any pair made up of an S k in S  (X k ) with S¯ k . 0 0 Hence S¯ k may be added to S  (X k ). Continue this process of changing the value of k in the pairs (x, k) until all 0 violations of Property α by c* have been eliminated and all sets of S(X k ) originally 0 0 0 outside of S  (X k ) have been modified and are now in S  (X k ). Because X k contains only a finite number of elements, K can always be made large enough to complete the process. 0 Observe, however, that although c* is defined on S  (X k ), c* is not yet a choice 0 function because its domain S  (X k ) is not a collection of all nonempty subsets of 0 a given set of objects. In particular, S  (X k ) is a proper subset of the collection of all nonempty subsets S(X × K) of X × K. But clearly c* can always be extended to S(X × K) in such a manner that Property α remains satisfied everywhere. By Theorem 2, then, there exists a relation rationalizing c* over S(X × K). Evidently, c is embedded in c* in the sense described previously. Q.E.D. Thus, by selecting an appropriate dimensional context, all choice behavior is explainable in terms of rationality as defined by (1). In particular, choices driven by, say, loyalty or obligation, and that lead to the selection of what might be regarded

Culture and the explanation of choice behavior 169 as less preferred alternatives in a different system of explanation, can be interpreted as perfectly rational actions. But what purpose is served by doing so? To explain the economic behavior of a Western individual, the general practice is to abstract from the myriad of motives that might determine his behavior and focus on that one force, namely self-interest or the pursuit of the most preferred, thought to be the most significant. Indeed, although, as described earlier, empirical confirmation is not possible, there is still a strong cultural justification for making self-interest, and the rationality reflecting it, the heart of such an analysis. The same procedure of honing in on what appears to be the most significant propellant of human behavior ought to apply to the explanation of behavior in every culture. And since choice behaviors in different cultures are clearly motivated by different forces, to preserve rationality in the sense of (1) when its preservation is unwarranted is not only to possibly misrepresent and mislead, but it also reduces understandings of choice behavior to one rather bland activity. The richness of human motivation in directing action is totally lost in the effort to condense and simplify. And a dubious or at least a secondary motive is substituted for one or more primary ones. Clearly the use of rationality in models when that motive is not, in actuality, the main driving force of action, substantially reduces the explanatory significance and relevance of the model, and its power to explain real choice activity.

II Consider, now, the possibility of explaining choice behavior in reference to a cultural imperative other than self-interest. Again let X be a finite set of objects, a preference relation defined on X , and S(X ) the collection of all nonempty subsets of X . Continue the assumption that X contains more than two objects, i.e. that I ≥ 3. Suppose each S in S(X ) is identified with a unique, nonempty S0 ⊆ S. When S contains only a single object x, the set S0 corresponding to it is S0 = S = {x}. Generally, the same S0 could be linked to more than one S in S(X ). Denote the collection of all sets S0 associated with some S in S(X ) by S0 (X ). Then S0 (X ) ⊆ S(X ). Define the function c on S0 (X ) as follows. For all S0 in S0 (X ), where S0 contains more than one object, c(S0 ) = x0 if and only if x0 is in S0 and x0 x for every x = x0 in S0 .

(7)

When S0 = {x} for some x in X , write c(S0 ) = c({x}) = x. As a rule, c on S0 (X ) cannot be a choice function because the domain S0 (X ) does not contain all nonempty subsets of a given set. However, for the special case in which the S0 = S, that is, in which S is identical to its associated set S0 for all S in S(X ), the collections S0 (X ) and S(X ) are the same, c is a choice function, and c is rationalized by . Returning to the general circumstance where S0 (X )  = S(X ), extend c to S(X ) by setting, for all S in S(X ), c(S) = c(S0 ),

(8)

170 Culture and the explanation of choice behavior where S0 is the set identified, as indicated above, with S. Then c defined on S(X ) is always a choice function, and S0 (X ) and are said to generate c. Typically, c on S(X ) is not rationalizable. But, obviously, for any c on S(X ) that is rationalizable, there exists an S0 (X ) and that generates c. It is clear that, in the present case, the choosing agent makes his choices from any S in S(X ) on the basis of a two-step procedure. First he selects the subset S0 of S identified with S; then he selects as his choice from S the most preferred element of S0 . The selection of S0 may be interpreted as an expression of the primary motive in determining choice action in that it eliminates from consideration those options that are less attractive, and therefore unacceptable, for cultural or other non-preference-related reasons. Moreover, it is only within the (possibly severe) cultural limitations or boundaries characterized by S0 that rationality or self-interest founded on preferences is permitted to operate. Thus, except for the extreme circumstance taken up in the next section where all of the S0 are single-object sets, self-interest is always present although, in general, it is secondary to and constrained by alternative and more significant cultural motivations. As it was described in Section I, the choice function relating to the cake example can be viewed as a special case of that defined by (7) and (8). This is accomplished through a particular specification of the sets S0 in S0 (X ): Recall S0 = {x0 } is associated with single-object sets S = {x0 }. When S contains two or more objects, take the associated S0 to be S0 = S − {x0 }, where x0 is in S and x0 x for all x in S0 .

