Worker satisfaction and economic performance : microfoundations of success and failure 0765605910

Citation preview

WORKER SATISFACTION AND

ECONOMIC PERFORMANCE Microfoundations of Success and Failure

MORRIS ALTMAN

oM.E. Sharpe Armonk, New York London, England

Copyright © 2001 by M. E. Sharpe, Inc. All rights reserved. No part of this book may be reproduced in any form without written permission from the publisher, M. E. Sharpe, Inc., 80 Business Park Drive, Armonk, New York 10504.

Library of Congress Cataloging-in-Publication Data Altman, Morris. Worker satisfaction and economic performance : microfoundations of success and failure / Morris Altman, p. cm. Includes bibliographical references and index. ISBN 0-7656-0591-0 (alk. paper) 1. Employee motivation. 2. Job satisfaction. 3. Performance. 4. Economics. 1. Title. HF5549.5M63 A57 2001 338.5—dc21

2001032066 CIP

Printed in the United States of America The paper used in this publication meets the minimum requirements of American National Standard for Information Sciences Permanence of Paper for Printed Library Materials, ANSI Z 39.48-1984.

BM(c)

10

9

8

7

6

5

4

3

2

1

To Louise

Contents

Foreword Shlomo Maital Acknowledgments 1. Introduction: Revising the Micro foundations of Economics

ix

xvii 3

2. Human Agency as a Determinant of Material Welfare

26

3. The Methodology of Economics and the Survival Principle Revised

50

4. A Behavioral Theory of Economic Welfare and Economic Justice

69

5. The Economics of Exogenous Increases in Wage Rates in a Behavioral/X-Efficiency Model of the Firm

103

6. A Behavioral Model of Endogenous Economic Growth

119

7. Interfirm, Interregional, and International Differences in Labor Productivity: Why Convergence Need Not Take Place

137

8. The Economics of Profitable Inefficiency and Market Failure: A Behavioral Model of Path Dependency

150

9. Economic Theory, Public Policy, and the Challenge of Innovative Work Practices

170

10. The Efficiency- and Welfare-Promoting Role of Labor Rights and Labor Power in a Market Economy

182 vii

viii

CONTENTS

11. A Revisionist View of the Economic Implications of Child Labor Regulations

199

12. How Discriminatory Pay Inequality Can Persist: Even in Competitive Markets

216

13. When Green Isn’t Mean: The Economics of Environmental Regulations

230

14. Big Is Not Always Better: A Critical Appraisal of the Transaction Cost—Economizing Paradigm

247

15. Culture as a Determinant of Material Welfare

260

References

271

Index

289

About the Author

299

Foreword

Morris Altman has written a timely book. His theme is this: “How workers are treated in the context of the firm’s general human relations or industrial relations environment plays a fundamentally important role in determining the economic performance of the firm and, thereby, of the economy as a whole.” Put simply: firms, economies, and societies who treat workers well and pay them good wages will ultimately thrive and prosper. While most of us find this proposition appealing on normative grounds, Altman uses be¬ havioral economics to refute neoclassical conventional wisdom and build positive foundations for his claim. For this reason, an additional subtitle for his book would be appropriate: “A Manifesto for Fairness and Equality.” There are two reasons why Worker Satisfaction and Economic Perfor¬ mance is today particularly relevant. First is the theory of the overwhelming triumph of American free market capitalism as the world’s sole principle of social organization. The theory that legitimizes and justifies enormous, in¬ tolerable injustice and inequality worldwide deserves refutation. Second is the reality of the insufferable inequality and unfairness in the distribution of wealth and income both within countries—^wealthy and poor alike—and among countries, which demands a manifesto showing why this is not only unmoral, but also, in the long run, inefficient and wealth-destroying. Altman focuses on the core of this inequality: how workers are paid, how well they are treated, and why it pays to pay them well.

Darwin or Isaiah? The fall of the Berlin Wall on November 9, 1989, signaled the final victory of American capitalism worldwide. Wall Street invented the term “emerging markets” and imposed a new set of mles: Developing countries must open ix

X

FOREWORD

their capital markets to foreign investments and eliminate controls. As Ameri¬ can capitalism triumphed, so did the Chicago neoclassical theorists who built its ideology. Miller, Coase, Becker, Fogel, and Lucas won five of six Nobel Prizes in economics between 1990 and 1995. Ironically, it was former secre¬ tary of the treasury Larry Summers, a modem Keynesian, who best expressed the triumph of the Chicago ideology. In an interview. Summers (2000) ex¬ plained how the world had changed. The old economic model, he explained, was based on Newtonian thinking, which says there are causal laws that govern markets and determine policies. The new economic model. Summers claimed, is based on Darwin. From time to time, “mutations” occur. Most are unsuccessful; a few are successful, and the successful ones generate enor¬ mous wealth. The economic system must be open and flexible to permit these mutations. In The Origin of Species, Darwin concluded by saying, “From the war of nature, from famine and death—the production of the higher animals [people] directly follows.” I reject Darwin both as a positive principle for social orga¬ nization and as a normative principle for advancing human well-being. Try instead Isaiah: “Learn to do good, seek justice, relieve the oppressed; defend the fatherless; plead for the widow” (Isaiah 1:17). Which society will be more prosperous: Darwin’s or Isaiah’s? Morris Altman shows Isaiah’s. All other things being equal, which society would you rather live in? Isaiah’s, hands down. The Rise of Global Inequality Globalization has generated enormous amounts of wealth for a tiny fraction of the world’s population. Of the world’s 6 billion people, the richest billion, living in about twenty-five countries, get 78 percent of the world’s GDP and enjoy average incomes of $25,000 yearly. There, only 7 of every 1,000 chil¬ dren die before age 5. The poorest 1.5 billion people get an average of $1 a day, or about $400 a year; in fifteen years there will be almost 2 billion people living at this starvation level. The poorest 3 billion people, living in sixty-one countries, get only 6 percent of the world’s GDP and have per capita GDP of $2 a day or less. There, 90 of every 1,000 children die before age 5. An entire continent—^Africa—has become deeply impoverished dur¬ ing the past generation. Since 1970, the wealthiest sixty countries doubled their per capita income. The middle sixty and poorest sixty countries gained little or no ground (World Bank 1999). Nothing is more symbolic than the twin problems of obesity and hunger. A billion people in the world are hungry. An equal number are obese. Both groups suffer. A transfer of resources from the fat to the thin would help

FOREWORD

xi

both. (In the United States, 30 percent of the population is overweight.) Yet in an age of supercomputers, gene therapy, and Internet, more people go hungry, even more people grow obese, and the absurdity is more tragic than comic. The good news is that globalization has created unprecedented wealth. The bad news is that only a handful have gotten it. Why? Inequality in the Home of Capitalism “Theory is the lens through which we perceive and make sense of the world around us,” Altman writes. The neoclassical theory showing that Darwinian survival of the fittest, among workers and firms, is a “given” has profound implications, Altman argues. It generates the belief that “living in a global and competitive world economy requires adherence to low-wage paths of eco¬ nomic development combined with minimalist government.” This is a sure¬ fire recipe for inequality and unfairness. It has created enormous inequality in the home of capitalism itself: America. America chose the “low road.” According to the World Development Indicators 2000 (World Bank 2000, 3), between 1990 and 1999, the total value of financial assets listed on the world’s stock markets tripled, rising from $9,400 billion to $24,458 billion. Fully two-thirds of that $15,000 billion increase in wealth was created in the United States, where market capitalization rose from $3,100 billion (1990) to $13,500 billion (1999). According to Forbes Magazine, America now has 189 persons whose net personal wealth exceeds $1 billion! America has 20 million persons whose net worth now exceeds $1 million. Many of America’s globalization policies were framed when Robert Rubin was U.S. treasury secretary. Recall that Rubin came to the Treasury from the leading Wall Street investment bank Goldman, Sachs. Joseph Stiglitz (2000), Stanford University professor and until recently chief economist and vice presi¬ dent of the World Bank, asks rhetorically, “Did America push its policies ... because we believed they would help East Asia (and other nations)? Or be¬ cause we believed they would benefit financial interests in the United States?” The answer lies in the numbers. Two-thirds of the increase in market value of financial assets between 1990 and 1999 occurred in American stock markets. Yet according to the UN Human Development Index, • America has the highest level of “human poverty”—life expectancy, illiteracy, and underemployment—among industrialized nations. In the United States one in every five adults is functionally illiterate (unable to read at a fifth-grade level). • Some 17 percent of the U.S. population has income less than half the country’s median.

xii

FOREWORD

• The wealthiest 1 percent of Americans own fully 40 percent of all the wealth. • The wealthiest 10 percent of Americans have average incomes seven¬ teen times higher than the poorest 10 percent. • Forty-two million Americans (one person in every seven) have no health insurance at all, up from 30 million just eight years ago. Why are both poverty and wealth rising in America? Because Americans, whose culture stresses fair play, seem unwilling to extend fairness to the business realm, by taxing the rich to help the poor. The United States has the lowest top marginal income tax rate of any industrial nation: 39.6 percent, a rate paid by Bill Gates and other billionaires. Capitalism Sinks Russia The triumph of neoclassical economic ideology, in its most single-minded form, has not only tom the fabric of American society, but ravaged the economy of Russia as well. According to Joseph Stiglitz (2000), after the fall of the Berlin Wall, two schools of thought emerged in the United States about how Russia should make a transition to a market economy. One group stressed that Russia must first build the infrastructure of a market economy: land titles offices, courts to enforce contracts, limited companies, and regulatory bodies. In this group were Nobel Prize winners, such as Kenneth Arrow. This group favored a gradual transition, with strong measures to fight inflation. A second group, comprised of American macroeconomists who knew nothing about Russia’s history or society, advocated “shock therapy”—^immediate, rapid transition to a market economy. They believed that the principles of market economics are universal and apply everywhere and anywhere. They won. The result? Russia today is a poor. Third World country. There was a lot of shock—but no therapy. Criminal elements plundered Russia’s wealth; the middle class had its savings destroyed twice—^in 1991 and again in 1998; and many fear Russia is ripe for a fascist dictatorship. It was not Russia’s economists who ruined Russia, but rather America’s. Russia was not ready for free markets and was misled by classroom economists whose neoclassi¬ cal ideology occupied brain cells normally dedicated to common sense. Low Wages? Or High Wages? Altman builds his argument methodically, ranging over such issues as labor relations, minimum wages, and child labor. Chapter 2 adopts Leibenstein’s simple but powerful insight that “individuals have some choice over their effort inputs . . . and have different capacities to realize these inputs.” He

FOREWORD

xiii

shows that “relatively low-wage economies should tend to experience rela¬ tively low rates of economic growth.” Higher wages will spur greater x-elficiency and higher effort from workers. He then shows, in chapter 3, that inefficient firms can survive even in highly competitive markets—a fact sur¬ prising to neoclassical economists but self-evident to anyone who has set foot in any real company, whether Fortune 500 or startup. In chapter 4, Altman shows that “redistributing income can increase both x-efficiency and the rate of technical change”—^happier workers simply improve the firm’s perfor¬ mance—another of those basic principles that are axiomatic in organiza¬ tional behavior but somehow exotic in economics. Chapter 5 refutes the neoclassical allergy to minimum wage laws by showing how “minimum wages and unions [can drive] firms into becoming more efficient.” In chapter 6, Altman shows how higher labor costs can explain part of Solow’s total factor productivity “residual.” Firms and institutions respond to higher wages in ways that produce faster economic growth. I have seen first-hand a living example of Altman’s “high road.” In Israel’s kibbutz (collective farm) sys¬ tem, the left-wing kibbutzim for years ideologically opposed solving their labor shortage by hiring outside workers at low wages. Instead they invested in labor-saving technology that in the long run proved far more profitable. The right-wing kibbutzim hired cheap labor without compunction and as a result performed poorly in the long haul. Why has labor productivity across firms and across countries not converged, as it should according to the neo¬ classical “survival of the fittest” axiom? Under Altman’s behavioral theory, high-productivity and low-productivity firms coexist, with identical unit costs; the former simply pay higher wages than the latter. The two types of kibbut¬ zim are a perfect example. Chapter 8 uses David and Arthur’s theory of path dependency to show how “inefficient products and regimes can survive over time if they can re¬ main cost competitive.” Competitive markets thus do not guarantee efficiency. Chapter 9 asks why innovative, productive industrial relations practices are “neglected by most firms.” Conventional wisdom believes that efficient management regimes will always be chosen. Behavioral economics believes they will not if they fail to produce lower unit costs. Empirical data support the latter view. In their re¬ cent book, Jeffrey PfefiFer and Robert I. Sutton (2000) note that some 1,700 business books are published each year, $60 billion is spent on training, and $43 billion is spent on management consultants; yet despite the mountain of knowledge, the bottom-line results are molehills. There is a gap, they note, between knowing what to do and actually doing it. That gap should not exist at all, let alone match the Grand Canyon, under neoclassical economics. In¬ deed, Harvey Leibenstein found the initial inspiration for his x-inefficiency

xiv

FOREWORD

theory in the pile of development consultant reports that were utterly ig¬ nored, though they could have done immense good for countries that chose to adopt their policies. The knowing-doing gap is a major source of x-inefficiency. Chapters 10 and 11 focus on labor practices, showing that better working conditions will lower, rather than raise, unit costs by eliciting higher-quality labor. Contrary to conventional wisdom, societies that take the high moral ground and ban child labor will be even more competitive than those that permit exploitation of young children. “Higher wages,” Altman explains, “compensate, at least in part, for a family’s loss in income that formerly flowed from the employment of children.” In this area, as in many other aspects of labor markets, doing good, under behavioral economics, also means doing well. Chapter 12 shows that Gary Becker’s well-known principle—market forces eliminate discrimination because it fosters inefficiency and higher costs— does not hold under the behavioral model, where low-wage, low-efficiency firms match unit costs of high-wage, high-efficiency ones. A similar argu¬ ment applies, in chapter 13, to pollution—^nonpolluting firms can offset the higher costs by achieving higher x-efficiency. Throughout Altman’s book, xinefficiency plays a key role: socially constructive policies that boost costs (higher wages, working conditions, labor relations) can always be offset, under the behavioral model, by lower x-inefficiency and generally will be when competitive pressures (Leibenstein’s concept of “pressure”) force firms to move closer to their production frontiers. Thus, he notes, “greener econo¬ mies may be competitive with the dirtiest economies in the marketplace, and the greener economies need not fear competition from their more pollution¬ intensive trading partners.” Statman (2000) has shown that between 1997 and 1999, assets invested in “socially responsible” portfolios of U.S. mutual funds almost doubled, to $2.2 trillion, and achieved at least comparable rates of return to “capitalist” portfolios. Chapter 14 focuses on size and economies of scale. Altman shows how monopoly power stemming from size can increase x-inefficiency; antitrust policy should always weigh economies of scale against diseconomies through higher inefficiency. Chapter 15 concludes the book by focusing on the fascinating issue of “whether culture can affect a firm’s productivity and thereby the relative wealth of nations,” a topic addressed by Max Weber. Culture disappears in the utility-maximizing, effort-maximizing conventional model. But in the behavioral model, “culture can affect effort choices.” Together with Shoshana Sharabani (Maital and Sharabani 1997), I have shown a direct link between wealth and culture variables, in a cross-section of countries, using empiri-

FOREWORD

XV

cally measured components of culture developed by Hofstede, HampdenTumer, and Trompenaar. Wise Choices Woody Allen once wrote, “The world is at a turning point. We face the holo¬ caust of nuclear destruction, or the hell of ecological disaster. May we have the wisdom to make the right choice.” Altman says it better, though less amusingly: “What type of market economy evolves, be it high or low wage, cooperative or antagonistic in labor-management relations, green or dirty, efficient or inefficient, depends on the choices made by individuals in society.” We have a choice. There is more than one kind of market economy. The Darwinian determinism of neoclassical economics tried to remove this choice. Behavioral economics has restored it. There are loads of free lunches lying around out there. By recognizing this, our theories hold water. By utilizing them, our pipes hold water, and our societies become better places in which to live. There is a kinder, gentler market economy than the one we now have. Read this book and learn why. Shlomo Maital Academic Director, TIM-Technion Institute of Management and Sondheimer Professor of International Economics Technion-Israel Institute of Technology, Haifa, Israel

.*v

*,jjar '’4AlyA,’4* : ■M

I'M—-|ir* v^flo

*1 ^1 .♦

• >

i■'’*■.■

|*|4

;

,ttiKt«5pi -'.itT

- '■• ■'-^

*•' v’M

T

.

■'•'■ •

r.

»V...

. *

VJy^m

Ni .!i ,*0rw. I* 4*brv»rii7%^'![jN:^^; . ;.t^ -1 • -li 3* . ■/< aii . t, V - Tj v.>-:> / 'J! •-. X' , * '

r

‘ y •>i,

^ '■'•..;

.

< .-'?;

--* -

ii j'j.

^(c

;.’.

uvj^r

-fr«

^ HiiJi>;Ti*' -jpc;

' - ■ F'

_ ^ i,^

!£ '-.HMUteW' * t«i* , ' -

■: ^ K*it ' ■

;)-. i-.-':.. - - ‘4

•.

'’

.rifeT-*

- * •*

'A

f...