(9)

Evidently, for each S having more than one object, S0 , in conformity with (9), discards the most preferred or best object in S under and hence, according to (7) and (8), c selects the second-best object from S. It has already been shown in the previous section that, with respect to two- and three-object sets, c violates Property α. Therefore c is not rationalizable. The formal mathematics of the second-best selection situation has been studied in general, and in a “first-best” (rationality) context, by Biagent and Gaertner [1]. In the present more specialized and simplified environment, choice functions arising in secondbest selection situations can be characterized in terms of the two properties introduced next. Property 1. For all distinct x , x , and x in X , if c({x , x }) = x and c({x , x }) = x , then c({x , x , x }) = x . Property 2. For all S in S(X ) containing two or more objects of X , if c(S) = x0 , then there is an x = x0 in S such that c({x , x}) = x for all x in S different from x .9

9 This property appears in Biagent and Gaertner [1] as their Axiom 3.3.

Culture and the explanation of choice behavior 171 Of course, Property 1 says that if, in pairwise choice, one object is selected over another and rejected over a third, then out of the set of the three objects, the first is selected. When S0 (X ) and generate c in a second-best selection situation, this property is, in part, a reflection of the transitivity of preference among the objects of X . Property 2 asserts that if x0 is selected from S, there is a different x in S that is always rejected whenever a pairwise choice is made between x and another object (including x0 ) of S. With S0 (X ) and generating c, this rejected object turns out to be preferred to all other objects in S. The precise relationship between S0 (X ) and on the one hand, and c on the other, is established by the two propositions below. Theorem 4. Let X and on X be given. Assume I ≥ 3 and suppose the S0 in S0 (X ) are characterized by (9). If c is generated by S0 (X ) and , then c satisfies Properties 1 and 2. Proof : Suppose c is defined by (7) and (8), and let c({x , x }) = x and for distinct x , x , and x in X . Then invoking (7) and (8), x x   and x x . By the transitivity of , x x . Hence, setting S = {x , x , x }, the S0 associated with S is, from (9), S0 = {x , x }. Now, using (7) and (8) again, c({x , x , x }) = x . This proves Property 1. To obtain Property 2, suppose c(S) = x0 for some S in S(X ), where S contains more than two objects. Then by (7)–(9), there is an x = x0 in S such that the S0 identified with S is S0 = S − {x }, and x x for all x (including x0 ) in S0 . For each of the latter x, a second application of (7) and (8) yields c({x , x}) = x. Q.E.D. c({x , x }) = x

Theorem 5. Let X be given with I ≥ 3. Suppose c is a choice function defined on S(X ) that satisfies Properties 1 and 2. Then there exists a preference relation on X and a collection, S0 (X ), of subsets S0 of X satisfying (9) that, together, generate c. Proof : With c given, associate to single-object sets S = {x} in S(X ) the set S0 such that S0 = S = {x}. When S contains two or more objects of X , observe that c(S) = x0 for some x0 in S. By Property 2, there is an x  = x0 in S such that c({x , x}) = x for all x = x in S. To this S associate the set S0 = S − {x } . Take S0 (X ) to be the collection of all S0 . Define on X by writing, for all x and x in X , x x if and only if there exists an S in S(X ) such that x and x are in S, and x is in the associated S0 while x is not. Then, since S(X ) contains all sets of distinct pairs x and x , where x and x are in X , and since the domain of c is all of S(X ), the relation is total on X . It is now shown that is irreflexive and transitive. First, irreflexivity of follows from the fact that x  = x in the definition of . To obtain transitivity, let x x and x x for some x , x , and x in X . Then there exists an S in S(X ) such that x and x are in S, x is in the associated S0 and x is not, and hence, for every x  = x in S, c({x , x}) = x. Likewise, there exists a

172 Culture and the explanation of choice behavior T in S(X ) such that x and x are in T , and for every x  = x in T , c({x , x}) = x. In particular, c({x , x }) = x and c({x , x }) = x .

(10)

By Property 1, c({x , x , x }) = x . Using this last fact together with S = {x , x , x } in Property 2, either (a) c({x , x }) = x and c({x , x }) = x , or (b) c({x , x }) = x and c({x , x }) = x . But the second part of (b) contradicts (10). Thus (a) must hold. It follows by using S = {x , x , x } in the definition of that x x and, hence, that is transitive. Finally, for any S where, say, S0 = S − {x }, observe that x x for all x in S0 , meeting, thereby, the requirements of (9). Moreover, (7) and (8) are also satisfied. Therefore and S0 (X ), together, generate c. Q.E.D. Just as Property α characterizes choice functions that are rationalizable, so do Properties 1 and 2 characterize nonrationalizable choice functions that are generated by a preference relation together with a cultural imperative expressed in the S0 of S0 (X ). It is the latter cultural imperative that prevents the selection of the most preferred object from any S in S(X ) and results in what may be referred to as a second-best choice function defined on S(X ). Of course, and in analogy with Theorem 3, one might still think of second-best selection as first-best selection subject to the constraint of deleting the most preferred choice. But since the latter constraint is nontraditional in its dependence on both the set from which the choice is made, as well as on the preferences among that set’s elements, such an approach does not reinstate the traditional model of rationality. Moreover, it reduces the significance of one motive in determining behavior (namely, that leading to the sets of S0 (X )) by treating that motive as a constraint, and it promotes another motive (namely, self-interest) as the driving force behind human behavior. Clearly, for circumstances in which self-interest is not the main propellant of behavior, the explanation that results is less accurate and relevant than it might be. Returning to the general context in which second-best selection has been discussed, it is also possible to similarly characterize third-best choice functions, fourth-best choice functions, and so on, as long as X contains a sufficiently large number of objects. To provide one illustration, consider a third-best situation. Assuming that I ≥ 5 and that is defined on X , obtain the sets of S0 (X ) from the following substitute for (9): To single-object sets S = {x }, associate S0 = S = {x } as before. When S = {x , x } and, say, x x , take the associated S0 to be S0 = {x }. Finally, when S contains three or more objects, set S0 = S − {x , x }, where x and x are in S, and x x and x x for all x in S0 .

(9 )

Culture and the explanation of choice behavior 173 Define the choice function c on S(X ) as in (7) and (8). Replace Properties 1 and 2 by their respective generalizations as described below: Property 1
. For all distinct x0 , x¯ , x , x , and x in X , if c({x0 , x , x }) = x , and c({¯x, x , x }) = x , then c({x0 , x , x }) = x , c({x , x , x }) = x and c({x0 , x , x , x }) = x . Property 2
. For all S in S(X ) containing three or more objects of X , if c(S) = x0 , then there exist x  = x0 and x = x0 in S such that x  = x and c({x , x , x}) = x, for all x in S (including x0 ) different from x and x . The counterparts to Theorems 4 and 5 are, respectively, Theorem 4
. Let X and on X be given. Assume I ≥ 5 and suppose the S0 in S0 (X ) are characterized by (9 ). If c is generated by S0 (X ) and , then c satisfies Properties 1 and 2 . Theorem 5
. Let X be given with I ≥ 5. Suppose c is a choice function defined on S(X ) that satisfies Properties 1 and 2 . Then there exists a preference relation on X and a collection, S0 (X ), of subsets S0 of X satisfying (9 ) that, together, generate c. In light of (7), (8), and (9 ), the proof of Theorem 4 is straightforward. The proof of Theorem 5 has much in common with that of Theorem 5. Of course, in this case the S0 in (9 ) have to be redefined in terms of c. Thus, with S = {x }, again define the corresponding S0 to be S0 = S = {x }. Associate to S = {x , x } the set S0 = {c({x , x })}. When S contains three or more objects, then c(S) = x0 , for some x0 in S. By Property 2 , there are x  = x0 and x  = x0 in S such that x  = x and c({x , x , x}) = x, for all x in S different from x and x . For such S, take S0 = S − {x , x }. Defining exactly as in the proof of Theorem 5, the structure of the demonstration of its transitivity is similar to that of the transitivity argument in Theorem 5.

III Choice functions that arise in reference to collections S0 (X ) , although characterized in terms of a primary cultural motive other than rationality or self-interest, still assign a limited role to self-interest within the sets S0 in S0 (X ). But for the case in which all S0 in S0 (X ) are single-object sets, the second step of the choosing agent’s two-step choice procedure is reduced to a triviality. Here rationality, in any significant sense, disappears completely from the decision-making mechanism, and a non-self-interest motive or cultural imperative is, for all practical purposes, the sole determinant of choice behavior. This remains so even when, given the preference relation on X , for each S containing two or more objects, S is identified with S0 = {x0 }, where x0 is in S and x0 x for all x  = x0 in S. For although

174 Culture and the explanation of choice behavior the choice function c generated by such S0 (X ) and satisfies Property α and is rationalizable, the definition of c in terms of (7) and (8) still does not allow the individual to make self-interested choices among two or more options. Since all sets of S0 (X ) contain only single objects, and since these objects are externally and culturally determined, there is no opportunity for rational or self-interested action on the part of the individual. To illustrate how a cultural imperative can completely eliminate self interest from the choice-decision process, imagine a person voting in an election for one of several candidates. Suppose the individual has specific preferences among the candidates. But suppose also that, for cultural reasons, this person feels he has an overriding obligation to a friend to vote for the candidate preferred by that friend. (Such obligations are not uncommon in, for example, Japan.) Under these conditions, the individual’s choice among candidates is attenuated to the selection of a candidate from a single-candidate set. A more interesting example of the reduction of choice to selecting from singleobject sets is the decision rule that requires the selection, from each S in S(X ), of the “median” or “central” object under . The formalities of this mechanism have been studied in general and in a setting of rationality by Gaertner and Xu [2]. Median-selection choice functions are now analyzed in the more specialized and simplified context of the present essay. Before doing so, however, it is worth illustrating how median selection might arise in a “real” setting. Imagine an individual voting for one of several alternatives in a situation in which his vote will be made public. Suppose that these alternatives line up on the political spectrum from left to right, and that the individual’s preferences line up similarly with the most preferred alternative at either one end of the spectrum or the other. Assume also that the individual’s cultural environment is such that there is considerable danger of condemnation and ostricization if he does not vote in the center of the political spectrum. Under such circumstances, the individual may well employ the median-selection rule when making his voting decision. With the preference ordering on X given, and with the number I of objects in X at least 3, consider an S in S0 (X ). Let n be the number of objects in S. If n = 1, then take, as usual, the S0 associated with S to be the single-object set S. When n ≥ 2, assume without loss of generality that x1 x2 · · · xn , and assign S0 to S as follows: ⎧ ⎨ {x(n+1)/2 }, if n is odd, (11) S0 = ⎩ if n is even. {xn/2 }, Refer to x(n+1)/2 or, as the case may be, xn/2 , as the median or central object of S. Note that when n is even and there are two objects in S that are “centrally” located along the ordering of S under , the median of S is designated as the more preferred of the two objects. Define c on S(X ) according to (7) and (8). Then c selects the central object out of each S in S(X ). Observe that, for the first