• ■• |Z ;... r, 'r# ,

I -.■

■'■. -.?>•I", '

'• -r- - ’ - ? -vii •"'.■^ . - - ^ 4!^* xo>rd I'-*- *1 I

‘

•- ' ‘ ,

‘

■'“">■-.»■ •- ;a V..V-V.;- V ‘

,-

-A -.)^qL,a

■ .,.

-

.

!»i

-Mil* Ji-'vic-■?j^v!;'’>'^»Mr. I

’■. ■ :^ t'< ^

:

- ■^r;t. i’>_» f*0 ) -^|((

•* •< •

. '

■

( .-%

1.1

.in- wVir -.i,^-

y 4. •'

■ ia.* "iwjfeiM*' :

•>

• . ,:»?»ecs«jw - ' •

Acknowledgments

Over the past twelve years or so I have been thinking through and writing about the implications for economic analysis and public policy of revising certain core assumptions of neoclassical microeconomic theory. The raison d’etre for this exercise is to better explain economic events that appear to ei¬ ther contravene the conventional wisdom or be ignored by it. This book has been enriched by the comments, criticisms, and suggestions of many individu¬ als. I would like to thank first and foremost Louise Edmee Lamontagne, who read the entire manuscript and with whom I have had endless and informed discussions on a multitude of issues related to this book and more. Our sixyear-old daughter, Hannah Rachel Altman, has always been a joy who has enriched my life tremendously. Her patience has been invaluable to the completion of this book. Also I owe John Tomer a debt of gratitude for our many conversations on behavioral economics to which he has made im¬ portant contributions. Many thanks to Rick Szostak for his thoughtful re¬ marks and suggestions on many of the chapters published here. The late Harvey Leibenstein contributed in many ways to this book. He encouraged my initial ventures into behavioral economics in spite of our various methodological and substantive disagreements. Moreover, his x-efficiency theory has impacted significantly on many of the basic arguments presented in this book. This book also owes much to the many discussants and re¬ viewers of my papers. Of vital importance have been my presentations and attendance of conferences sponsored by the Society for the Advancement of Behavioral Economics (SABE), the International Association for Research in Economic Psychology (lAREP), and the Association for Social Econom¬ ics (ASA). These organizations play a pivotal role in providing an open and vibrant forum for discussion and debate to those scholars interested in exxvii

xviii

ACKNOWLEDGMENTS

tending the bounds of the conventional wisdom or even providing alterna¬ tive analytical frameworks. This book was completed while I was a Visiting Scholar with the Department of Economics at Cornell University and with the Center for North American Studies at Duke University. I would like to thank both institutions for the facilities and resources made available to me. The University of Saskatchewan Publication Grant covered some of the ex¬ penses incurred in preparing the manuscript for publication. Many thanks to Sean Culhane, my initial M.E. Sharpe editor, for bringing this project to fruition and to my current editor, Elizabeth Granda, for seeing this manu¬ script into print. I would also like to thank the staff of M.E. Sharpe for all their efforts. *

*

*

All of the chapters in this book, with the exception of the introduction, are based on previously published papers that have been reprinted here, with varying degrees of revision, with permission of the publishers. The original references are as follows; M. Altman, “Interfirm, Interregional, and Interna¬ tional Differences in Labor Productivity: Variations in the Levels of‘X-Inefficiency’ as a Function of Differential Labor Costs,” in Studies in Economic Rationality: X-Efficiency Examined and Extolled, ed. K. Weiermair, and M. Perlman, 323—350 (Ajin Arbor: University of Michigan Press, 1990) (chap¬ ter 7); M. Altman, “A Critical Appraisal of Corporate Bigness and the Trans¬ actions Cost Economizing Paradigm,” in Handbook on Behavioral Economics, vol. 2A, ed. Roger Frantz, 217—232 (Greenwich, Conn.: JAI Press, 1990) (chapter 14); M. Altman, “The Economics of Exogenous Increases in Wage Rates in a Behavioral/X-Efficiency Model of the Firm,” Review of Social Economy 50 (1992): 163—192 (London: Routledge) (chapter 5); M. Altman, “Human Agency as a Determinant of Material Welfare,” Journal of Socio-Economics 22 (1993): 199-218 (Amsterdam: Elsevier Science) (chapter 2); M. Altman, “Labor Market Discrimination, Pay Inequality, and Effort Variability: An Alternative to the Neoclassical Model,” Eastern Economic Journal 21 (1995); 157—169 (Eastern Economics Association) (chapter 12); M. Altman, “A Behavioral Model of Endogenous Economic Growth,” in M. Altman, Human Agency and Material Welfare: Revisions in Microeconom¬ ics and Their Implications for Public Policy, 53—68 (Boston: Dordrecht, and London: Kluwer Academic Press, 1996) (chapter 6); M. Altman, “The Meth¬ odology of Economics and the Survival Principle Revisited and Revised: Some Welfare and Public Policy Implications of Modeling the Economic Agent,” Review of Social Economics 57 (1999); 427^29 (London:

ACKNOWLEDGMENTS

xix

Routledge) (chapter 3); M. Altman, “Labor Rights and Labor Power and Welfare Maximization in a Market Economy: Revising the Conventional Wisdom” IntemationalJoumal of Social Economics 27 (2000): 1252—1269 (Bradford, England: MCB University Press) (chapter 10); M. Altman, “A Behavioral Theory of Economic Welfare and Economic Justice: A Smithian Alternative to Pareto Optimality,” International Journal of Social Econom¬ ics 27 (2000): 109S-1131 (Bradford, England: MCB University Press) (chap¬ ter 4); M. Altman, “A Behavioral Model of Path Dependency: The Economics of Profitable Inefficiency and Market Failure,” Journal of Socio-Economics 29 (2000): 127—145 (Amsterdam: Elsevier Science) (chapter 8); M. Altman, “Culture, Human Agency, and Economic Theory: Culture as a Determinant of Material Welfare.” Journal of Socio-Economics 30 (forthcoming) (Amsterdam: Elsevier Science) (chapter 15); M. Altman, “When Green Isn’t Mean: Economic Theory and the Heuristics of the Impact of Environmental Regulations on Competitiveness,” Ecological Economics 36: 31^4 (forth¬ coming) (Amsterdam: Elsevier Science) (chapter 13); M. Altman, “A Revi¬ sionist View of the Economic Implications of Child Labor Regulations,” Forum for Social Economics (Association for Social Economics, forthcom¬ ing) (chapter 11); M. Altman, “Economic Theory, Public Policy and the Chal¬ lenge of Innovative Work Practices,” Economic and Industrial Democracy: An International Journal 22 (forthcoming) (Beverly Hills: Sage Publica¬ tions) (chapter 9).

‘jftCT-.

r

'■

^‘di 'J-Jtjjd'r:*!!'-ifC/-j ^

.0jS4&a3^

VS jTi

'

/ r.VJi^ *ii

. vrH? ; tei ;i)dh^96tidl9 - ,-!>.

j.' A ,>.:•

■»h:'.*,-: a JOSSiirT’l- aJr '■ -_^1. ^

t' i’

I f.* >- • .-Lti

•vl-

, «

-T ■'

«

.

■V.

4

--:#J

WORKER SATISFACTION AND

ECONOMIC PERFORMANCE

Tb-*

^ n-

>iP!OW

» «

1

Introduction Revising the Microfoundations of Economics

Although Worker Satisfaction and Economic Performance is much to do about economic theory, it is also fundamentally a work about public policy and economic welfare. In this book I attempt to lay more realistic micro¬ foundations for analyzing a wide array of public policy questions. This alter¬ native theoretical infrastructure subsumes the mainstream neoclassical worldview along with its related analytical predictions and their implica¬ tions for public policy. The alternative analytical framework is akin to a gen¬ eral theory wherein the neoclassical narrative becomes one possible analytical tale, or a special case, among a variety of alternatives that are articulated in Worker Satisfaction and Economic Performance. I argue that whether or not the mainstream view holds critically depends on the behavioral assumptions that underlie the theory and upon the social and institutional context of the events that the economic theory is designed to explain. If the underlying assumptions of the mainstream theory prove to be the exception to the rule, which is what I argue, then the public policy recommendations that flow from this theory would be faulty, and this would have significant implica¬ tions for public policy and economic welfare. Therefore, unlike the main¬ stream worldview, I maintain that the assumptions underlying economic theory matter. Faulty assumptions yield faulty analytical predictions or at best faulty conclusions relating to cause and effect, with potentially disas¬ trous consequences for society. The general theory presented in this book focuses on one set of alternative and more realistic behavioral assumptions that yield analytical predictions that are more in tune with the stylized facts of economic life and that recom3

4

CHAPTER

1

mend public policy and private choices that often stand in stark contrast with what flows from the neoclassical worldview. In particular, I make the case that what happens within the firm, in particular how workers are treated in the context of the firm’s general human relations or industrial relations envi¬ ronment, plays a fundamentally important role in determining the economic performance of the firm and, thereby, of the economy as a whole. Moreover, I argue, what happens within the firm is subject to the discretion of the deci¬ sion makers inside of the firm. This is true even in a perfectly or extremely competitive product market. Individuals can choose the type of industrial relations system to be adopted by the firm. Market forces do not dictate what system must be adopted. Much ultimately depends on the power relationship among workers, management, and owners, which is in turn affected by mar¬ ket forces and labor market-u-elated institutions, as well as by the prefer¬ ences of the economic agents. There is no unique solution to what happens within the firm. This conclusion—that individuals have a certain degree of freedom in deciding how workers are treated within the firm even in a com¬ petitive environment—^stands in sharp contrast to what is predicted by main¬ stream neoclassical microeconomic theory. Indeed, this contrary perspective also is opposed to the worldview held by many economists of a radical or more institutional bent, for whom market forces or noneconomic factors pre¬ determine the choices made by individuals inside the firm. In the neoclassi¬ cal world workers, managers, and owners ultimately construct the inner workings of the firm so that the firm is maximizing output per unit of input. In the neoclassical firm the best that can be done is accomplished. All oppor¬ tunities for gain are exhausted. Economic agents and the firm are performing efficiently given the various constraints that they face. These constraints, most economists agree, include transaction and information costs.' Technically speaking, the firm is operating along its production possibil¬ ity frontier, where this outermost boundary for maximum obtainable output is determined by traditional inputs such as labor, human capital, capital, land, and technology. Practically speaking, individuals are therefore assumed to be working as hard and as well as they can. This is the manner in which individuals operating within the firm must behave to survive in a reasonably competitive environment. This is the classic analytical position of what has been dubbed the “Chicago School,” whose core is clearly articulated in Milton Friedman’s essay on the methodology of economics, an essay that has found its way into the very heart and soul of contemporary economic theory (chap¬ ter 3 below; Friedman 1953a; Reder 1982). Market forces make economic agents perform invariably in an economically efficient way. In effect, indi¬ viduals are assumed to be working as hard and as well as is possible by dint of market forces. Any other choice spells disaster for the firm. Therefore, no

INTRODUCTION

5

choice is truly afforded to individuals as to how they do behave within the firm when it comes to deciding on the level of efficiency. Economic effi¬ ciency is a product of an economic imperative. Part and parcel of the as¬ sumption that economic efficiency is a given is the notion that individuals always choose the most efficient technology. They must. This again is a prod¬ uct of an economic imperative. Any other choice yields relatively high and uncompetitive production costs. Moreover, implicit in economic theory in general and most especially in its neoclassical variant is the assumption that individuals behave efficiently as a product of a moral imperative, one that is ultimately geared toward maximiz¬ ing the material welfare of the firm (Reder 1982). Deirdre McCloskey has eloquently expressed this fundamental principle of contemporary economic theory as the “American Question” or the “Axiom of Modest Greed”; “The Axiom of Modest Greed involves no close calculation of advantage or large willingness to take risks. The average person sees a quarter and slides over it... he sees a $500 bill and jumps for it. The Axiom is not controversial. All econo¬ mists subscribe to it, whether or not they believe in the market . . . and so should you” (1990, 112). Therefore, even in the absence of serious competi¬ tive pressures, even in a world dominated by monopolies, few economists would expect or predict that the typical individual will pass over reasonable opportu¬ nities for gain. A bit of small change might be overlooked on the sidewalk but certainly not a $500 bill or bills of even larger denominations. This particular axiom also presumes, quite implicitly at that, that all economic agents within the firm are interested in maximizing the material welfare of the firm and that they all stand to gain by working as hard and as well as they can. From this perspective, the existence of a relatively competitive product market is not a necessary condition for the existence of economic efficiency. Nevertheless, it goes without saying that the existence of severe and increasing competitive product markets serves to reinforce the view that economic agents have no choice but to be efficient in the market economy. For those who do not necessarily buy into the argument that economic efficiency is a product of a moral imperative, the notion that it is an eco¬ nomic imperative, whence there is no escape, has been made ever more stri¬ dently in more recent years as globalization has become the catchword for a much more intense competitive environment forcing individuals to behave efficiently or perish in the whirlwind of market forces. Economists, other scholars, and experts from a wide spectrum of political proclivities, ranging from the left to the right, repeatedly make this point. Moreover, in the realm of economic theory, the theory of contestable markets developed by Baumol (1982) suggests that product market competition is even more severe than traditional measures of competitive productive markets imply since what

6

CHAPTER 1

counts is not simply the number of firms in a market and their market share, but also the credible threat of entry of new firms into a particular product market, even in a world dominated by oligopolies and monopolies. Only by protecting an economy from unbridled market forces can the economic im¬ perative be somewhat circumvented, but only at a loss of material welfare to society at large. However, once exposed to market forces, individuals have no choice but to behave in a prescribed manner. Individuals are forced to behave efficiently. It is important to note that according to mainstream theory, efficient be¬ havior is expected to take place by dint of a moral or economic imperative irrespective of the specific institutional framework within which economic agents work unless, of course, institutions are designed to protect the ineffi¬ cient, relatively high-cost economic entities from market forces or to prevent economic agents from realizing their ingrained moral imperative to perform efficiently. Indeed, it is expected that institutions evolve that are compatible with and conducive to economic efficiency, for otherwise the economies con¬ tained within the bounds of efficiency-impeding institutions will fail the test of the market. Therefore, institutional convergence is expected, specifically with regard to institutions that encourage or facilitate economic efficiency. Within these or even less than optimal institutional constraints individuals are assumed to behave in an economically efficient manner.^ Making the case in theory that what happens inside the firm is not a prod¬ uct of individual choice, but is rather a product of a moral and economic imperative, is not a moot esoteric point. It is fundamentally important for public policy. Whether we admit it or not, theory is the lens through which we perceive and make sense of the world around us. Theory helps both de¬ termine those facets of reality to which we pay attention and how we per¬ ceive this reality. If the theoretical lens is not the best of fits, it can provide us with only a distorted picture of how the world really works. This, in turn, affects public policy (chapter 3 below). Therefore, that economic theory pre¬ dicts that economic efficiency is a necessary prior to the process of produc¬ tion, that it is a given, has profound implications for public policy. Under these assumptions, it is argued that living in a global and competitive world economy requires adherence to low-wage paths of economic development combined with minimalist government. The latter would be especially true when government participation in the economic realm refers to institutional or legal support for labor or the labor market writ large. This proposition at least implicitly presumes that a successful market economy, especially one that finds itself in the midst of an increasingly competitive global economy, comes at the expense of either the absolute or relative material well-being of the majority of any given society. The point is made, given the efficiency

INTRODUCTION

7

assumptions of mainstream theory, that efforts to improve the level of mate¬ rial well-being of the working population can only increase production costs, reducing the competitive position of the firm and the economy as a whole. This argument flies in the face of the worldview of Adam Smith that workers should be the immediate beneficiaries of market production and that their material success positively affects the productivity of the economy (chapter 4 below). However, this argument is ironically consistent with Marx’s long disputed contention that the evolution of capitalism goes hand in hand with the immiserization of the working class. Of course, those leaning toward the left then call for efforts to protect society from the unbridled effects of glo¬ balization. Alternatively, many on the left argue that the key to success in an increasingly competitive market economy is to invest appropriately in both human and public capital, so as to allow a nation to take advantage of its comparative advantage, whence will eventually flow material benefits to all by wit of competitive market forces, albeit with some delay (Cohen 1998; Reich 1992). In stark contrast, those leaning to the right make the case that evenmally the benefits of the market economy will trickle down and accrue to all as a product of unfettered market forces. Challenging the view that improving the material well-being of workers (be they skilled or unskilled, blue or white collared, low or high tech) need not damage the economy and may even further contribute to the material well-being of society at large is not simply a matter of presenting a set of facts, however rigorously derived and presented, that appears to demonstrate that the mainstream worldview is wrong. It is a fact that facts per se have convinced few economists or others dealing with economic policy to shift worldviews. Facts must make sense in terms of a particular worldview to win the day. As Thomas Kuhn points out, “Anomalous observations . . . cannot tempt [a scientist] to abandon his theory until another one is sug¬ gested to replace it. ... In scientific practice the real confirmation questions always involve the comparison of two theories with each other and with the world, not the comparison of a single theory with the world” (cited in Coase 1994a, 27). Moreover, it is theory that is our fact-finding machine. It both provides us with the means to search for facts and determines the type of facts we search for. A theory that is wrong fails to ask the right questions. It acts as a blinder, preventing the scholar or the activist from seeing facts that speak to the issues at hand. In addition, facts that appear to contravene theory can be and are dismissed as exceptional or simply a product of poor empiri¬ cal analysis. To quote Kuhn once again. The road from scientific law to scientific measurement can rarely be traveled in the reverse direction. To discover quantitative regularity one must normally know what regularity one is seeking and one’s instruments must be designed accordingly, even then