Culture and the explanation of choice behavior 175 three objects, say, of the sequence identified above, namely, x1 , x2 , and x3 , this median-selection choice function has the characteristic, in violation of Property α, that c({x1 , x2 , x3 }) = x2 and c({x1 , x2 }) = x1 . Therefore c is not rationalizable. Two possible properties of c different from Property α are of interest here: Property 3. For all distinct x , x , and x in X , if c({x , x }) = x and c({x , x }) = x , then c({x , x }) = x . Property 4. For all S in S(X ) containing three or more objects of X , if c(S) = x0 for some x0 in S, then there exist x = x0 and x  = x0 in S such that x  = x , c(S − {x , x }) = x0 , c({x , x}) = x for all x = x in S, and c({x , x}) = x for all x  = x in S. Note Property 3 asserts that in pairwise choice, if one object is selected over another, and if the rejected object is selected over a third, then the first is selected over the third. Like Property 1 (but in a different context), when S0 (X ) and generate c in a median selection situation, Property 3 is partly an expression of the transitivity of in terms of c with respect to pairs of objects in X . Property 4 says that, with S0 (X ) and generating c, the choice from S does not change when both the “highest” object above x0 in the ordering of S under and the “lowest” object below x0 are removed from S. It is now possible to characterize median-selection choice functions in terms of Properties 3 and 4. The proof of the first of the two propositions required, involving obvious restatements of the above-stated interpretations of the two properties, is rather straightforward and not given here. Theorem 6. Let X and on X be given. Assume I ≥ 3 and suppose the S0 in S0 (X ) are characterized by (11). If c is generated by S0 (X ) and , then c satisfies Properties 3 and 4. Theorem 7. Let X be given with I ≥ 3. Suppose c is a choice function defined on S(X ) that satisfies Properties 3 and 4. Then there exists a preference relation on X and a collection, S0 (X ), of subsets S0 of X satisfying (11) that, together, generate c. Proof : Let c be given. Then for each S in S(X ), associate with S the set S0 = {c(S)}. Take S0 (X ) to be the collection of all such S0 . For all distinct x and x in X , write x x if and only if c({x , x }) = x . Since S(X ) contains all sets of distinct pairs x and x , where x and x are in X , and since the domain of c is all of S(X ), the preference relation is total on X . It is irreflexive because the definition of requires x = x , and its transitivity follows immediately from Property 3. It remains to show that (11) is satisfied. Now, evidently, (11) holds for all S in S(X ) containing exactly two objects of X . Consider any S with n ≥ 3 objects. By repeated application of Property 4 there is a nested sequence of subsets of S, say S 1 ⊆ S 2 ⊆ · · · ⊆ S J , such that S 1 contains

176 Culture and the explanation of choice behavior 



one object if n is odd and two objects if n is even, S j = S j+1 − {x j , x j } for some   distinct x j and x j in S j+1 , the largest set S J = S, and c(S j ) = x0 for j = 1, . . . , J .   Moreover, if n is odd, then x j x0 x j , and if n is even so that S 1 contains   two objects, say, S 1 = {x0 , x } , then x j x0 x  x j . Since this is so for each j = 1, . . . , J − 1, object x0 must be the median object in S. Hence (11) holds, as do (7) and (8). Therefore S0 (X ) and generate c. Q.E.D. The main thrust of this essay has focused on the properties of behavior that relate to rational, and certain forms of nonrational choice. Since the number of choice options is finite, the problem of testing for evidence of any of these properties is straightforward. It is only required to present all possible subsets of the collection of choice options to the individual and ask him to select from each. The results of his selections can then be compared with the relevant properties. However, deviations of such observed data from meeting the requirements of, say, Property α as it was originally stated above could arise from cultural imperatives that prevent it from being satisfied, from other violations of the assumptions of the rational choice model that, with respect to the case in question, are not thought to be culturally linked (perhaps, for example, intransitivities in preferences), from modifications in the individual’s preferences that occur between his selections, and from nonsystematic errors or inconsistencies in decision-making. That is, deviations of observed behavior from Property α do not necessarily imply deviations from rationality and the cultural imperative of self-interest. Likewise, an absence of observed differences in behavior across cultures does not necessarily signify that there are no cultural differences – only that whatever cultural differences are present are not sufficient to produce differences in observed behavior. Thus, although it is possible, at least in principle, to see if (or how well) observed choice satisfies Property α or any of the pairs of Properties 1 and 2, 1 and 2 , or 3 and 4, it is still not possible, as suggested with respect to preference optimization at the outset, to know if satisfaction (or “near” satisfaction) of any of these means that the motivational hypotheses generating them or the cultural backgrounds relating to them are actually in force. Moreover, even if the data appear to be inconsistent with Property α, it is always possible, in accordance with Theorem 3, to expand the dimensionality of the data (perhaps through questioning of the individual involved) and interpret it as a manifestation of rational behavior. Regardless, from the results of this section and those of Section II, it is clear that nonrationalizable choice functions can be characterized in terms of concrete cultural imperatives defined with respect to the S0 of S0 (X ) that have little or nothing to do with rationality or self-interest. And the latter cultural imperatives are “equivalently” realized in terms of the properties of the observable choice functions they spawn. In particular, Properties 1 and 2, 1 and 2 , and 3 and 4 represent choice behavior derived, respectively, from second-best, third-best, and central-option selection. There is no reason, then, why models based on such cultural imperatives cannot be used, when appropriate, in explaining observable