8

CHAPTER 1

nature may not yield consistent or generalizable results without a struggle” (cited in Coase 1994a, 27). The mainstream point of view will also not be easily overturned by arguments for a more just society. If it cannot be shown that such a society is economically sustainable or what the opportunity costs, if any, for a differently structured market economy are, such calls will fall largely on deaf ears. The construction of an alternative economic theory is therefore critical in making the case that improved working conditions or an ecologically sustainable economy, for example, are possible within the bounds of a competitive market economy. Theory is required to demonstrate the con¬ ditions under which a more “just” society is economically viable or not and the potential costs of realizing such a society. This is the gist of this book. The mainstream worldview does not provide such an analytical framework. I hope to contribute to the construction of Kuhn’s other theory, which is both consistent with the facts and which can serve to confront the analyses and public policy recommendations that flow either explicitly or implicitly from the mainstream analytical framework. Although this is a book that is largely theoretical in nature and is therefore concerned with the logical consistency of the arguments presented, the theo¬ retical arguments presented in this book are not presented largely in math¬ ematical prose. Nevertheless, there is math, as well as analytical diagrams, found throughout this book. Math and analytical diagrams often go a long way toward clarifying the logic of particular argimients. But the objective of the presentation is to expound the theory in a manner that is understandable to a wide audience without losing the necessary rigor required of theory. Moreover, a necessary condition for the formulation of good economic theory is not the extent of the math contained in the articulation of the theory, the capacity of an argument being convertible into mathematical form, or the extent to which logical-miathematical proofs are provided for the “existence” of certain axioms or basic propositions contained in the theory. Some of the most influential works in economics include the contributions of Adam Smith, John Maynard Keynes, Joseph Schumpeter, Milton Friedman, Ronald Coase, and Douglass North, whose works, although logically argued, lack mathematical form and rigor. In addition, I am not concerned with presenting theory without regard to the economic reality to which it must relate if the theory is to be economic theory as opposed to an exercise in mathematical logic or model building for the sake of model building. It is the latter that so much of econom¬ ics has become (Blaug 1998; McCloskey 1996; Nelson 1995; Szostak 1999). This is a book in economic theory grounded and infused by real economic issues and problems and with a steadfast concern for the realism of the as¬ sumptions underlying the economic theory. At this point, it is essential to summarize and further clarify the distin-

INTRODUCTION

9

guishing features of the model articulated in this book as compared to the mainstream theory and to the key alternative economic theories that directly relate to the analytical framework presented below, such as x-efficiency theory and efficiency wage theory. The theoretical framework provided in this book differs from the mainstream neoclassical framework in terms of three key assumptions. For the rest, I remain consistent with the mainstream model. This forces us to focus our attention on those assumptions that are critical both to the mainstream model in terms of the economic problems discussed in this book and to the alternative theory presented here. For example, I as¬ sume that given the constraints that they face, individuals are rational in the sense that they make best use of the information at hand and that they are forward looking in their decision making with the end in mind of maximiz¬ ing their utility or general well-being.^ I also assume perfect product market competition. Deviations from these assumptions can only strengthen the ar¬ guments made in this book. In contrast with mainstream theory, however, I assume that individuals are typically not maximizing the quantity and qual¬ ity of effort inputted into the process of production. Moreover, I assume that labor markets are imperfect in the sense that supply and demand factors will not typically result in identical wage rates for identical types of labor. Fi¬ nally, I make the assumption that individuals within the firm hold different preferences with regard to conditions of work and that it is best to model the firm as composed of at least two groups of individuals or economic agents, such as workers and managers or owners, where these different groups have different preferences. An important footnote to this discussion is to underscore that the analysis presented here is consistent with the major role of institutional factors as deter¬ minants of economic output. It is also important to note that the mainstream view that institutions should converge toward their efficiency-facilitating ideal is rejected here. However, this book is not about institutions per se or about institutional change. The focus here is on what happens inside the firm. I ar¬ gue that the level of efficiency and the rate of technical change are affected by the work environment and by the overall choices made by economic agents within the firm irrespective of the institutional environment in which the firm finds itself and given other constraints, such as transaction and informa¬ tion costs, that it faces. The right rnacro-instimtional environment—such as reasonable guarantees of the related variables of law and order, property rights, and the enforcement of contracts—is no guarantee, and it is not a sufficient condition for the realization of efficiency in production. One set of institutional parameters that is significant to this book and that is typi¬ cally ignored in the literature relates to the labor market and overall condi¬ tions of work. One hypothesis that flows from the analysis presented here

10

CHAPTER 1

is that institutions that strengthen the labor market—that is, improve the overall bargaining power of labor—can serve to improve the overall efficiency of the economy through their impact on the choices made by individuals within the firm. Related to this, the behavioral model articulated in this book serves to provide at least one explanation as to why institutional convergence need not take place. Using Harvey Leibenstein’s term, economic agents are assumed to be typi¬ cally x-inefficient. Only under ideal conditions of work will individuals work their best. This ideal is characterized by a relatively cooperative system of industrial relations.'^ Otherwise, economic agents and the firms that employ them are performing x-inefFiciently. But when individuals are not doing their best, labor productivity is not maximized. In this case, firms and therefore economies would be performing below potential unless conditions of work are ideal. All else remaining the same, production costs would be higher (see chapters below). How can one address the fundamental question asked by good, logically thinking economists: how can such firms survive in the mar¬ ketplace, and why are individuals knowingly forfeiting opportunities for gain, why are they passing over all this cash waiting to be picked up? According to Leibenstein, who pioneered x-efficiency theory, x-inefficient firms can survive in spite of the fact that they are relatively high-cost producers only if they are protected by monopolistic markets, subsidies, tar¬ iffs, and the like. Economic agents are willing to be x-inefficient because maximizing utility involves things other than profits or income, such as lei¬ sure and less stress. So if you can get away with being x-inefficient, why not, if it serves to increase your utility. Moreover, not all individuals in the firm have the same goals and aspirations. Therefore, Leibenstein argues, the con¬ flict in the objectives of economic agents can result in x-inefficient behavior by all, one that even degenerates into a worst-case productivity scenario or a Prisoner’s Dilemma solution to the productivity question. Under these con¬ ditions, it is simply not worthwhile for individuals to chase after all this potential productivity. Of course, Leibenstein considers x-inefficient behav¬ ior to be only quasi-rational, making his work the bane of so many in the economics profession. Needless to say, so long as utility maximization’s con¬ siderations about one’s nonmaterial wants and desires are considered ratio¬ nal, there is no reason to define x-inefficiency as a product of irrational or even quasi-rational behavior. X-inefficiency is not possible when product markets become reasonably competitive since, in this case, x-inefficient firms cannot survive in the marketplace. So long as there are relatively x-efficient firms around, they will be the lower-cost producers, and they will success¬ fully drive out of the market or force into x-efficiency the formerly laggard firms. Most economists are of the view that at least over time markets are

INTRODUCTION

11

competitive enough to impose a certain discipline upon firms. But in this case, x-inefficiency in production would be an impossibility. I argue that even under perfectly competitive product markets rational eco¬ nomic agents can be expected to act x-inefficiently unless ideal conditions of work prevail. How is this possible? Should not the x-inefficient firms be driven by the wayside? This is in no way the case once it is recognized that increasing the level of x-elficiency is not a free ride. It comes at a material or economic cost to the firm. Improved conditions of work typically involve increasing workers’ income through wages, bonuses, or profit sharing and improved work¬ ing conditions, such as safety and health standards. In addition, often firms need to invest in plant reorganization and management reorganization, where the latter involves a relatively smaller layer of management and more worker participation at different levels of the decision-making process within the firm. At the end of the day, the increased productivity accming to the x-efficient firm might very well be simply offset by the increased production cost of achiev¬ ing more worker satisfaction. In other words, there might be no cost advantage to the firm in becoming x-efficient, nor need there be any material advantage accming to members of the firm hierarchy. Moreover, they experience a rela¬ tive reduction of power and prestige in the x-efficient firm, where the hierar¬ chical power stmcture becomes relatively flatter. In contrast, in remaining relatively x-inefficient, although the firm is relatively unproductive, it need not be uncompetitive to the extent that the lower productivity is compensated for by lower labor costs. Indeed, what is argued throughout this book is that the most appropriate approach to modeling the firm in terms of addressing a wide range of public policy questions is to assume a world where a unique unit cost of production is associated with a wide array of rates of labor compensation and overall working conditions. Unit costs remain constant as changes in labor productivity coinciding with changes in the level of x-efficiency offset any increases in labor costs. When x-efficiency is realized—that is, given technol¬ ogy, the firm is maximizing output per unit of labor input—^further increases in labor costs yield increases in unit production costs, making the firm relatively uncompetitive. By way of contrast, in the mainstream model, given technol¬ ogy and where the quantity and quality of effort is not a variable and the firm is by definition x-efflcient, the firm has no room to maneuver when faced with increased labor costs, no matter their point of origin. Unit costs must increase. The static x-efficiency model, where technology is given, is extended in this book to incorporate technological change that is induced by changes in costs of production and where the choice of technology can be x-inefficient (chapters 6 and 8). In a nutshell, it is argued that technological change is in part a product of efforts within the firm to remain cost competitive in the face of rising production costs, especially when reductions in the level of

12

CHAPTER 1

x-inefflciency are not a viable option. In this context, to the extent that the available or new technology is capable of just compensating for higher pro¬ duction costs by increasing labor productivity or total factor productivity, only those firms that would otherwise face higher unit costs of production will adopt the higher productivity technologies. More specifically, low-wage firms, for example, need not adopt the more productive technologies if they can remain competitive on the bases of low wages and poor working condi¬ tions. On the other hand, the relatively higher-wage, x-efficient firms will tend to adopt such technologies. In this scenario, firms are afforded an addi¬ tional degree of freedom when confronted by potentially higher production costs than is possible in the static x-efficiency model and, of course, in the traditional modeling of technological change. In the model developed in this book, where x-efficiency is not assumed a priori and where the benefits of becoming x-efficient fall in the hands of workers and x-efficiency need not yield a competitive advantage to the firm, rational decision makers within the firm can be expected to decide against building a relatively more x-efficient firm. In so doing, they would not be forfeiting material opportunities for gain for themselves, but only for their employees. Moreover, remaining x-inefficient or adopting a system of labor relations and labor compensation that yields x-inefficiency in production, need not result in any penalty to the decision makers or to the firm. Indeed, it might yield to them relatively more power, prestige, and even income. Re¬ maining x-inefficient costs workers and society at large in terms of lost in¬ come, and of course this has significant welfare implications. But the x-inefficient solution to the productivity problem might very well be the equi¬ librium one, given the preferences of the decision makers. There is no eco¬ nomic imperative embodied in the analytical framework presented in this book that yields either an x-efficient or x-inefficient and correlated high- or low-wage solution to the productivity problem. This is unlike what is found in the conventional wisdom, where x-efficiency is assumed from the get-go, and more often than not this is correlated with a low-wage regime. On the other hand, the results generated by my model also speak against the recent research in the cooperative game theoretic literature that argues for an economic imperative toward a high-wage cum x-efficient solution to the productivity problem (Fehr and Schmidt 1999; Fehr and Gachter 2000). According to this analytical framework, an x-inefficient course of firm de¬ velopment will be chosen unless the decision makers, in particular the own¬ ers and other members of the upper echelons of the firm hierarchy, have a preference for a system of labor relations that is more egalitarian. This of course assumes that workers and their representatives hold a similar set of preferences. Such preferences appear to be few and far between among mem-

INTRODUCTION

13

bers of the firm hierarchy. It is therefore fundamentally important to the theory presented below that one recognize the extent to which individuals within the firm are not homogeneous in preferences. Of critical importance are the differences between those who make the decisions for the firm and those who are affected by these decisions. In the analytical framework presented here the decision makers within the firm, to the extent that they can exercise their preferences, can significantly affect the state of labor conditions within the firm. On the other hand, if a more egalitarian work structure is forced upon the firm or if market forces, labor organization, or legislation require improved working conditions and rates of labor compensation, the behav¬ ioral model of the firm presented in this book suggests that improvements in the level of x-efficiency within the firm can compensate for the increased labor costs, maintaining the competitiveness of the affected firm. Of course, if the organizational strucmre of the firm, combined with organizational in¬ ertia and lack of trust among economic agents, prevents the firm from reduc¬ ing the level of x-inefficiency or from introducing new technology to sufficiently compensate the firm for factors that would otherwise cause in¬ creases in unit production costs, then the firm could fall by the wayside in the face of market forces and organizational failure within the firm. A critical point that stems from the theory enunciated in this book is that individuals can make choices that determine the level of material well-being in society and that such choices are not completely constrained even under conditions of perfect product market competition. Moreover, these choices need not be based on irrational preferences on the part of the decision makers. The possibility that preferences and the capacities to realize a given set of preferences can affect the level of x-efficiency in a competitive market economy not only stands in contrast to the conventional wisdom and to traditional xefficiency modeling of the firm, but it is also at variance with the analysis that stems from the efficiency wage literature. Of course, effort is a variable in the model presented here, as it is in efficiency wage theory, which was also pio¬ neered by Leibenstein (1957,1974). However, in the standard efficiency wage literature the wage rate is, in effect, the efficiency wage and is therefore not a choice variable. Rather, the efficiency wage is the unique wage rate, given technology and the human capital endowment of the economic agents, that yields minimum unit production costs by maximizing effort, where the latter is subject to diminishing returns with respect to productivity. The efficiency wage is therefore technologically and mechanically determined, and the system of industrial or human relations within the firm does not affect the efficiency wage. A wage higher than the efficiency wage will yield higher unit costs, as will a wage rate that is lower. In this model, rational decision makers must choose the efficiency wage if their firms are to remain competitive. In a com-

14

CHAPTER 1

petitive world, the only wage rate that can be chosen is the efficiency wage, and therefore all wages must be efficiency wages. Under these circumstances, ceteris paribus, wages are inflexible both upward and downward. Moreover, one would not expect wages to differ among like workers and firms since they all must be paid the same efficiency wage if firms are to survive in the competi¬ tive market (Akerlof 1984; Akerlof and Yellen 1986; Shapiro and Stiglitz 1984; Stiglitz 1976, 1987). In fact, wage rates can differ in these circumstances only if workers have different preferences for working harder so that work unit costs are minimized at different wage rates (Albrecht and Vroman 1998). In this case, workers are effectively choosing their wage by choosing their prefer¬ ences for effort inputs. Unlike efficiency wage theory, the behavioral model presented here does not assume that there is a unique efficiency wage that yields a unique mini¬ mum unit production cost. The uniqueness given by efficiency wage theory is given by the assumption of diminishing returns to effort inputs. In the behavioral model, I argue that it is best to model the firm through the as¬ sumption that there is a one-to-one proportional, or linear, relationship be¬ tween labor costs in general (inclusive of wages) and effort inputs. This type of relationship would hold up to a point, after which diminishing returns to effort inputs kick in. Such a potential modeling of the economic agent is clearly admitted by Stiglitz (1987) and Akerlof and Yellen (1990), among the leading contemporary efficiency wage theorists. Given the proportional¬ ity assumption of the behavioral model, there need not be a unique wage, but rather there is a unique unit cost of production associated with an array of working conditions, inclusive of the wage. In this model, the wage rate, for example, is not technologically determined, nor is it simply or necessarily a function of workers’ preferences. Much is determined by the preferences of the decision makers. These can be affected by many variables, including culture, the law, the state of the labor market, and the relative power of labor organizations. Even workers who want to work hard and well will not do so if the decision makers have the capacity to opt for a low-wage regime. On the other hand, a high-wage regime can be realized with the same type of workers when decision makers opt for such a regime. Therefore, in the be¬ havioral model, decision makers are not forced by circumstances to adopt a unique wage or unique set of working conditions. Decision makers and the economic agents within the firm in general are afforded a greater degree of freedom in determining wage rates and working conditions than in the effi¬ ciency wage model. Unlike in efficiency wage theory, in the behavioral model one can have, in competitive long-run equilibrium, similar workers using the same technology being paid different wages in different firms. Moreover, wage rates need not