Culture and the explanation of choice behavior 177 choice behavior. Indeed, where rationality in the sense of (1) is not the primary driving force behind human action, models such as these would provide a more relevant, more significant, and more interesting explanation than that obtained from any model that would require the artificial imposition of rationality.

References 1. Biagent, N. and W. Gaertner, “Never Choose the Uniquely Largest: A Characterization,” Economic Theory 8 (1996), pp. 239–249. 2. Gaertner, W. and Y. Xu, “On Rationalizability of Choice Functions: A Characterization of the Median,” Social Choice and Welfare 16 (1999), pp. 629–638. 3. Greif, A., “Cultural Beliefs and the Organization of Society: A Historical and Theoretical Reflection on Collectivist and Individualist Societies,” Journal of Political Economy 102 (1994), pp. 912–950. 4. Hampden-Turner, C. and A. Trompenaars, The Seven Cultures of Capitalism (New York: Doubleday, 1993). 5. Katzner, D.W., “Western Economics and the Economy of Japan,” Journal of Post Keynesian Economics 21 (1999), pp. 503–521. This volume: Essay 2. 6. Lancaster, K.J., “A New Approach to Consumer Theory,” Journal of Political Economy 74 (1966), pp. 132–157. 7. North, D.C., “Institutions and Economic Theory,” American Economist 36, no. 1 (Spring, 1992), pp. 3–6. 8. Roth, A.E., V. Prasnikar, M. Okuno-Fujiwara, and S. Zamir “Bargaining and Market Behavior in Jerusalem, Ljubljana, Pittsburgh, and Tokyo: An Experimental Study,” American Economic Review 81 (1991), pp. 1068–1095. 9. Sen, A.K., “Choice Functions and Revealed Preference,” Review of Economic Studies 38 (1971), pp. 307–317. 10. ——, “Internal Consistency of Choice,” Econometrica 61 (1993), pp. 495–521.

Index

Abe, S. 100, 118 Abegglen, J.C. 38, 41, 91, 99, 106n, 107n, 111n, 118, 148, 149, 154, 159 Acceptance of less-than-maximum profit (Japan) 111, 113; empirical support for 113, 116–117 Activities (in firms) 150 Advertising 75 American economy 63–64; understanding of 97 Analysis in response to problems 3–4 Ando, A. 116, 118 Aoki, M. 18n, 19, 31, 32, 100, 149, 154, 159 Arrow, K.J. 124, 129 Auerbach, A.J. 116, 118 Axomatic-deductive approach; see model-building approach Bacon, N.T. 31, 130 Bailey, D. 82n Bardhan, P. 16, 60 Barrett, J. 131n Becker, G. xii, 4, 16, 18, 32, 43, 59, 84, 99, 136n, 145 Bellah, R.N. 19n, 24, 31, 38, 41, 50n, 54n, 59, 85, 99, 117, 118, 126, 129 Benedict, R. 25n, 31, 55n, 59, 85, 99, 105n, 117, 118, 148, 159 Bentham, J. 23, 31 Best next move 145 Biagent, N. 163n, 164n, 170, 177 Blinder, A.S. 148, 159 Boards of directors 90–91, 111 Bowles, S. 15, 42n, 59, 130, 148n, 159 Boyd, R. 15, 59, 130 Brain 44, 47 von Buquoy, G. 121 Bureaucracy 156n