INTRODUCTION

15

be inflexible downward or upward. Owners and managers might try to pres¬ sure for low-wage regimes when this will not increase unit costs. On the other hand, efforts to develop high-wage regimes with high-quality working conditions can prove competitively viable if these efforts do not negatively affect unit production costs. The behavioral model is also consistent with firm decision makers showing reluctance to cut wages during an economic downturn if this involves no improvements in costs and if this, moreover, involves costly short-run adjustments in the firm.^ Therefore, I would argue that this relatively more flexible model is better able to explain the stylized facts of economic life in a competitive environment, even when one assumes rational economic agents, than are either the efficiency wage or x-efficiency wage models. The fact that individuals can exercise their preferences in the behavioral model presented here does not imply that market forces play no role in the determination of wages. Market forces affect the mean wage in the different labor markets of a particular economy. Labor shortages can be expected to yield higher wages, whereas surpluses provide firms with the opportunity to reduce wages. Rather, what is implied in the analytical framework articu¬ lated in this book is that given the existence of x-inefficiency, a whole array of wage rates is consistent with some unique and competitive cost of produc¬ tion. Firms can competitively pay workers above what is required by market forces per se given that such relatively high rates of labor compensation are covered by corresponding increases in the level of x-efficiency. In this case, there can exist, in equilibrium, a wide array of wage rates for any given state of labor supply and demand. But relatively high wage rates, above the marketdetermined mean, can create a surplus labor supply in the high-wage labor markets. In this scenario, however, the low-wage workers cannot easily com¬ pete down the wages or working conditions in the high-wage firms since lowering the wage rate need not provide the firm with any cost advantage if productivity moves with wages. Indeed, if there are any team effects in the firm accruing over time, replacing high-wage with low-wage workers might reduce firm productivity to such an extent that unit costs will increase until the team component of productivity is restored. Nevertheless, the firm’s de¬ cision makers can adopt the low-wage option afforded to them if that is their preference and if this preference can be realized. This would be especially true in a nonunion environment. On the other hand, shifts upward in the economy wide mean wage by market forces or government policy need not result in higher unit production costs, as would be true in the conventional model, when x-inefficiency exists and such wage increases motivate produc¬ tivity increases. The behavioral model presented here opens the door to preferences, insti-

16

CHAPTER 1

tutions, and public policy as critical determinants of cost-competitive wage rates. As is argued in Galbraith (1998), the extent of unionization, legislation on minimum wages, unemployment insurance, welfare, and the like, as well as unemployment and related macroeconomic policy, can affect the wage rate in the economy as a whole, as well as differentials in pay across labor markets within a particular economy. In the behavioral model of the firm such exogenous influences on labor market outcomes are consistent with a firm’s ongoing competitiveness even in a highly competitive product mar¬ ket. Moreover, in the behavioral model, rational preferences, however formed, can affect the wage regime in the firm and in the overall economy. Certainly, unlike in the conventional model, where the quantity and quality of effort is not a variable and where firms are assumed to be performing x-efficiently, especially in competitive markets, preferences for a high-wage regime can be realized without harming a firm’s competitive position. This would be true even in a world dominated by firms where low-wage regimes mle the roost. Throughout this book the behavioral model is elaborated upon as it is applied to a variety of social, public policy, and theoretical issues. The driv¬ ing force behind this book is a concern for the capacity of economic theory to explain economic reality and guide empirical research, as well as the role played by theory’s underlying assumptions in achieving these critically im¬ portant tasks. This book begins (chapter 2) with the presentation of a basic theoretical framework wherein individuals have some choice over their effort inputs, different preferences with regard to effort inputs, and different capacities to realize these inputs. Moreover, in this model, what happens inside of the firm matters for firm productivity. Of vital importance is the relationship between workers and employers. Under these assumptions, preferences, cul¬ ture, firm organization, and the institutional constraints faced by the firm affect the level of material well-being in society through their effect on the level of x-efficiency and the rate of technological change. One of the central analytical predictions of this chapter is that relatively low-wage economies should tend to experience relatively low rates of economic growth and rela¬ tively low levels of material well-being. It is here postulated that labor costs represent a significant spur to reductions in the level of x-inefficiency and to increases in the rate of technical change. Some of the basic implications of my economic methodology are explored in chapter 3. I focus on Milton Friedman’s survival principle, which has dominated economic discourse either explicitly or implicitly for generations. Friedman marginalizes the importance of behavioral assumptions for the construction of economic theory and focuses on the predictive powers of the theory, irrespective of the theory’s underlying assumptions. In his type of

INTRODUCTION

17

methodology, although facts matter, they are always under deep suspicion when in conflict with the predictions of the theory. However, critical to Friedman’s methodological approach is the assumption that individuals per¬ form efficiently (x-efficiently) inside the firm. This, he argues, must be so, for if human behavior were otherwise, firms could not survive in a competi¬ tive market. Of course, inefficient firms could survive when protected from market forces, but at a cost of material welfare to society at large. However, the methodological legacy of Friedman’s contribution is the presumption that if firms survive, they must be efficient. I argue that once one admits to the possibility of effort discretion, this version of the survival principle is inappropriate and misleading. In the behavioral model presented in this book, inefficient firms can survive even in highly competitive markets, so they pass the critical test of survival. But survival of the firm can no longer be proof of efficiency and what happens inside of the firm, and the assumptions one makes about the behavior of economic agents take on critical importance in understanding the dynamics of a competitive market economy. In this sce¬ nario, Friedman’s narrative holds true only under very special and unrealis¬ tic assumptions. Still, it remains important to appreciate that firms must remain cost competitive to survive in the marketplace. However, in accordance with the behavioral model, becoming efficient is not the only means of achieving this end. This opens the door to questions relating to the survivability of xinefficient firms and the dynamic process involved in building more x-efficient firms and economies. In chapter 4, the methodological discourse is carried one step further. The traditional economic approach to so many public policy issues remains im¬ mersed in the notion of Pareto Optimality, which presumes that society’s well-being is somehow maximized when it is not possible to make one indi¬ vidual better off without reducing the well-being of at least one other individual. Well-being is typically taken to mean economic or material well-being that is tied to society’s measurable output. Underlying the contemporary rendition of Pareto Optimality is the assumption that individuals are efficient from the start and that the level of efficiency and the rate of technical change are independent of events that redistribute income from one individual to an¬ other. Moreover, it is presumed in the conventional wisdom that a dollar yields the same level of welfare to all economic groups in society so that, for example, a dollar transferred from the rich to the poor has the same welfare effect as a dollar transferred from the poor to the rich. Pareto Optimality has been central to economists avoiding public policy analyses that speak to the distribution of income and its redistribution. The behavioral model suggests that redistributing income can have a positive impact on both the level of xefficiency and the rate of technical change. Moreover, redistribution of in-

18

CHAPTER 1

come need not result in any one individual being worse off at the end of the day. The argument is simple. If, for example, working conditions are improved in the firm and the costs of so doing are recuperated through improvement in the firm’s efficiency and the adoption or development of more cost-competi¬ tive technology, income is in effect redistributed to labor but not at the ex¬ pense of managers or owners. In the behavioral model there is a dynamic relationship between income redistribution and the level of economic wel¬ fare and therefore there need not necessarily be a trade-off between equity and econorhic efficiency. Moreover, equity can contribute to efficiency and economic growth. The behavioral model opens the door to public policy questions that speak directly to the redistribution of income as one potential tool for promoting efficiency and growth as opposed to benefiting one group at the expense of another. In chapter 5, the behavioral model is applied to the issue of minimum wages and unions. The standard economic model predicts that legislative and union interference in the marketplace can only damage the economy, especially its poorest, least skilled members. Empirical studies that suggest that this is not so are typically dismissed as being necessarily flawed, as they are out of whack with the analytical predictions of mainstream theory. How¬ ever, critical to mainstream theory is the assumption that all firms are neces¬ sarily efficient, for otherwise how could they survive in the marketplace? In the behavioral model, of course, no such assumption is made a priori. There¬ fore, contrary to the conventional wisdom, it is quite possible for institutions such as minimum wages and unions to have positive effects on the economy by driving firms into becoming more efficient. On the other hand, the ab¬ sence of such institutions can add to persistent inefficiencies in the economy. Empirical studies suggesting that minimum wages and unions are not harm¬ ful to the economy should therefore not be so easily dismissed. Indeed, in the behavioral framework they might serve as proof of a positive and dy¬ namic relationship between labor costs and economic efficiency. The behavioral model has significant implications for growth theory (chap¬ ter 6). The traditional growth model developed by Robert Solow serves as the backbone to the behavioral model developed here. The essence of the behavioral model is to offer an explanation for what Solow refers to as the “residual,” which in his model remains unexplained. Most scholars agree that the majority of increases in per capita growth are not a product of in¬ creases in capital and labor inputs. Rather, they are a function of technical change and other factors that generate increases in per capita output. Most recently, some economists have argued for an endogenous theory of eco¬ nomic growth whereby per capita growth is driven by investments in knowl¬ edge, such as research and development, suggesting that such investments

INTRODUCTION

19

generate per capita growth that have a ripple effect on growth that is sus¬ tained over time. Most economists remain highly dubious that technical change, no matter how constructed, can have an effect on growth, which is much more than a one-shot deal. The behavioral model retains this basic assumption of the conventional growth literature but offers a causal explana¬ tion of sustained growth that is wanting in the conventional worldview. Sim¬ ply put, in the behavioral model, growth is a function of increasing costs of labor, which provoke improvements in the levels of x-efficiency and the rate of technical change. Even in an economy where there are significant invest¬ ments in research and development and education, if the economy is charac¬ terized by low wages and poor working conditions, the incentives for firms to become more efficient and more technologically advanced are lacking. In the behavioral model, per capita growth is induced or hindered by labor mar¬ ket conditions that are filtered through the modus operand! of the firm. This model suggests that economies where labor market pressure is brought to bear and where firms and institutions respond to these pressures appropri¬ ately are the economies most likely to grow at the fastest pace and to achieve the highest level of real income per capita. In chapter 7, the behavioral model is used to address the persistent lack of convergence in labor productivity and thereby per capita output across the economies of the world. This failure of convergence stands starkly against what the conventional wisdom predicts should occur—that is, market forces would pressure the less productive firms into becoming more productive for reasons of survival. The relatively less productive firms should be character¬ ized by relatively higher unit costs, making them uncompetitive. In the be¬ havioral model, convergence is not inevitable since the more productive firms need not be any more competitive than their less productive counterparts when productivity is positively and causally related to labor costs. In such a world, relatively low- and high-productivity firms can produce at the same unit cost and therefore be cost competitive. In this scenario, market forces per se cannot force convergence in productivity. However, in the behavioral model convergence in labor costs, through its effect on the level of x-eflficiency, can contribute to convergence in labor productivity and therefore per capita output. An important and controversial theory, one that is marginalized by the conventional wisdom, is the theory of path dependency, pioneered indepen¬ dently by Paul David and Brian Arthur. In essence this theory argues that inefficient products and economic systems can persist over time if they were chosen by happenchance. Future economic inefficiencies become path de¬ pendent or products of historical choices, and these inefficiencies become locked into an economy’s future development even with known and more

20

CHAPTER 1

efficient products and regimes. The conventional wisdom rejects this theory, which presumes that economic inefficiency can persist over time in the face of competitive market forces. But with the behavioral framework (chapter 8), it is shown that economically inefficient products and regimes can sur¬ vive over time if they can remain cost competitive. Moreover, both efficient and inefficient products and regimes can exist simultaneously over time. In a world where effort is variable, the existence and persistence of both efficient and inefficient products and regimes can be path dependent. Therefore, a competitive market is no guarantee of the dominance of efficient products and regimes. Of course, this was the central message of the original propo¬ nents of path dependency theory. The persistence of economic inefficiencies is again raised in chapter 9, where the paradox of the dominance of traditional, apparently less efficient work practices over more innovative, apparently more efficient work practices is examined. There is a large empirical literature suggesting that relatively more productive systems of industrial relations or firm management are known to firm decision makers but remain neglected by most firms. These systems are typically more cooperative and egalitarian in nature. The conventional wis¬ dom rejects the notion that efficiency can be determined by firm organization per se, but to the extent that it can, it is assumed that the management regimes that are chosen will be the efficient ones. Therefore, what appear to be more efficient work practices are, in effect, not that at all. The behavioral model suggests, however, that the innovative work regimes might very well be the relatively more x-efficient ones, but all or even most firms need not adopt them if they do not generate lower unit costs than their less x-efficient counterparts. And this might very well be the case since the more innovative regimes must be developed at a cost. As long as the decision makers do not see benefit to themselves in constructing a more irmovative, efficient, but more costly sys¬ tem of industrial relations, it will not be done. According to the behavioral model, the more efficient work regimes will be adopted if there is labor market or institutional pressure to do so or if the preferences of the decision makers become biased toward the innovative systems. In chapter 10, the question of work practices is examined from the per¬ spective of the potential impact of improved working conditions and en¬ hanced power of labor on economic welfare and upon the competitive position of the firm. This takes us back to some of the issues raised specifically in chapter 5 and more generally to one of the central themes of this book, which addresses the role that worker satisfaction plays in determining the efficiency of the firm. The conventional wisdom contends that improvements in labor standards, be they imposed institutionally or through the enhanced bargain¬ ing power of workers, will increase production costs, thereby damaging the

INTRODUCTION

21

long-run viability of the firms and economies that are so affected. In con¬ trast, the behavioral model suggests that improvements in labor standards might have no such negative impact on the economy given the positive ef¬ fects that such investments can be expected to have on the firm’s economic efficiency. Therefore, economies with high labor standards need not fear competition from economies with substandard working conditions when superior working conditions contribute to improve firm productivity and when inferior working conditions tend to have the opposite effect. The arguments presented in chapter 10 are narrowed down in chapter 11 to examine the economics of child labor. Most economists, no matter their political perspective, accept the view that using child labor provides econo¬ mies with a competitive edge given the poor working conditions and low wages that characterize the employment of children. Restrictions to child labor, therefore, will tend to damage the economies that employ such labor, while economies that place restrictions on child-labor for moral reasons will find themselves at a competitive disadvantage with child-labor-intensive economies. Moreover, it is also argued that restricting child labor tends to damage the already fragile level of material welfare that characterizes fami¬ lies whose children are employed. In the behavioral model it is argued that economies employing child labor need not have any competitive advantage. Child labor is typically less productive than adult labor. This productivity differential can serve to neutralize any cost advantage that a child-laborintensive economy might otherwise have. Moreover, a diminution in child labor would have the effect of reducing the overall supply of labor, thereby placing upward pressure on the adult wage rates. The conventional wisdom predicts that will invariably damage the economy. But the behavioral model suggests that the induced higher adult wages can serve to increase an economy’s level of x-efficiency and the rate of induced technical change. Moreover, the increased wages can compensate, at least in part, for a family s loss in income that formerly flowed from the employment of children. Fi¬ nally, I argue that in a world where adults are income maximizing and real target income increases over time, one cannot expect parents to reduce child labor without some institutional intervention in the decision-making pro¬ cess. Still, the important economic and social question remains what to do when a reduction or elimination of child labor has at least the short-run ef¬ fect of reducing a family’s already low level of real income. The sustainability of pay discrimination in a competitive market economy is the subject of chapter 12. The long-standing argument in economic theory, following upon the work of Gary Becker, is that discriminatory pay inequality cannot be sustained over time since the firms doing the discriminating—against women, for example—^will be the high-cost firms, and they will be driven out

22

CHAPTER 1

of business by their relatively low-cost, nondiscriminating competitors. I ar¬ gue that in a world where wages and working conditions affect productivity, nondiscriminating firms need not have any competitive advantage over the discriminating firms by hiring the relatively low-wage women. To the extent that low levels of efficiency match the low wages, both the discriminating and nondiscriminating firms will be cost competitive. In this case, market forces in themselves cannot be expected to rid society of pay inequality due to discrimi¬ nation, as the conventional wisdom would have us believe. In chaptfer 13, the behavioral model is applied to environmental policy, addressing the issue of whether or not efforts to reduce pollution can be expected to increase production costs and reduce the regulating economy’s level of material welfare. The conventional wisdom argues that although less pollution is no doubt a good thing, it must come at a cost, and individuals must decide if the benefits outweigh the costs. But the argument that pollu¬ tion abatement necessarily increases production costs assumes that firms are always x-efficient and that pollution abatement policy cannot affect the rate of technical change. If these two assumptions do not hold, pollution abate¬ ment need not generate higher unit costs. In the behavioral model the cost involved in pollution abatement can serve to pressure firms into becoming more x-efficient and developing and adopting more productive but greener technology. These productivity-enhancing behaviors represent cost offsets to the cost involved in making firms greener. In this scenario, greener firms need not be the higher-cost firms, as is predicted by the conventional wis¬ dom. Indeed, greener economies may be eompetitive with the dirtiest econo¬ mies in the marketplaee, and they need not fear competition from their more pollution-intensive trading partners. However, there is no reason to expect, on theoretical grounds, that firms will be greener on the basis of the free will of their decision makers when pollution abatement provides no cost advan¬ tage to the greener firms. Ajiother issue addressed in this book (chapter 14) is whether or not economie efficiency and corporate bigness go hand in hand. The standard ap¬ proach in contemporary economics is to argue that bigger is just about always better. Mergers take place so as to exploit various economies, including the reduction in transaction costs, that corporate bigness affords firms. There¬ fore, bigger firms mean more efficient and lower-cost firms than would oth¬ erwise be possible. Efforts to prevent mergers, therefore, can be expected to typically be harmful to the economy. However, these expectations stem from modeling that presumes that firms are always x-efficient irrespective of the market stmcture in which they find themselves. In the behavioral model, it is argued that one variable omitted from the conventional analysis is the possi¬ bility that any market power garnered by the larger corporations will, through