Cake example (Sen) 127, 165–166 Caldwell, B.J. 21, 31, 136, 145, 147, 159 Camerer, C. 15, 59, 130 Catastrophism 132, 141 Cavalli-Sforza, F. 44, 59 Cavalli-Sforza, L.L. 44, 59 Chamberlin, E.H. xi Characteristics of choice objects 165–168 Choice 142; and uncertainty 142; best option (most preferred) 165, 169; median- (central-) selection 174; role of culture in 142, 162; second-best 170–171; third-best 172; worst option 165 Choice behavior 163; and culture 162–163, 166; Japanese 162; rational 164 Choice function 163; embedded 165–168; generated 170; median- (central-) selection 174–176; nonrationalizable 170, 172, 175–177; rationalized 163–168; second-best 172; third-best 172 Clark, R. 105n, 118 Coffey, D. 39, 40, 41, 82n Cole, J.W. 42n Company family (Japan) 92 Comparative advantage 112; irrelevance for Japanese economy 112 Competition 67; among Japanese firms 87–88; non-ideal forms of 70, 75–77; perfect 69–71 Competitive ideal 83; see also Ideal economic system Consumer behavior; see also Economic behavior; in absence of information 128; model of 125–128, 136–137; satisficing 128

180 Index Consumption 53; American 51–53; Japanese 55 Cournot, A. xi, 3, 23, 31, 122, 123, 130 Cultural continuity 19 Cultural development 44–47; and physical maturation 45, 47 Cultural patterns 1, 44–46, 48 Cultural relativism 17n Cultural traits 1; dominant 6 Culture 1, 18, 43–44, 138; and academic disciplines 1–2; and aggregate activity 137; and American consumption 51–53; and choice 142; and consumer behavior 126, 137; and decisions 47; and discipline of economics 2–4, 56–57; and economic behavior 49, 58–59, 64, 162; and economic efficiency 117; and economic growth 117–118; and economic laws 138; and explanation 161, 169; and firm behavior 147–148, 156; and models 48; and motivations 47; and replication across space 144; and societal goal formation 129; and symbols 44, 46; and the brain 44; and the mind 44, 46–47; and thought 46; as a control mechanism 45 Darwin, C.R. 132 Davidson, P. 42n, 144, 145 Debreu, G. 3, 124, 129 Decision making 47, 142, 144–145; American consumers 64–65; American firms 65–66; American informational requirements for 66–67; and experience 139; history-creating 145; Japanese consumers 85–86; Japanese firms 86–87; Japanese informational requirements for 87 De Long, J.B. 18, 31 Directed reasoning 47, 58 Discipline of economics 2–4, 56–57, 119 Dispositions 46–47 Distribution system 94–95, 97 Dostoyevsky, F. xii Economic behavior 49, 147; American 58–59, 64–66, 126; and culture 49; and models 49; and motivations 49, 169; and situation 49; Japanese 58–59, 126–127, 162 Economic concerns of Americans 61–62 Economic concerns of Japanese 82

Economic growth 133; and culture 117–118; and inefficiencies 117; in firm models 148–149 Economic laws 133–134; across space, 143; across time 143; and culture 138; relating to aggregate activity 137; relating to individual action 135–137 Economics and physical science (history) 131–132 Efficiency 69, 103–104, 106; absence of in Japanese economy 97–98, 106, 111–112, 157, 159; absence of in US economy 75–76; and culture 117; economic 65; from production at minimum unit cost 75; input-mix 65, 104, 112; of allocations 72–74, 157–159; output-level 104, 112; profit 152; technological (productive) 65, 104 Empirical testing 21, 136, 176 Enkhbat, R. 119n Enterprise group (Keiretsu) 89–90, 96 Ergodic environment 144 Euler, L. 121 Explanation 5, 36, 43, 50, 134–135, 161, 169; and culture 161, 169; correct 5, 15, 36; erroneous 5, 15, 36–37, 40 External referent 162 Externality 77–78 Falacy of misplaced concreteness 124n Fallows, J. 28, 31, 84, 100 Fehr, E. 7n, 15, 59, 130, 148n, 159 Fermat, P. de 121 Firm 147–149; clan 156–159; differences between Japanese and Western 148, 154; Japanese 156, 159; selection function 152; targets 154, 157; Western 156–159 Fischoff, E. 33 Fisher, I. 131, 145 Frisi, P. 120, 121, 123 Fundamental analogy 48, 58 Gächter, S. 7n, 15, 148n, 159 Gaertner, W. 161n, 163n, 164n, 170, 174, 177 Geertz, C. 44, 45, 46, 47, 59 Gerlach, M.L. 19, 26n, 31, 148, 159 Gintis, H. 15, 59, 130, 149, 152, 154, 155n, 157, 160 Gossen, H.H. 122, 130 Gould, S.J. 133, 135, 145 Government 95; and Japanese economy 95–96; and US economy 71, 95–97