INTRODUCTICn

23

the protection this affords to such firms, result in increased levels of x-inefficiency. Monopolistic power yields x-inefficiency. To determine whether or not corporations are getting too big, according to this analysis, one has to incorporate the possibility of x-inefficiency and weigh it, among other costs, against any possible gains corporate bigness can be expected to generate. The final section of this book (chapter 15) brings us back in some sense to chapter 2, as it focuses on the role of culture in economic development and thereby at a more general level on the role of choice or human agency as a determinant of material welfare or economic well-being. Culture has been long neglected by economists as a possible cause of differentials of income among individuals and among economies, although most recently cultural factors have been shot into the limelight by Francis Fukuyama (1995), David Landes (1998), and Thomas Sowell (1994). In previous chapters, I have emphasized the role that work culture can play, directly and indirectly, in affecting the level of material welfare. In this chapter the focus is on the more traditional definition of culture in terms of norms and social context and their effect on the choices made by economic agents in the workplace.^ In the tradition of Max Weber, this chapter addresses the question of whether culture can affect a firm’s productivity and thereby the relative wealth of nations. The conventional wisdom, assuming effort to be always maximized, assumes away the possibility of culture affecting productivity. Moreover, it is argued that individuals choosing for cultural reasons to work x-inefficiently would be eliminated by market forces. In other words, the conventional wis¬ dom assumes that all individuals are constrained by the same cultural pre¬ cepts, at least as they relate to effort choices. In the behavioral model, culture can affect effort choices. Also, under reasonable assumptions individuals choosing to work at different levels of intensity and quality can all survive even in a competitive marketplace. To the extent that firms are culture-bound to be x-inefficient, they can survive if they can keep labor and other costs low enough to compensate for their relatively low levels of productivity. Of course, individuals who are culture-bound to behave x-inefficiently need be paid a wage low enough to compensate for their low level of productivity. The alternative is for the cultures of x-inefficient societies to be transformed so as to facilitate increasing the level of x-efficiency to the extent desired by individuals in these societies. Culture itself can be affected and changed by the firm’s organization, education, and the political environment, where these factors have a dialectical relationship with their culhiral milieu. The point is that culture matters and that different cultural settings and choices carry with them certain material costs and benefits. The fundamental focal point of the theoretical framework presented and applied in Worker Satisfaction and Economic Performance is the revision of

24

CHAPTER 1

one basic premise of mainstream theory with important implications for the direction of theoretical and empirical analysis, as well as for public policy. The conventional wisdom, by assuming that firms are efficient in terms of effort inputs, as well as in the use of technology, assumes away significant potential explanations for economic problems, dilemmas, and paradoxes. In particular, it assumes away problems that derive from how the firm is organized for pro¬ duction and the social context in which the firm operates. The behavioral model presented here, by allowing for effort discretion, as well as for different prefer¬ ences across individuals, opens the door to a better understanding of a series of important economic questions that cannot be effectively addressed by the con¬ ventional wisdom. The mainstream model holds sway only when efficiency is ever present and a firm’s achievement is independent of its organization, mar¬ ket structure, factor prices, and the institutional framework of society. The latter, I argue, holds only as a clear exception to the rule. Using the conven¬ tional model or allowing its reasoning to dominate one’s thinking implicitly or explicitly when one deals with issues that relate to efficiency can create highly distorting and misleading pictures of economies, resulting in public policy that can only pretend to cope scientifically and objectively with economic and so¬ cial problems. The behavioral model presented in this book provides no defini¬ tive or exact answer to any particular problem. However, this model suggests that market economies can be varied and multifaceted. One cannot, therefore, predict a priori which type of economy will survive or dominate over time. Many types of capitalist economies can exist simultaneously, even in a highly eompetitive world. What type of market eeonomy evolves—be it high or low wage, cooperative or antagonistic in labor-mianagement relations, green or dirty, efficient or inefficient—depends on the choices made by individuals. There is in the behavioral model no eeonomic imperative that forces individu¬ als to necessarily choose one route over another. Preferences and the capacity to realize these are, for this reason, of fundamental importance to understand¬ ing how an economy evolves. The behavioral model hopefully opens the door to such an understanding. Notes 1. Mainstream economic theory has, for the most part, incorporated into its ana¬ lytical core the reality that transaction and information costs are positive and, more¬ over, that individuals are imperfect processors of information. However, this has in no way altered the conventional wisdom’s core assumption that economic agents per¬ form efficiently. Only now productivity and output are less than they would be in a world with no transaction and information costs and with individuals with the compu¬ tational capacity of computers. See Williamson (1985, 1986) and De Alessi (1983) on transaction costs and Simon (1959,1978,1987) on information and processing costs.

INTRODUCTION

25

2. A growing body of economists now argues that institutional convergence need not occur, largely for reasons of transaction costs, following upon the work of North (1990). Still, individuals are assumed to behave efficiently given the constraints that they face. However, suboptimal institutions can make efficient economic agents less productive than they might otherwise be. North concludes from his analysis of the development of economic institutions through time that these institutions created en¬ vironments that “overwhelmingly favor activities which promote redistributive rather than productive activity, that create monopolies rather than competitive conditions, and that restrict opportunities rather than expand them. . . . The organizations that develop in this institutional framework will become more efficient—but more effi¬ cient at making the society even more unproductive and the basic institutional struc¬ ture even less conducive to productive activity” (1990, 9). For further discussion of the role of institutions in the macroeconomic growth process see, for example, Lamoreaux, Raff, and Temin, eds. (1999) and Olson and Kahkonen, eds. (2000). 3. It is important to note, however, that a growing body of economics literature suggests that a significant amount of individual behavior is inconsistent with the neo¬ classical modeling of rational behavior (see, for example, Maital 1982; Maital and Maital 1993; Shiller 1993; Thaler 1992a). 4. The conclusion that a relatively cooperative system of industrial relations will yield economic efficiency not otherwise achievable has also been reached more re¬ cently by Miller: “Cooperation will be defined as occurring when individuals in a social dilemma select alternatives that are not rewarded by the formal incentive sys¬ tem but that result in Pareto-efficient outcomes. Cooperation will offer efficiency gains that short-term hierarchical incentives cannot promise. In the limit, indeed, coopera¬ tion will reestablish ideal efficiency as a benchmark for organizations that can never achieve it otherwise” (1992, 177). A similar point is also made in Fehr and Schmidt (1999), Fehr and Gachter (2000), and Gordon (1996, 1998). 5. The most recent contributions to efficiency wage theory have been directed toward explaining the downward inflexibility of money wages during economic down¬ turns (Akerlof 1984; Akerlof and Yellen 1986; Bewley 1999; Shapiro and Stiglitz 1984;Yellen 1984). 6. On a more microeconomic level, Lamontagne (2001) shows how cultural fac¬ tors, often mediated through community organizations, can help explain how one group of individuals can become economically successful while another group, fac¬ ing similar economic constraints, might fail. See Yang and Lester (2000) for a discus¬ sion of the relationship between unemployment and cultural factors.

2

Human Agency as a Determinant of Material Weifare

Introduction The conscious actions of purposeful individuals can help determine the level and rate of growth of gross national product (GNP), yet standard economic theory pays little or no attention to human agency as a primary determinant of society’s material welfare. In any firm where goods and services are produced, economic agents, workers, and members of the firm hierarchy alike are typically assumed to behave “optimally”—that is, in a manner consistent with profit maximization and cost minimization—and that such behavior is the flip side or dual of maxi¬ mizing productivity. In this sense, human agency as a determinant of GNP is simply assumed away. Whether or not and under what circumstances individu¬ als can and do choose to behave differently is not a subject for discussion. But economic theory, if it is to be effective, must better incorporate human agency as a determinant of society’s material well-being. In order to do so, it becomes necessary to allow for the possibility that economic agents can choose to behave in a manner consistent with either optimal or suboptimal economic results. The behavioral model of the firm, unlike standard economic theory, al¬ lows for nonoptimal economic behavior. It assumes that economic agents within the firm, including members of the firm hierarchy, can hold prefer¬ ences or perform objective functions that can lead to goals other than pro¬ ductivity maximization. Moreover, the preferences of economic agents are profoundly affected by the system of industrial relations within the firm, be the firm small or large. In this way, suboptimal results become a possibility when modeling firm behavior (chapter 9).' 26

HUMAN AGENCY AS A DETERMINANT OF MATERIAL WELFARE

27

Of particular interest to this chapter is the x-efficieney model developed by Leibenstein (1966, 1973a, 1978a, 1979, 1983b, 1987), where it is argued that, typically, the preferences of economic agents differ. In particular, the objective funetions of workers and members of the firm hierarchy are as¬ sumed to be both different and ineompatible. Only under exceptional cir¬ cumstances, when all economic agents have preferences consistent with optimal behavior, are costs minimized and productivity maximized. These eircumstanees are closely tied to the firm’s system of industrial relations. Even if all members of the firm hierarchy have preferences for optimal be¬ havior, they will not be realized unless workers have the same preferenees, and viee versa. X-efficiency theory assumes that unless the motivational struc¬ ture and the culture of the firm yield very particular preferences, firms will not tend to produce optimally. Under sueh eonditions the results predicted by standard eeonomic theory become the exception to the rule (Frantz 1988, 1997; Schwartz 1998). Writing earlier in the twentieth century, Joseph Schumpeter (1974) was eoncemed with the general performance of economies. He stands preemi¬ nently as one economist who understood that individual entrepreneurs, or more precisely certain key entrepreneurs, are major determinants of a society’s standard of material well-being through their contribution to technical change. Max Weber (1958), stands as another well-known thinker who argued that certain mental attributes incorporated into the ideal-typical Puritan work ethic were conducive, and indeed necessary, to capitalist development. The Puri¬ tan work ethic, as an ideal type, holds that individuals who did not prefer to work hard would be relegated to an economic wasteland. It also holds that cultures that incorporate the basie ingredients of the Puritan work ethic will encourage growth and development. From the perspective of Schumpeter and Weber human agency affects economic change since all economic agents are not assumed to behave optimally at all times. An Alternative Model of the Economic Agent The premise of differential preferences or objective functions among economic agents and effort discretion, elaborated upon by Leibenstein, establishes the basis for a model in which human agency is clearly a determinant of material welfare. The model developed in this chapter is one variant of the behavioral model detailed in this book. It demonstrates how conscious human action can determine differences in productivity among firms and societies and thereby differences in the level of material well-being. The basic premise of this model is that economic agents are not all the same and are therefore characterized by different objective functions. These functions, in turn, can be affected by the

28

CHAPTER 2

work culture within the firm and, of course, by the larger socioeconomic envi¬ ronment in which the firm evolves and is embedded. These different objective functions cause differences in material wealth through their impact upon the quantity and quality of effort inputted into the process of production and tech¬ nological change. This is not to say that other factors do not also play deter¬ mining roles in generating differences of productivity among firms or differences in the wealth of nations. Rather, this model simply attempts to isolate the role of human agency in affecting these differences. As with all economic models, simplifying assumptions are required here. To begin, I assume that there are n industries, each producing a unique and homogeneous product. Within each industry, there are n firms, where n is a large enough number to prevent any one firm from affecting the market price of its output. In other words, perfect product market competition is assumed. It is important to note that this assumption differs from the product market assumption of behavioral economics in general, as well as that of x-efficiency theory, where it is assumed that imperfect product markets exist. Im¬ perfect product markets, it is argued, allow for economic agents to engage in nonoptimal behavior. Perfect product market competition would force opti¬ mal behavior on the firm’s economic agents. I further assume that each firm, within each particular industry, is characterized by identical production func¬ tions. It is therefore also assumed that each firm is producing the same prod¬ uct with the same technology. I also presuppose that there exist no external economies or diseconomies and that there are constant returns to scale, with diminishing returns to each factor input. Moreover, all firms are assumed to face the same transaction cost and institutional environment, and each firm’s economic agents are assumed to possess identical limitations with respect to their capacity to process information. Thus important exogenous variables that can affect a firm’s level of output are the same for all firms within a partieular industry. I also assume, for simplicity, that economic agents at¬ tempt to maximize their utility. Utility maximization simply refers to the assumed desire of individuals to do their utmost to maximize their personal level of spiritual and/or material well-being given the constraints they face. Utility maximization does not imply that an individual will necessarily work as hard or as well as she or he can. Rather, utility maximization is consistent with both x-efficiency and x-inefficiency. Whether x-efficiency is achieved depends on the economic agents’preferences. X-efficiency requires that eco¬ nomic agents possess very particular preferences or utility functions. Finally, I assume rational behavior on the part of economic agents in the sense that they are assumed to be consistent, calculating, and forward looking in their behavior given the constraints they face. These assumptions are not unlike those typically made by standard economic theory.

HUMAN AGENCY AS A DETERMINANT OF MATERIAL WELFARE

29

Standard theory also typically assumes that the firm’s economic agents have the same objective functions, and/or that the arguments of these objec¬ tive functions are such that each firm will behave in a manner consistent with optimality in production. Implicit in the notion of optimal economic behavior are two behavioral assumptions (chapter 3). First, it is assumed that the firm’s economic agents behave in ways consistent with the equating of marginal cost to marginal revenue. Furthermore, the same economic agents— workers, managers, and owners alike—must all work as best they can. In effect, each firm’s economic agents maximize their pace and quality of effort per unit of time and thereby minimize a firm’s unit costs, given technology. In this type of modeling of the economic agent, effort is not a variable input since it is always maximized.^ Therefore, firms are presumed to be techni¬ cally efficient or, in Leibenstein’s more meaningful terminology, x-efficient. More precisely, x-efficiency can be defined as a firm maximizing output per unit of input and therefore operating along its technically efficient produc¬ tion isoquant. This isoquant incorporates any transaction costs that result from technical considerations, such as the limited capacity of individuals to process unlimited amounts of information. Such costs are a function of ex¬ isting information technology. This view, of course, would be consistent with Herbert Simon’s view of economic man (1959,259—273; 1978,8—9,12—14). Also, the technically efficient isoquant can be assumed to incorporate any technical monitoring and metering costs generated by the need to coordinate production from an engineering perspective (Alchian and Demsetz 1972). This definition of x-efficiency is the most general one possible and is consis¬ tent with the traditional textbook ideal of economic efficiency (Leibenstein 1969,600; 1978a, 18; Blois 1974,685-686). However, socially related trans¬ action, monitoring, or metering costs are not considered to be part of the xefficient production isoquant. These costs are related to work cultures or work environments within the firm that generated higher unit costs than are objectively necessary. As discussed below, Leibenstein argues that the ideal work environment and its concomitant incentive structure is one that is rela¬ tively cooperative, wherein all economic agents have identical (or similar) objective functions with regard to the quantity and quality of effort inputted into the process of production. This definition of x-efficiency is illustrated in Figure 2.1.0, represents the x-efficient production isoquant. In this case the quantity and quality of effort is maximized. If Qq equals 0, of output, 0^ represents the x-inefficient isoquant in that at 0, the same level and quality of output can be produced with fewer labor inputs since at 0, economic agents are working harder and better than they are at 0q. A cost-minimizing firm would produce at the point of tangency between the production isoquants and the price lines (given here by a'c' and ac) such as at point e" or e'?

30

CHAPTER 2

X-efficiency, as defined here, represents an ideal level of efficiency that can be achieved by the firm. In the behavior model discussed here, just as in Leibenstein’s modeling of the firm and in efficiency wage literature, it is assumed that firms can be x-inefficient and that economic agents can therefore choose a wide range of alternative pace and quality of effort combinations per job for any given compensation package or rate of pay (AkerlofandYellen, eds. 1986; Frantz 1997; Leibenstein 1969, 602-603; 1978a; 1979, 484-486, 498; 1980, 96102; 1983a, 833; Stiglitz 1987). This assumes, of course, that contracts are incomplete with respect to pace and quality of effort per unit of time. Work¬ ers are contracted to provide a specified number of hours, weeks, or months to the firm. However, they have some discretion with respect to the pace and quality of the effort bundles that they choose to supply given the posi¬ tive economic costs involved in writing, specifying, monitoring, and en¬ forcing contracts. Because of effort discretion, at one extreme rational, utility-maximizing economic agents are characterized by preferences that yield maximum pace and quality of effort bundles. At the other extreme, their preferences yield minimum pace and quality bundles. The work envi¬ ronment of the firm affects these preferences, with different environments and cultures yielding different sets of preferences with respect to x-efficient behavior. Thus both x-efficiency and x-inefficiency in production are consis¬ tent with the behavior of rational, utility-maximizing economic agents. In the model presented here the simplifying assumption that economic agents behave as if they are attempting to equate marginal revenue to mar¬ ginal cost is not rejected. But even if the firm behaves in this manner, it will not be maximizing profits if economic agents are not behaving in a manner consistent with x-efficiency in production, assuming that the level of x-effi¬ ciency is independent of labor and related costs. Of course, the conventional wisdom assumes x-efficiency in production. In Figure 2.2, total profits are given by the area between the marginal cost curves (MCj and MC^ and mar¬ ginal revenue curve (Pq). An x-inefficient firm operates with marginal cost curve, MC,, whereas the x-efficient firm operates with marginal cost curve MCq. If a firm is more x-efficient, more effort per unit of labor time is input¬ ted into the process of production. This increases labor productivity. This, in turn, reduces marginal cost, as marginal cost equals the wage rate times the inverse of marginal product of labor when labor is the only factor input. Total profits, therefore, are greater when the firm is x-efficient and marginal costs are lower, although in both of the above cases the firm’s economic agents behave in a manner consistent with equalizing marginal cost and mar¬ ginal revenue.