Index 181 Government failure 81 Grattan-Guinness I. 121n, 130 Greif, A. 17n, 31, 43, 59, 84, 100, 162, 177 Groenewegen, P.D. 120, 121n, 130 Gudeman, S. 8, 15, 17n, 31, 33, 37, 48, 59 Guiso, L. 4n, 15 Hampden-Turner, C. 43, 59, 162, 177 Harris, M. 17n, 31 Hashimoto, R. 56n Hatta, T. 17n Hay, D.A. 148, 160 Hayami, Y. 115n, 118 Henrich, J. 11, 15, 43, 54n, 59, 127, 130 Higa, T. 98, 100, 107n, 118 Higgott, R. 31 Historical Time 139–141; and choice 142; and knowledge 141; and laws of human behavior 141; and replication 144; and uncertainty 141–142; and uniformitarianism 141; and unusability of probability 144 History creation 142, 145 Horioka, C.Y. 17n, 19n, 31, 37n, 42n, 82n, 91, 100, 101n Hours worked 91–92 Hudson, A.B. 17n Ideal economic system 69, 71–72, 88; see also Competitive ideal; and cultural values 72, 74; efficiency of 72–74; nonwastefulness of 73; unbiasedness of 74 Igawa, K. 100, 118 Incentive function 151; nondecreasing 155 Incentive motivation 154 Inefficiencies and economic growth 117 Institutionalization of values 22; in Japan 24–25; in West 22–23 Insufficient information 79–80, 127–128 Internal equilibrium (firm) 153, 157–159 Internalize the values of the firm 155 Internalize the worth of productivity 155 Isoperimetric problem 120 Ito, T. 19n, 26n, 31, 40n, 41, 89, 91, 94, 96, 100, 148, 160 Jaffé, W. 130 Japanese economic miracle 101; causes 103 Japanese economy 84; understanding of 97 Jevons, W.S. 23, 31, 122, 130, 131, 145 Johnson, C. 28, 31, 84, 96, 100

Kachelmeier, S.J. 20n, 32 Kaleidic change 141 Katz, R. 102n, 118 Katzner, D.W. 5n, 15, 27n, 32, 33, 36, 37, 41, 43, 51n, 54n, 60, 64, 81, 84, 85, 98n, 100, 105n, 117n, 118, 124n, 127, 128n, 130, 137n, 141n, 144n, 145, 147, 149, 152, 154, 155, 156n, 157, 159, 160, 162, 177 Keiretsu; see Enterprise group Keynes, J.M. xi, xii, 3 Keynesian-cross American model 52, 55–56 Kinzley, W.D. 25, 32 Kirby, S. 17n, 32 Kline, M. 121n, 130 Knowledge in historical time 141–143; and the unusability of probability 144 Koike, K. 29, 32, 84, 100 Kusner, R. 119n Labor unions 71, 93–94 Lagrange, J.L. 121 Lancaster, K.J. 166n, 177 Large vs medium and small firms 65, 90, 96 Lau, L.J. 18, 32 Leaver, R. 31 Leff, N.H. 18, 32 Lévy-Garboua, L. 160 Lifetime employment 91, 96 Lines of authority (firm) 150 Localism 11–12, 33, 57 Logical time 139; and laws of human behavior 139; and philosophical materialism 139–140; and uniformitarianism 139; and use of probability 143 Long-term relationships 86, 95 Lupardus, K. 98, 100, 107n, 118 McElreath, R. 15, 59, 130 McGilvray, B. 130 Machlup, F. 22, 32, 124n, 130 Madsen, R. 59, 99 Mak, J. 100, 118 Market failure 76–81, 97 Markets in American economy 67–69 Marshall, A. 3, 21n, 32 Median (central) object 174 Mind 44, 46–47; and dispositions 46–47 Mirowski, P. 131, 132n, 145 Misallocation of resources (Japan) 98, 111 Miyazaki, H. 17n

182 Index Model-building approach 4 Models 48–49, 58; and cultural patterns 48; and economic behavior 49; and fundamental analogies 48; and thought 48; “as being” formulation 48–49; “as if ” formulation 48–49 Morishima, M. 29, 30, 32 Morris, J.D. 148, 160 Morris-Suzuki, T. 17n, 18n, 29n, 32 Motivations 47, 50, 169; and culture 47, 169; and economic behavior 49, 169 Motive 169, 172; as a constraint 172; primary 170 Movshuk, R. 114, 115, 188 Murakami, Y. 29, 30 Nakane, C. 25n, 26, 32, 38, 41, 85, 87, 88, 100, 105n, 117, 118, 148, 160 Nisbet, R.E. 6, 15, 49, 60 Non-ideal forms of competition 70, 76–77, 88 North, D.C. 17n, 32, 43, 60, 162, 177 Obara, T. 17n Obuchi, K. 56n Offsets to Japanese inefficiencies 98–99, 107–112 Ogasawara, J. 115n, 118 Okuno-Fujiwara, M. 60, 177 Optimization in economics 119; as defining an ideal 123, 128–129; in consumer and firm models 123–128; preference 162; see also utility maximization; used for convenience 123–124 Optimization in economics (history) 120–123; future importance 129; under constraints 122–123; why it became important 127 Oshima, H.T. 93, 100 Ouchi, W.G. 156, 160 Overhiring of workers (Japan) 98–99, 106–111, 113; empirical support for 113–115 Pálsson, G. 12, 15 Pardalos, P.M. 119n Pareto optimal (allocative efficiency) 72–74; internally in production 157–159 Parsons, T. 32, 60, 130, 146 Pepper, S.C. 48, 60 Philosophical materialism 138; and knowledge 142; and methodological uniformitarianism 139; and time 140