HUMAN AGENCY AS A DETERMINANT OF MATERIAL WELFARE

Figure 2.1

31

32

CHAPTER 2

It is also true that if a firm’s output per unit of input is not being maxi¬ mized, if it is x-inefficient, ceteris paribus, unit costs to the firm will be higher than they would otherwise be. Unit or average cost equals factor input prices times the quantity of relevant factor inputs required to produce a unit of output. The latter term is the inverse of average factor productivity. In x-inefficient firms, average factor productivity is lower than in relatively x-efficient firms. Therefore, unit costs must be higher in the x-inefficient firms. It follows that in firms where economic agents are working harder, and better, pVofits will be higher and unit costs will be lower, ceteris paribus. This argument can be illustrated by a simple equation (2.1), where for simplic¬ ity it is assumed that labor (Z,) is the only factor input and the wage rate (w) is the only factor price. Of course, average cost (AC) is a product of the weighted input costs deflated by the weighted average productivity of factor inputs: w

0

’

(2.1)

L

where (Q) is output and (Q/L) is labor productivity. Another implication of x-inefficiency in production is that a lower level of real per capita output must characterize economies where x-inefficient firms produce a larger per¬ centage of output, ceteris paribus. This must be the case since per capita output or gross domestic product (GDP) per person is a function of labor pro¬ ductivity and labor inputs, where the level of x-efficiency affects labor produc¬ tivity. The higher the level of x-efficiency, the greater is the level of labor productivity. This can be expressed as follows:

(2.2) P

P

’

where Q/P is per capita output, Q/L is labor productivity, L is labor inputs, and P is population. That x-inefficient firms produce at higher unit costs is central to the xefficiency literature. But these relatively high-cost firms can survive in the marketplace only if imperfect product markets exist or if higher-cost produc¬ ers are sheltered by tariffs, subsidies, and/or tax breaks from the competitive pressures imposed by the more x-efficient firms (Leibenstein 1979). When sheltered from competitive pressures, economic agents can trade off higher costs and lower profits for higher prices. Leibenstein (1966,408-410; 1979, 488, 490, 492; 1973a, 772—775; 1983b) argues that increasing competitive pressures reduce the level of x-inefficiency, such that when firms are ex¬ posed to severe market pressure, the economy will become x-efficient in

HUMAN AGENCY AS A DETERMINANT OF MATERIAL WELFARE

33

production. Economic agents will, in fact, be forced or pressured into behav¬ ing x-efficiently, absent an ideal industrial relations environment. But in this case, economic agents are maximizing utility only in a very narrow sense. They are simply doing what they must to survive. Any relaxation of pressure reverts the firm back to x-inefficient levels of production consistent with the broader utility-maximization preferences of the firm’s economic agents. However, contrary to what is maintained in traditional x-efficiency theory, relatively x-inefficient firms can also survive in the marketplace when prod¬ uct markets are perfect, even if protection is not afforded to them. A central theme running through this book is that x-inefificiency can per¬ sist over time and x-inefficient production can be an equilibrium solution to the productivity problem, even under the extreme assumption of perfect prod¬ uct market competition. X-ineflficiency yields relatively high unit costs and x-efificiency yields relatively low unit costs when there is no strong relation¬ ship between wages and working conditions and the level of x-efficiency. However, low wages and poor working conditions both compensate for and can cause the low productivity of the x-inefficient firms. At relatively low wages, ceteris paribus, unit costs can be lower than they would otherwise be (Equation 2.1). But, as illustrated in Figure 2.1, the marginal cost curve of the x-inefficient firm can be driven toward MC^ by reducing wages, just compensating for any upward shift in the marginal cost curve generated by an increase in the level of x-inefficiency. In this case, the marginal cost curve of the x-inefficient firm need not be any higher than in the x-efficient firm. Bear in mind that where higher wage rates cause more x-efficiency, any down¬ ward shift in the marginal cost curve resulting from increasing the level of xefficiency might be neutralized by the increased wage rate. Given the state of the product market and the extent of protection available to x-inefficient firms, low wages can allow for the persistence of x-inefficient firms in the market¬ place. Alternatively, relatively high-wage firms can thrive in the marketplace to the extent that the high wages are compensated for by higher productivity. Firms where economic agents work harder and better, firms that are rela¬ tively x-efficient, can afford to pay higher wages. Moreover, firms that pay higher wages and provide a better work environment are more x-efficient. Both high wages and better working conditions are viewed here as contribut¬ ing to increasing the firm’s level of x-efficiency (chapter 9). In both of these cases, the level of x-efficiency does not affect the marginal cost curve since any movement in the level of x-efficiency is both caused and compensated for by changes in the rate of labor compensation. For this reason, in this scenario, and unlike in the traditional x-efficiency literature, firms can be both profit maximizing and x-inefficient. Moreover, under these assumptions, it is possible for there to be a whole

34

CHAPTER 2

range of firms, from the relatively x-inefficient to the relatively x-efficient, that produce at the same average or unit cost. This point is illustrated in Figure 2.3, where there is a unique average cost (C*) to (W*) of wages. Up to this point improvements in the level of x-efficiency serve to keep unit costs from rising in the face of rising labor costs. In terms of a firm’s com¬ petitiveness, there are here multiple equilibrium points for wage levels and labor costs relative to unit costs. Therefore, under these conditions it would be possible for both x-efficient and x-inefficient firms to exist simultaneously in competitive product market equilibrium. However, further increases in labor costs yield an increasing average cost as one enters the realm of dimin¬ ishing returns to effort inputs, given technology. In addition, wages and work¬ ing conditions can reach a low point beyond which any further reductions yield a more than proportional reduction in productivity and thereby an in¬ crease in unit costs. Such firms could not persist under severe competitive conditions. They would have to be protected to survive. In this scenario, members of the firm hierarchy have no incentive in terms of unit costs or profits to either improve or reduce the level of x-efficiency. The economic benefits of more x-efficiency accrue to workers without damaging the mate¬ rial position of the firm hierarchy. Moreover, to the extent that improvements in the level of x-efficiency require structural changes to the firm that gener¬ ate a lower level of utility of these economic agents, there exists a clear incentive for them not to build a more x-efficient firm (chapter 9). For one-person firms, for example, the assumptions listed here yield a distribution of firms by level of x-efficiency that flows directly from the pref¬ erences of the owner-operator of the firm. For the multiperson firm, the re¬ sults are similar although more complex. For simplicity, I assume a firm comprised of only two groups of economic agents: workers and members of the firm hierarchy. Leibenstein (1973) argues that x-efficient production is possible only when the preferences of all economic agents are biased toward maximum pace and quality of effort bundles. Such preferences are, in turn, at least partially a product of a cooperative workplace environment, where workers have a clear incentive to produce x-efficiently (chapter 9; Alcaly 1997; Appelbaum and Batt 1994; Ichniowski et al. 1996; Kochan, Katz, and McKersie 1986; Leibenstein 1973; Levine 1995; Levine and Tyson 1990; Rozen 1985, 1990). Otherwise, firms will produce x-inefficiently. For a multiperson firm, x-inefficiency can result simply from members of the firm hierarchy’s biases against x-efficient pace-quality effort bundles.^ In this case, even if workers are willing to produce x-efficiently, firms will be xinefficient as long as members of the firm hierarchy are not willing to invest the time and effort required to generate x-efficient production.^ The latter decision can be related to the rate of time preference relative to any expected

HUMAN AGENCY AS A DETERMINANT OF MATERIAL WELFARE

35

Figure 2.3

Unit costs

economic gains from increasing the level of x-efficiency, as well as to the leisure preference function of members of the firm hierarchy or to the eco¬ nomic gains expected (or not) from increasing the level of x-efficiency (see chapter 9). The point is that workers may be willing to behave x-efficiently only under certain working conditions that are not acceptable to members of the firm hierarchy for utility reasons, where the utility function of members of the firm hierarchy may include leisure preferences and higher incomes. Even given the preferences of workers for x-efficiency in production, the distribution of firms by their level of x-efficiency flows directly from the preferences of members of the firm hierarchy constrained by the bargaining power of labor and state intervention. In this instance, it is assumed that the preferences of the firm hierarchy are consistent with maintaining the firm’s competitiveness. This argument is illustrated in Figure 2.4, where the prefer¬ ences for levels of x-inefficiency are mapped against levels of x-efficiency. Assume that all levels of x-efficiency yield the same level of unit costs. There is a negative relationship between preferences for x-inefficiency and the level of x-efficiency in the firm. However, improving the bargaining position of labor or legislating more cooperative forms of work cultures shifts the xeffiiciency-hierarchy preference function {XP) to the right such that a given preference for x-inefficiency on the part of the hierarchy yields a higher level

36

CHAPTER 2

Figure 2.4

XP 1

of x-efjficiency, increasing fromX£'Q X.o XEy In this scenario, all economic agents are maximizing utility given the constraints that they face, no matter the level of x-efficiency actually realized. However, the utility of workers would be higher in the more cooperative work environment, where they would be realizing an improved work environment at a higher level of x-efficiency. The utility of members of the firm hierarchy might be lower in the coopera¬ tive scenario. But given that workers outnumber members of the firm hierar¬ chy, the weighted average utility in the firm increases in this case, as would the material wealth generated by the firm (chapter 4). Even if members of the firm hierarchy prefer their firm to perform xefficiently, this will not occur if workers do not have similar preferences, even in a relatively ideal work environment. Dissimilar preferences yield higher monitoring and enforcement costs with respect to contracts as the firm hierarchy attempts to force workers to work x-efificiently in a world of effort discretion. In addition, efforts to pressure workers to work harder enter into the realm of diminishing returns.^ Thus even when members of the firm hierarchy attempt to equate marginal costs to marginal revenues, the higher

HUMAN AGENCY AS A DETERMINANT OF MATERIAL WELFARE

37

Figure 2.5

contractual transaction costs and the limitations of pressure as an effort¬ enhancing device in a world of effort discretion yield lower pace and quality of effort bundles than what would be generated in a world where preferences of both the workers and the members of the firm hierarchy are oriented to¬ ward x-eflficient behavior. Moreover, the level of x-efficiency is further re¬ duced if the preferences of members of the firm hierarchy are affected by any conflict that might ensue from forcing workers to work harder than they would otherwise desire. If such conflict yields disutility, utility-maximizing members of the firm hierarchy will place less pressure on workers, and workers will work less x-efficiently than they would otherwise. In this case, the dis¬ tribution of firms by their level of x-efficiency is related to the preferences of workers and the manner in which these preferences affect the transaction costs of managing the firms’labor force (Leibenstein 1979,1982,1984,1987). X-inefficiency in production, no matter its source, must be consistent with the binding constraints faced by firms that they must be cost competitive. Either the relatively x-inefficient firms are high-cost producers sheltered from competitive pressures or they can produce at the same unit cost as the rela¬ tively x-efficient firms. This behavioral model has implications for the long-mn industry supply curve. In the conventional model, where all firms are x-efficient and assum¬ ing no external economies or diseconomies and perfect product market com-

38

CHAPTER 2

petition, the long-run industry supply curve is horizontal, such as in Figure 2.5. All identical firms are operating on the minimum point of the standard U-shaped average cost curve. In this model, factor prices are con¬ stant, and costs vary with changes in factor inputs. In one variant of the behavioral model presented here, levels of x-efficiency vary positively with changes in the wage rate and the work culture of the firm. To the extent that changes in the level of x-efficiency simply offset changes in labor costs, positive or negative, the long-run industry supply curve is not affected. How¬ ever, in this case, the horizontal long-run supply curve is composed of a range of x-inefficient and x-efficient firms producing at the same average cost. The percentage of firms in the industry that is relatively x-elficient is in part determined by preferences of the economic agents within the firm and the work environment of the firm. The greater the proportion of x-efficient economic agents, which must include the decision makers within the firm, the greater will be the percentage of firms that are x-efficient. In both of these cases no economic profits are earned in the long run. However, in an economy where there is a larger percentage of relatively x-efficient firms, society realizes a higher level of per capita income. In another variant of the behavioral model, in line with Leibenstein’s xefficiency theory, the relatively x-efficient firms produce at a relatively lower average cost than the relatively x-inefficient firms. Given factor prices, aver¬ age costs increase when economic agents are characterized by relatively xinefficient preferences such that less effort is inputted into the process of production. In Leibenstein’s analytical framework the level of x-efficiency can be determined independently of labor and related costs, or any increases in the level of x-efficiency more than compensate for changes in labor costs and other investments required to increase the level of x-efficiency. In this sce¬ nario, the level of x-inefficiency increases when members of the firm hierar¬ chy have a preference for organizational slack. They trade off increasing leisure and other perks for higher levels of x-inefficiency and higher unit costs (Leibenstein 1966,1979). This generates a positively sloped long-run industry supply curve under certain assumptions to be discussed below. The more xinefficient firms can supply output only at a higher price given their higher unit cost of production. In Figure 2.5, the more x-inefficient firms constitute the highest portion of the supply curve, while the more x-efficient firms comprise the lower portion. Given a market demand curve and market supply curve Sq, an industry price of OP^ is generated. Firms that are relatively x-efficient earn economic profits. The more x-efficient the firm, the greater the economic profits. As seen in Figure 2.5, only the /ith firm, the most x-inefficient firm given by of output, earns no economic profit. But how can the high-cost producers survive on a competitive product market in the long run?

HUMAN AGENCY AS A DETERMINANT OF MATERIAL WELFARE

39

Economic profits, according to standard economic theory, should induce more firms to enter into a given industry. In so doing, the x-inefficient firms are pressured into becoming more x-elficient. This line of reasoning is ac¬ cepted by Leibenstein, who argues that x-inefficiency can be sustained in the long run only if x-inefficient firms are protected from competitive pressures. But this need not necessarily be the case. Only firms that are relatively xefficient (more x-efficient than the nth firm) and can produce at a lower average cost than the nth firm will be attracted into the industry. The rela¬ tively more x-inefficient firms will operate at an economic loss. To the ex¬ tent that a sufficient number of x-efficient firms can be established in a given industry to supply the entire output and these firms are identical in the mini¬ mum average cost at which they can produce, it is possible that the long-run industry supply curve will be horizontal, as, for example, is PS^ in Figure 2.5, yielding an equilibrium price of OP^. This process can be illustrated by an outward shift of what would be equivalent to a short-run industry supply curve, from to 5",. The long-run supply curve (P^S^) can be comprised of both x-inefflcient and x-efficient firms, as discussed above, where both types produce output at the same unit cost. In other words, in the short run, output is supplied by both the relatively x-efficient and inefficient firms. But in the long run, the most x-efficient firms drive out the most x-inefficient firms. If, however, not enough of the relatively x-efficient firms can be established in the industry due to a scarcity of x-efficient economic agents, the long-run supply curve will be positively sloped. The relatively x-efficient firms will simply earn economic rents on their relatively x-efficient forms of firm orga¬ nization and their economic agents’ x-efficient utility functions. The x-inefficient firms can survive only if the market pressures introduced by the new and relatively x-efficient firms result in the x-inefficient firms becoming pro¬ gressively more x-efficient up to the point that they can produce at the same unit cost as the relatively x-efficient firms. The relatively x-efficient firm serve to increase the level of x-efficiency of existing firms, making the entire economy more x-efficient. Ceteris paribus, this increases a given society’s level of material well-being by increasing the economy’s per capita real out¬ put (see Equation 2.2). Relatively x-inefficient firms need not respond to the increased market pressures brought to bear by the introduction of more x-efficient firms into an industry by becoming more x-efficient. Take the case, for instance, where increasing x-efficiency requires that members of the firm hierarchy work harder and more diligently, or also necessitates that they change the organi¬ zational structure of the firm and/or reduce hierarchical perks. Owners and managers may opt not to become more x-efficient if such behavior runs con¬ trary to their prevailing preferences, and if an alternative exists that is more

40

CHAPTER 2

consistent with their preferences (see chapter 9). One such alternative might be to reduce wage rates in order to reduce costs. This would perhaps involve breaking unions or lobbying the state to reduce protection for labor organi¬ zations.^ If wage reductions do not involve reductions in labor productivity such that unit costs rise as a consequence of wage cuts, lower wage rates can result in relatively x-inefficient firms remaining competitive in the industry. To the extent that labor organizations and/or the legal infrastructure of soci¬ ety preclude wage cuts as an option, the firm hierarchy’s options are cur¬ tailed. Reorganizing for increased x-efficiency becomes the next best alternative to eventual bankruptcy. Apart from wage cuts, firms can seek protection from the state through subsidies and tax breaks. When foreign firms introduce increased competitive pressures, protection can be afforded through increased tariffs and quotas, and so on. Yet another option that is available to relatively high-cost x-inefficient firms is the construction of new plants in regions or countries where low wages and few nonpecuniary ben¬ efits to labor are the rule. This option also serves to reduce the pressure on xinefficient firms to become more x-efficient. But this option is effective only if the lower wage environment is not characterized by such high levels of xinefficiency that unit costs would be even higher than they are in the rela¬ tively higher wage environment. Nevertheless, once wage rates and state policies become variables in the model, the introduction of more relatively low-cost x-efficient firms need not result in a more x-efficient and thereby more productive society. Whether or not increased x-efficiency is generated through an increase in such competitive pressures depends on options avail¬ able to the relatively x-inefficient firms. Human Agency and Technical Change I have thus far assumed that technology remains constant for all firms. How¬ ever, the logic of the model specified here applies pari passu to a world with technical change. By technical change I refer to changes in technology con¬ sistent with inward shifts of the production isoquant or with outward shifts of the production function. The assumption that individuals are character¬ ized by different objective functions, with only a minority characterized by a Puritan work ethic, is consistent with the notion of Schumpeterian entrepre¬ neurs. Certain individuals drive technical change through their application of new and old technology to particular industries in the expectation of eco¬ nomic profits. Other individuals drive technical change by developing new technologies that can eventually be applied by the entrepreneur or, more generally, by the firm hierarchy. My concern is with the process of applying new and old technology, as opposed to the process of innovation.^ By adopt-