Pilat, D. 113, 114, 118 Polanyi, K. 22n, 32 Policy (economic) 55–56, 99, 144–145; decision-making and implementation 134–135, 144–145; formulation 134–135 Potential surprise 128 Prasnikar, V. 60, 177 Prediction 36–37, 43, 50, 134–135, 141–142 Preference optimization 162 Preference relation 163; irreflexive 163n; total 163n; transitive 163n Prestowitz, C.V. jr. 28, 32, 84, 100 Prices 9; in American economy 66–68; in Japanese economy 87–88 Principle of Least Action 121 Principle of Least Time 121 Principle of Uniformity; see Uniformitarianism Probability 143–144; and logical time 143; and methodological uniformitarianism 143; frequency 143; subjective 143 Production function 150 Profit efficiency 152 Profit function 151 Profit maximization 121–122; absence of in Japanese firm 39–41, 86, 105–112, 148, 152, 157; in American firm 65–66, 86; in Western firm 152, 154–157 Property α, 164–168 Property 1, 170–172 Property 1 , 173 Property 2, 170–172 Property 2 , 173 Property 3, 175 Property 4, 175 Pure capitalist economy 104 Pure capitalist firm 104 Public good 78–79 Quasi-capitalist economy 106 Quasi-capitalist firm 106 Questions 58 Ramseyer, J.M. 28, 32 Rational behavior 20–21, 29, 35, 125, 164, 169 Rationality 35, 164, 165, 168, 170, 173–174; see also Rationality postulates; Self-interest

Index 183 Rationality postulates 21–22, 25, 27, 29–30, 50–51, 147–148, 156, 159; cultural basis for 50 Ravenhill, J. 31 Reed, S.R. 30n, 32, 49, 60 Replication over time and space 144 Ricardo, D. 3 Richerson, P.J. 44, 59 Rivera, A. 48, 59 Robbins, L. xiin Robertson, R.M. 120, 121, 130 Robinson, J. xi, 42, 60 Rosenbluth, F.M. 28, 32 Ross, L. 49, 60 Roth, A.E. 43, 60, 162, 171 Rules (in firms) 150 Sahlins, M. 17n, 22n, 32 Salaries 92–93 Samuelson, P.A. 3, 123n, 130 Samuelsson, K. 22n, 32 Sapienza, P. 15 Sato, K. 18, 32 Saving (Japan) 55, 86 Schultz, T.W. 4, 16, 42, 60 Self-interest 20–30, 35, 51, 64, 86, 93, 125–126, 137, 147–148, 155, 162, 164, 169–170, 173–174; see also Rationality Sen, A.K. 25n, 27, 32, 43, 60, 162, 165, 177; cake example of 27, 165–166 Shackle, G.L.S. 128, 130, 140, 141, 145n, 146 Shehata, M. 20n, 32 Shepherd, W.G. 70, 81 Simon, H.A. 128, 130 Skott, P. 147n Smith, A. xi, xii, 2, 3, 16, 23, 32, 51n, 60 Spirit of capitalism 23 Stalk, G., jr. 38, 41, 91, 99, 106n, 107n, 111n, 118, 148, 149, 154, 159 Stigler, G.J. 4, 16, 18, 32 Stiglitz, J.E. 4, 16, 42, 60 Subculture 2, 44; in Economics 2–3 Sullivan, W.M. 59, 99 Sunder, S. 100, 118 Swidler, A. 59, 99 Symbols 44, 46–47, 58

Thorne, S. 59 Thought 47, 58; affective 47; and models 48; directed reasoning 47 Time; see Historical time; Logical time Tipton S.M. 59, 99 Tomlinson, P.R. 39, 40, 41, 82n Transformation function 124n Trompenaars, A. 43, 59, 162, 177 Tseveendorj, I. 119n Tylor, E. 1, 16 Uncertainty in historical time 141 Uniformitarianism 132–133, 143; and economics 133–135, 143; and philosophical materialism 140; methodological 14, 133, 135, 137, 139, 143; substantive 14, 133, 139 Universalism 8–11, 33, 57 Utility function 153 Utility maximization; see also Preference optimization; absence of in Japanese firms 153; by consumers 21, 125, 136; in Western firms 153–154 Values (cultural) 54; American 50–51, 64, 66, 72, 74, 76, 85, 91, 93, 137; Japanese 24–25, 54–55, 85–87, 91, 126, 137, 148 Verri, P. 120, 121, 130 Vickers, D. xiii, 1n, 17n, 37n, 42n, 61n, 82n, 101n, 119n, 131n, 140, 141n, 145, 146, 147n Walras, L. xi, 3, 122, 130 Waste in Japanese economy 112 Waste in US economy 75–81; as unemployed resources 80; from government failure 81 Weber, M. 22, 23, 24n, 32, 33, 51n, 60, 126, 130, 146, 147 Weierstrass Theorem 124 Western economics 20 White- vs blue-collar workers 92, 97 Window watchers 92 van Wolferen, K. 18n, 33 Women workers 92, 97 Xu, Y. 163n, 164n, 174, 177 Yotopoulos, P.A. 18, 32

Takeuchi, K. 17n Temin, P. 28, 32 Thiruvadanthai, S. 33, 34, 35, 36, 37

Zamir, S. 60, 177 Zingales, L. 15