HUMAN AGENCY AS A DETERMINANT OF MATERIAL WELFARE

41

ing higher productivity technologies, some firms shift the industry supply curve outward, forcing technically backward firms to adopt previously un¬ used technology. If all economic agents possessed identical x-efficient objective functions and technical change were costless in terms of money and/or utility, new technology would be adopted by all firms simultaneously, provided the new technology yielded lower unit costs. In a Schumpeterian world, the process of technical change is neither automatic nor instantaneous. In the first in¬ stance, technical change can be expected to be distributed in patterns similar to the distribution of the Schumpeterian entrepreneur. Once the new technol¬ ogy is adopted by the more aggressive Schumpeterian owners or managers, the less aggressive, less highly motivated owners and managers are forced by competitive pressures into adopting the lower-cost technology. In the Schumpeterian world, therefore, technical change is adopted willingly only by the few. But in the absence of the Schumpeterian entrepreneur what drives the process of technical change? What drives non-Schumpeterian entrepre¬ neurs to initiate the process of technical change? In the model presented here, pressure on economic agents is an important causal factor driving technical change. Technical change need not take place, even if it increases productivity, unless it reduces unit costs. Moreover, if the decision makers within the firm must apply more time and effort into the process of production, at least in the short run, while the new technology is being introduced without any expectation of economic gain for themselves (in terms of increased income, for example), one can assume that they will prefer not to engage in such change. In this scenario, increasing wage rates, together with the downward inflexibility of wage rates, can induce and speed up the process of technological change. Assume that a firm is operating along isoquant as in Figure 2.6. Total costs of production are given by the isocost line a j. The relative price of labor and capital is given by the slope of the isocost line a'j, which yields a unique equilibrium combination of capital and labor given by point e of isoquant Q^. If a firm faces higher wage rates, given, for example, by the slope of isocost line ac, of output can be pro¬ duced only at a higher unit cost. The higher wage rate will result only in a profit-maximizing firm using a different factor input combination, given by point e' of isoquant Q^. A change in technology can result in of output being produced at the same unit cost as when wage rates are lower. This result is generated when the change in technology is such that isoquants Qq and represent the same amount of output. The profit-maximizing firm then operates at point e" of isoquant g,. Here unit costs are the same for the high-wage firms using the new technology embodied in isoquant as they are for low-wage firms using the old technology embodied in isoquant Q^.

42

CHAPTER 2

Figure 2.6

The high-wage firms will therefore be the ones using the new technology and producing at higher levels of productivity; the process of technical change will commence even in the absence of a Schumpeterian entrepreneur. If the low-wage firms adopt the new technology, their unit costs will fall below those of the high-wage firms. Nevertheless, the disutility of engaging in the costly process of technical change might outweigh the utility of lower unit costs. Moreover, if the new technology’s productiveness is tied to higher wages and improved working conditions and an overall restructured work culture, there is no economic reason for the new technology to be adopted. In this sense, low-wage firms face no pressing economic incentive to engage in tech¬ nical change unless a new technology that is clearly dominant is developed, such as is the case represented by isoquant Q^. The level of output represented by Q2 is equal to that represented by and Qy Unit costs at Q^, faced by the high-wage firms, are therefore lower than at (the isocost curve shifts in¬ ward from a c' to d'd) and are also lower than the unit costs faced by the low-

HUMAN AGENCY AS A DETERMINANT OF MATERIAL WELFARE

43

wage firms employing the technology embodied in isoquant Q^. However, if low wage firms can force wage rates down even further, yielding isocost curve d'g, the low-wage firms need not adopt the new technology to survive—their unit costs would once again equal those of the high-wage firms employing the most advanced technology (^2)- The development and adoption of new tech¬ nology can be regarded as a substitute for lowering wage rates (or lowering factor prices in general). Of course, this new technology need not be adopted if the state provides protection to the low-wage x-inefficient firm. Unless an al¬ ternative technology clearly dominates in terms of unit costs, low-wage firms have little incentive to engage in technical change or innovation since both exercises involve, at least initially, the investment of additional time and effort. Technical change must occur if firms cannot keep wage rates or other factor prices from rising, if firms cannot force factor prices downward when faced with competition from more cost-competitive firms, or when shelters cannot be located for relatively high-cost firms. X-inefficiency can also affect the rate or presence of technical change. In the example above, the low-wage firms engage in technical change if the alter¬ native technology is clearly dominant at the prevailing wage rate. If this alternative technology is embodied in isoquant Q , it is adopted by the lowwage firms characterized by the relative price of labor given by isocost line a'j. However, if members of the firm hierarchy expect that the potential shift of the production isoquant from Qq to Q cannot be realized, if the firm’s internal organization promotes significant levels of x-inefficiency, the alter¬ native technology need not be dominant; and it would not then be adopted by the low-wage firms. If, due to x-inefficiency, the production isoquant is expected to shift inward only to or, even worse, somewhere between and Qp unit costs will remain lower for the low-wage x-inefficient firms under the old technology. For technical change to occur in the low-wage firms, the level of x-inefficiency expected to be embodied in the new tech¬ nology would have to be significantly reduced. The existence of x-inefficiency can therefore impede or even prevent technical change. Increasing wage rates can both induce technical change and reduce the level of x-inefficiency. The reduction in x-inefficiency itself has the effect of making the introduction of previously unused technology a more economical proposi¬ tion to members of the firm hierarchy.’® In this model, technical change is an economic event induced through pres¬ sure upon the firm’s economic agents—^in particular the members of the firm hierarchy." The clear exception is that of the Schumpeterian entrepreneur, who executes technical change without being pressured to do so. However, the expectation of economic profits must be present for technical change to take place. I have emphasized the critical role played by increasing wage rates and 2

2

44

CHAPTER 2

the downward inflexibility of wage rates and, more generally, by increasing factor prices and their downward inflexibility in inducing technical change. Once technical change is induced in some firms and the industry supply curve shifts outward or downward, it coaxes hierarchies of low-wage firms to intro¬ duce the new technology and/or reduce the level of x-inefficiency to the extent necessary to allow these firms to survive at the lower equilibrium market price (see Figure 2.5). Of course, there are alternatives to technical change. The firm hierarchy can attempt to lower wage rates for other factor prices or attempt to engage in activity that guards the older technology firms from increased com¬ petitive pressures. However, it is technical change, not reductions in factor prices, that yields increases in the potential material well-being of society. It would be useful here to reconcile the results of this model with better known conclusions drawn from Ester Boserup’s (1965,1981) model of tech¬ nical change and with those of the more recent model of Reuven Brenner (1983). Both argue that population growth serves to stimulate innovative activity and to encourage the adoption of new or hitherto unused old tech¬ nologies. In Boserup’s model, population growth serves to reduce the land to labor ratio, which in turn reduces output per unit of land and, ceteris paribus, also reduces output and, finally, income per peasant farm family. In order to keep income per family from falling, more labor-intensive techniques of pro¬ duction need to be applied. An alternative, not necessarily new, technique of production is adopted. Increasing wage rates play no role in affecting techni¬ cal change or changes in techniques of production. If anything, the Boserup model is most consistent with falling real wages. In the model presented in this chapter, higher wage rates encourage technical change because they threaten the competitive position of the firm and even its very survival. In the Boserup model hired hands are not important; the family farm is analo¬ gous to the firm in my model. Boserup sees the survival of the family farm as threatened by an increasing population. The pressure, thus brought to bear by external factors, forces the farm family to work harder and longer. According to Brenner’s causal hypothesis on technical change, increas¬ ing population coerces economic agents to engage in innovative activities. Brenner argues that innovation is analogous to gambling. It is a risky activ¬ ity. Increasing population, he argues, results in some economic agents find¬ ing themselves in a relatively worse economic position. Since Brenner assumes that economic agents prefer to be higher rather than lower on the wealth or income distribution ladder, any deterioration in an economic agent’s relative position induces him to engage in more innovative or risky activity. Such activity can shift the firm’s production isoquant inward if the firm adopts the newly developed technology. In the Brenner model, higher wage rates serve no positive role in inducing innovative activity.'^

HUMAN AGENCY AS A DEtERMINANT OF MATERIAL WELFARE

45

In the model presented in this chapter, though, any event threatening the economic well-being of the firm or economic agents, writ large, can produce a reaction that is directed toward at least preserving the ex ante economic status. For innovative activity to be precipitated, an economic agent’s or firm’s position on the income distribution ladder need not be affected. On the other hand, neither innovative activity nor technical change need result from pres¬ sure against economic agents or firms in my model. Any activity that yields similar results with less work, effort, or risk (where economic agents are adverse to risk) can instead be adopted. Sheltering activity or reducing wage rates are possible strategies. Nevertheless, according to both Boserup’s and Brenner’s models, as well as mine, pressure plays a crucial role not only in affecting, but also in initiating positive economic change since in all three cases the typical economic agent is behaving x-inefficiently in the absence of pressure. Utility, Leisure, and Welfare One implication of the model presented here is consistent with standard eco¬ nomic theory: there is no such thing as a free ride. A more x-efficient economy characterized by higher rates of technical change requires economic agents to work harder and to take more risks than economic agents in a less xefficient society with less technical change. Moreover, to the extent that eco¬ nomic agents prefer to be x-inefficient, pressure that forces x-efficient behavior or technical change can be expected to reduce the utility of most individuals. This situation can change over time if the preferences of these individuals change toward more x-efficient behavior and behavior that is more condu¬ cive to technical change, resulting in a different work ethos or work culture. As discussed above, dynamic entrepreneurs in the economy force less dynamic entrepreneurs to become more cost efficient. If this results in the latter group’s engaging in more x-efficient behavior, the group’s utility might be reduced. What of the employees? If it is possible to force employees to become more x-efficient without compensating them in a manner consistent with their preferences, their utility can be expected to diminish. However, if more x-efficient behavior ean be induced only by providing the employees with incentives (such as increased real income or improvements in working conditions), it is possible for the utility of employees to increase or, at the very least, not to diminish. Alternatively, if firms can become more cost effi¬ cient by cutting the price of labor—^and this option is preferable to the mem¬ bers of the firm hierarchy—the utility of the employees can be expected to fall, whereas the utility of the entrepreneurs need not change either way. The potential level of society’s material well-being, though, will

46

CHAPTER 2

increase only if firms become more x-efficient. Whose material well-being increases depends, of course, on how the increases in output are distributed (see chapter 4). Technical change, it is assumed, forces itself upon all but the most dynamic, energetic entrepreneurs, those characterized by the ideal-typical Fhiritan work ethic. Higher wage rates become one source of pressure-inducing technical change. Since technical change is a costly process, at least for the firm hierar¬ chy, it can be expected to reduce the utility of members of the hierarchy. To the extent that if also involves increased risk (Brenner 1983), there is an additional cost that can further reduce the hierarchy’s utility. But technical change ulti¬ mately results in increased output per worker and thereby contributes, at least potentially, toward increasing the material well-being and thereby the utility of employees. Once technical change is realized in the firm, it is unlikely that the new technology will be more costly to maintain than the old, less productive technology. Therefore, once adopted, the new technology can yield the same utility to members of the firm hierarchy as did the old technology. To compete with the more technically advanced firms, the less advanced firms can choose to cut wage rates instead of opting for technical change. This option may be chosen if wage rates are flexible downward and if the process is less costly, at least in the short mn, than the process of technical change, though wage reduc¬ tion diminishes both the material well-being of workers and their utility. More¬ over, this strategy may fail when the wages cuts yield proportional or more than proportional reductions in productivity. Since increasing the level of x-efficiency and engaging in technical change are assumed to be costly processes, who benefits from increasing x-efficiency and technical change depends largely upon the dispersion of the increased output and on the marginal utility that economic agents associate with the increase in material welfare. Such welfare is relative to the marginal disutility associated with the increased quantity and quality of effort required to in¬ crease x-efficiency. Improvements in the level of x-efficiency and technical change can also cause short-term job loss for both employees and manage¬ ment. These costs must also be included in any assessment of the net benefits of efficiency improvements and technical change. Conclusion In standard economic theory, a society’s level of material well-being is deter¬ mined mainly by the production function and by exogenously given technical change. Institutions can also play a role in the process of wealth generation through the exogenous constraints they impose on firm behavior. Economic agents are also assumed to behave x-efficiently. In this sense, individuals are

HUMAN AGENCY AS A DETERMINANT OF MATERIAL WELFARE

47

devoid of free will. Economic agents cannot choose to deviate from x-efficient behavior, and for this reason, they cannot be induced or pressured into becom¬ ing more x-efficient. In an economic world where individuals are understood to have some freedom in imposing their preferences with respect to the quality and pace of effort in their job performance and these preferences are seen to deviate from x-efficient behavior, human agency can have significant effects on the efficiency and growth of an economy. Individuals imbued with a strong work ethic can strongly influence the level and rate of growth of material pro¬ duction, given the economic and institutional constraints faced by firms. More¬ over, the work culture of the firm, which can affect and help mold the work ethic within the firm, can have the same effect. The model presented in this chapter implies certain fundamental testable hypotheses. If a strong work ethic characterizes economic agents, this can be expected to make their firms, industries, and societies materially wealthier than they would otherwise be. Such firms, industries, and societies impose competitive pressures on the economic agents of other firms, industries, and societies to engage in more x-efficient behavior. Such behavior can be avoided by sheltering the threatened economic entities from competitive pressures or by cutting factor prices such as wage rates. The latter strategies will keep firms, industries, and societies relatively less productive and thereby rela¬ tively impoverished. Another hypothesis suggested by my model is that, ceteris paribus, low-wage industries and societies should tend to experience lower levels of x-efficiency and slower rates of technical change than those achieved by higher-wage industries and societies. The model of human agency and differential preferences presented in this chapter strongly suggests that an important facet of any inquiry into the “wealth of nations” should be a focus on the work culture of firms and societies in general. Cultures emphasizing hard work contribute to improving a society’s level of potential material well-being and its competitive position in the mar¬ ketplace. In addition, institutions that either encourage increases in labor com¬ pensation or impede its reduction act as mechanisms for realizing improvements in a firm’s, industry’s, or society’s level of x-efficiency and even for improving their competitive position in the marketplace. Such hypotheses and queries into economic development can be relevant only in an economic world where individuals are understood to be able to choose to behave x-inefficiently. In the real world, of course, such choices are actually made. Notes 1. See Leibenstein (1979, 69-74) for an excellent summary of behavioral eco¬ nomics and Kaufman (1989, 1990) for a recent expression of the behavioral view of

48

CHAPTER 2

the firm. Schwartz (1998) provides a detailed summary and discussion of behavioral economics. See Shen (1991) for a mathematical expose of x-inefficient behavior of economic agents. J.L. Simon (1991) treats effort as a variable, thereby allowing for xineflficiency that is a function of an economic agent’s wealth and the opportunities to realize more wealth. On the latter subject, see also Brenner (1983). 2. More recent literature in efficiency wage theory, argues that effort per unit of labor time is a variable and need not be typically at some maximum. However, this literature assumes that members of the firm hierarchy will behave in a manner consis¬ tent with minimizing the efficiency wage or the wage to marginal product of labor ratio. In that model, a unique real wage minimizes the efficiency wage, and this unique wage will, at a minimum, be searched for by members of the firm hierarchy. See Akerlof and Yellen (1986a), Katz (1986), Shapiro and Stiglitz (1984), Solow (1986), and Stiglitz (1987). 3. This view of economic efficiency is rejected by conventional theorists such as George Stigler (1976) and Louis De Alessi (1983) since they maintain that economic efficiency simply means cost and output results that are consistent with constrained cost-minimizing and profit-maximizing behavior. Thus, by definition, all behavior becomes minimizing or maximizing once we specify the prevailing constraints. It follows, therefore, that costs are being minimized at all times. According to such a view, there can be no unique minimum unit cost since minimum costs can vary as constraints change. This perspective is, as Leibenstein points out (1973a, 210; 1987, appendix), one that makes neoclassical microtheory into a tautology. One cannot use the x-efficiency definition of efficiency in this general form to determine what is or is not efficient production since production is always by definition efficient. 4. It should be noted that Leibenstein makes the case, quite controversially, that economic agents behave only quasi-rationally in the firm when they do not behave in a fashion consistent with x-efficiency. However, although Leibenstein does argue that for his purposes dropping maximization and minimization from one’s analytic tool box is best, he does not contend that these concepts are necessarily incompatible with x-inefficiency. Indeed, Leibenstein (1985,12—13) is largely concerned that these con¬ ventional concepts should not be used in a tautological fashion. If, in fact, economic agents are largely partially maximizers-minimizers, as Leibenstein would have it, then the case for the existence of x-inefficiency becomes even stronger (see also Leibenstein 1986). 5. Scitovsky (1943-44) demonstrates that for profit maximization to take place, the utility of the entrepreneur can be maximized only when profits are maximized, irrespective of the sacrifice incurred in terms of time and effort. 6. Tomer (1987) points out that x-efficient production requires the investment of organizational capital, which includes a sacrifice in terms of time and effort on the part of members of the firm hierarchy. 7. Referring to the Yerkes—Dodson Law, which relates “performance” to stress, Leibenstein (1987,18-20, ch. 8) argues that pressure or stress will affect performance positively only up to a limit. Thereafter, one would expect performance (effort in this case) to diminish. Simply trying to force workers to work harder will eventually hit a wall. 8. See Bluestone and Harrison (1988, 1990) for a discussion of low-wage strate¬ gies toward regaining a competitive position in the marketplace. 9. See Mokyr (1990) for an elaborate discussion of the process of technological innovation through time.

HUMAN AGENCY AS A DETERMINANT OF MATERIAL WELFARE

49

10. See Leibenstein (1973) for a discussion of the relationship between x-inefficiency and technical change. 11. The arguments presented here can be seen as part of the induced technological change literature, although the model here emphasizes the causal role of increasing labor costs to shifts in the production isoquant as opposed to movements along it (see chapter 6; Hicks 1932; Hayami and Ruttan 1971; Ruttan 1997). 12. Although increasing wages explicitly play no positive role in inducing techni¬ cal change, such a positive causal relationship is consistent with the logic of Brenner’s model. As long as higher wage rates cause a fall in the relative economic position of members of the firm hierarchy, one would expect that the hierarchy would be induced to engage in technical change if this served to restore or improve its standing in the distribution of income or wealth.

3

The Methodology of Economics and the Survival Principle Revised

Introduction Following the publication of Milton Friedman’s classic work on the method¬ ology of economics in 1953, few economists have challenged the assertion that the realism of assumptions is of little analytical consequence to eco¬ nomic theory and therefore to the formation of economic policy. This chap¬ ter looks at the potential substantive significance of the realism of assumptions for economic welfare and economic justice by way of its effect on the ana¬ lytical predictions of economic models and the policy implications of these predictions, as well as the effect it might have on the scope and breadth of economic analysis.' Economic analysis and policy flow from economic models either explic¬ itly or implicitly. Economic models establish the roadmap and the signposts for economic discourse. As Friedman put it, “A theory is the way we per¬ ceive ‘facts,’ and we cannot perceive ‘facts’ without a theory” (1953a, 34). Theory is an engine designed to analyze the world as it is (Friedman 1953a, 35). See also Thomas Kuhn’s remarks in Coase (1994, 27). The construction of economic theory is therefore of fundamental analytical importance to the design of public policy, as well as in gamering an understanding of the costs and benefits of choosing one policy over another. Here it is argued that to the extent economic models are misspecified as a result of inappropriate behavioral assumptions, the analytical predictions gen¬ erated by the models will be incorrect. Even if the predictions are reasonable 50

METHODOLOGY AND SURVIVAL PRINCIPLE

51

with respect to the stylized facts, causality will be misspecified. Moreover, the roadmap or frame of reference provided by the models will be inaccu¬ rate, thereby directing attention away from true causality. By assuming that the realism of assumptions is not of substantive importance, potential ana¬ lytical problems are in turn assumed away. This is not a costless process. The opportunity cost of building models upon rocky foundations and false be¬ havioral premises can be measured by the economic cost of the policy errors derived from such theories, as compared to what could have been generated by more realistic theories. My focus is on one particular set of behavioral assumptions underlying much of the conventional wisdom and their implications for modeling the economy and constructing public policy. The conventional wisdom assumes (as noted above) that economic agents tend to work as hard as they can and that work effort is determined independently from the wage rate and the industrial relations system of which the rate of labor compensation is but one component. These assumptions underlie the survival principle, which, in turn, informs so much of the theoretical and empirical work in economics. I argue that these assumptions are largely incorrect or, more specifically, represent an extreme, atypical scenario. The more general case is one where effort is variable and a product of the behavior of individuals inside the firm. Build¬ ing models predicated upon more reasonable assumptions yields analytical predictions that lend support to causal explanations that differ from the stan¬ dard fare, with important analytical and policy implications.

Friedman’s Methodological Core Empirical Instrumentalism In Friedman’s methodological perspective the analytical predictions of eco¬ nomic theory must be consistent with the facts. Economic theory consists of a set of generalizations that yield predictions “about the consequences of any change in circumstances.” The efficacy of a theory must be evaluated in terms of the “precision, scope and conformity with experience of the predictions it yields” (1953a, 4). Whether or not a theory is appropriate or valid is therefore an empirical question. It is the precision of the prediction as compared to the known data, to which the analytical prediction speaks, that allows one to dif¬ ferentiate between a false (refuted) theory and an apparently tnie theory. The best theory is simply one that is not refuted in terms of the match between prediction and data. And the best theory is true only in the sense that its predic¬ tions prove to be a good match with the relevant data. Better theories yield more precise predictions so that progress in economic modeling is measured

52

CHAPTER 3

by getting the analytical predictions more closely attuned with the relevant data (Freidman 1953a, 5-6). Clearly, for Friedman economics is an empirical science insofar as analytical predictions must always relate to real world eco¬ nomic questions and be tested in terms of their fit with the facts. Friedman does not recommend an economics that focuses upon the logi¬ cally consistent derivation of analytical predictions from fundamental axioms or assumptions but that pays little heed to the relationship between the logi¬ cally derived predictions and the data—^what some have derided as blackboard economics.^ The latter has become the bread and butter of much of contempo¬ rary economics. Friedman argues: “Logical completeness and consistency are relevant but play a subsidiary role; their function is to assure that the hypoth¬ esis says what it is intended to say and does so alike for all users. They play the same role here as checks for arithmetical accuracy do in statistical computa¬ tions” (1953a, 10). He adds: “Formal logic and mathematics, which are both tautologies, are essential aids in checking the correctness of reasoning, discov¬ ering the implications of an hypothesis, and determining whether supposedly different hypotheses may not really be equivalent or wherein the differences lie. But economic theory must be more than a structure of tautologies if it is to predict and not merely describe the consequences of action; if it is to be some¬ thing different from disguised mathematics” (1953a, 11—12). For Friedman (1953a, 12) it is critically important to continuously develop data sets so as to determine the practical utility of particular theories, as well as to contribute to the development of new theories. Moreover, new theories should force scholars to develop new data sets so they can determine the extent to which the theories’ predictions are in concordance with the data. More numerous and more refined data sets should force us to revise our theories so that the analytical predictions that emanate from them better fit the data. Theories must be revised or at least refined when confronted with facts that either refute the predictions of the theories or cast some doubt on their predictive powers. However, Friedman maintains that predictions might also simply be off the mark because of the inadequacy of the available data. It is therefore possible that what appears to be a poor theory in terms of its predictive power might actually turn out to be a good theory if the appropri¬ ate data set could be developed to test it (Friedman 1953a, 30). Friedman is clearly an empiricist in his recommended approach to economics in the sense that facts matter and are of fundamental importance to the testing and devel¬ opment of economic theory. However, he is also an instrumentalist in the sense that theory is a tool designed to generate analytical predictions, where the adequacy of a theory is determined by its predictive power (Caldwell 1980).3

METHODOLOGY AND SURVIVAL PRINCIPLE

53

Behavioral Assumptions and Economic Theory

The assumptions that underlie a theory are not of analytical significance in Friedman’s recommended methodology. Indeed, efforts to determine the re¬ alism of a theory’s assumptions are thought to be fruitless endeavors with respect to enriching our understanding of an economy. As long as the predic¬ tive power of the theory is strong, the realism of the underlying assumptions is unimportant. Testing for the realism of a theory’s assumptions, therefore, can neither replace nor supplement testing a theory’s predictive power. Determining the assumptions’ realism, Friedman argues, “is fundamentally wrong and pro¬ ductive of much mischief Far from providing an easier means for sifting valid from invalid hypotheses, it [testing for the realism of assumptions] only confuses the issue, promotes misunderstanding about the significance of empirical evidence for economic theory, produces a misdirection of much intellectual effort devoted to the development of positive economics, and impedes the attainment of consensus on tentative hypotheses in positive eco¬ nomics” (1953a, 14). Assumptions, Friedman continues, can never be realis¬ tic since they are abstractions from reality and therefore must be descriptively false. Important theories contain assumptions that might even be “wildly inaccurate descriptive representations of reality, and, in general the more significant the theory, the more unrealistic the assumption (in this sense)” (1953a, 14). This must be tme since successful theories and their hypotheses must abstract “the common and crucial elements from the mass of complex and detailed circumstances surrounding the phenomenon to be explained and permit valid predictions on the basis of them alone” (1953a, 14). In a nutshell, Friedman argues that “the relevant question to ask about the ‘as¬ sumptions’ of theory is not whether they are descriptively ‘realistic,’ for they never are, but whether they are sufficiently good approximations for the pur¬ pose at hand” (1953a, 15). The latter is determined by the predictive power of the theory. Hence the test for the validity of the assumptions is collapsed into the test for the predictive power of the theory. If the theory works, it follows that the theory’s underlying assumptions must work and are suffi¬ ciently realistic. This would be the case even if the behavioral assumptions were quite explicitly incorrect since it is possible for incorrect assumptions to generate correct predictions (Friedman 1953a, 20).^^ A new set of behav¬ ioral assumptions can be determined to be better than the old only if it yields more powerful predictions. This would be true even if it were known that the new set was wildly unrealistic (Friedman 1953a, 30—34). How individuals actually behave or typically behave is not important for

54

CHAPTER 3

Friedman. Moreover, his preferred methodology can give rise to a situation where false assumptions yield correct predictions since there can be an array of theories based on assumptions ranging from completely false to com¬ pletely true that yield predictions of the same power. In sum, Friedman’s preferred methodological approach boils down to correlation analysis. It does not seek to determine the potential causes of a particularly high correlation. It recommends statistical correlation over and above causal explanation.^ If the prediction is of an acceptable power, one can simply assume that the individual behaves as if he or she followed the dictums of the theory’s be¬ havioral assumptions. Friedman’s methodological stance is well illustrated by the now famous examples he provides in his essay. These offer an excellent vantagepoint from which to access the analytical consequences of his instrumentalist approach to economic theory. The first example is that of predicting the shots made by an expert billiard player (Friedman 1953a, 21). Friedman argues that one can predict such shots by assuming that the player behaved as if he or she knew and applied the mathematical formulas consistent with producing the shots generated on average by an expert player. Such an assumption, Friedman ad¬ mits, is wildly unrealistie. However, the power of prediction generated by it is high. This is a case of an unrealistic assumption yielding good predietions. Friedman accepts the “as if’ hypothesis since “unless in some way or other they [expert billiard players] were capable of reaching the same result, they would not in fact be expert billiard players.” This approach sidesteps the se¬ quence of causality in terms of how billiard players actually became experts. In terms of recommending how to develop expert billiard players Friedman’s “as if” hypothesis would be a poor substitute for determining how billiard players become experts. A training program for billiard players concentrating on math and engineering courses would by itself not produce expert billiard players. If true assumptions generate predictions of the same power as the “as if’ assumptions, then the hypothesis containing the true assumptions would be superior from a causal and policy point of view. The Survival Principle and the Conventional Wisdom

Friedman’s example of his methodological approach from the economic realm, proffers a similar moral to the one provided in the parable of the expert bil¬ liard player. In this case, however, Friedman brings to the fore the survival principle, which has played, either explicitly or implicitly, a critical role in theory-building in neoclassical economics.® A conventional and fundamen¬ tal economic hypothesis, the maximization-returns hypothesis, stipulates that “firms behave as if they were seeking rationally to maximize their ex-

METHODOLOGY AND SURVIVAL PRINCIPLE

55

pected returns . . . and had full knowledge of the data needed to succeed in this attempt; as if, that is, they knew the relevant cost and demand functions, calculated marginal cost and marginal revenue from all actions open to them, and pushed each line of action to the point at which the relevant marginal cost and marginal revenue were equal” (Friedman 1953a, 21). Friedman admits that business people “do not actually and literally solve the system of simultaneous equations in terms of which the mathematical economist finds it convenient to express this hypothesis” (1953a, 22). What is important, however, is that a set of admittedly unrealistic assumptions yields predictions of economic performance that, according to Friedman, are con¬ sistent with the conditions necessary for the survival of the firm on the mar¬ ketplace. If individuals behaved in a fashion that generated performance results inconsistent to those predicted by the “as if’ theory of the firm, the firms managed by such individuals would not survive. Friedman argues that the prediction of this theory “summarizes appropriately the conditions for sur¬ vival” (1953a, 22). The process of natural selection in the marketplace in this sense serves to validate the theory even if the theory is predicated upon false behavioral assump¬ tions. As with the expert billiard player example, Friedman here too sidesteps the question of the type of behavior that allows business people to survive and even prosper in a competitive environment. If business people apply a different set of rules for survival than those specified in the “as if’ theory of the firm, then this theory should not form the basis for policy, nor can the theory explain how firms manage to survive in a competitive market. More generally, if individual behav¬ ior that is inconsistent with maximizing expected returns is compatible with longrun survival, the theory would fail to predict or explain actual firm behavior. Moreover, as with the expert billiard player, if a hypothesis of firm behavior based on tme behavioral assumptions generates predictions of the same degree of accuracy as those flowing from the maximization-returns hypothesis, the hy¬ pothesis containing the tme assumptions would be superior from a causal and policy point of view. Nevertheless, as Melvin Reder points out, the principle remains funda¬ mentally important to contemporary economic reasoning, although it takes on two specific forms. In the weak, relatively evolutionary form, the prin¬ ciple claims that “competition will lead to the selection of lower cost meth¬ ods of performing any task so that in the long mn the surviving method(s) will be those associated with the lowest cost.” In its strong form, which is more closely linked with many adherents of the Chicago School, it “holds that at any moment the expected cost of the method actually in use will be ‘close’ to the least cost method available. In individual cases, deviations of actual from least cost methods are interpreted as reflections of changes in

56

CHAPTER 3

technology or factor prices that had been unanticipated when the productive technique was adopted. Thus the strong form of the survival principle says that, at any moment in time, the decision rules being followed come close to minimizing the expected cost of doing whatever each decision maker is at¬ tempting” (Reder 1982, 383).The flip side of cost minimization is typically held to be output maximization per unit of labor input. In other words, it is assumed that effort discretion does not exist, and therefore all firms that sur¬ vive must be x-efficient in production. More to the point, cost minimization is viewed, ill the modem variant of the conventional wisdom, to be an exer¬ cise of constrained utility optimization on the part of individuals whose con¬ straints include transaction costs and limited information-gathering and processing skills. In both its weak and strong forms the survival principle generates the ana¬ lytical prediction of least cost behavior as the only possible pattern consistent with firm survival in a competitive market. Moreover, the conventional wis¬ dom (especially its Chicago variant) typically assumes that currently observed prices and quantities are good approximations of their long-mn competitive equilibrium values, so that competitive conditions mle over the long haul even if monopolistic conditions and government intervention produce impediments to the competitive process. In this sense, only one type of economic behavior is possible because only one type is economically viable. And economically vi¬ able firms are those that are x-efficient. Theoretical models that generate pre¬ dictions that are apparently in contradiction with the survival principle or empirical work of the same tenor tend to be rejected because such results are not plausible in a market that is competitive, at least over the long haul. This approach is partially a product of the fact that in economics, “The necessity of relying on uncontrolled experience rather than on controlled experiment makes it difficult to produce dramatic and clear-cut evidence to justify the acceptance of a tentative hypothesis” (Friedman 1953a, 40). Following upon the survival principle, the conventional wisdom views the predictions of the optimizing model of the firm to be of the highest power, to be the most accurate available irrespective of the evidence. In this vein, the conventional wisdom treads closely upon Hume’s (1993) ought/is fal¬ lacy: observers of a phenomenon often maintain that what is ought to be because it is. Proof of the inevitability of a phenomenon is contained in its existence. In this case, the assumptions of the model subsume the empirical side of Friedman’s recommended methodology. There is a proclivity to re¬ ject contrary empirical evidence as flawed so that the theory and its underly¬ ing assumptions become the determinants of what is true or false. This allows conventional economists to make reasonably strong and unambiguous pre¬ dictions with respect to the impact of changes in particular economic vari-

METHODOLOGY AND SURVIVAL PRINCIPLE

57

ables. The survival principle holds when economic behavior relates to out¬ comes that are subject to the discipline of the marketplace, when a choice by an individual has meaningful consequences for the survival of the firm. It does not apply when choices have no or only negligible consequences for the survival of the firm. This relates to the broad and important questions of consumer choice, ranging from the determination of a consumption basket or consumption patterns, the purchase of financial instruments, the choice of marriage partners, and the number of live children desired to the consump¬ tion of addictive products.^ The Survival Principle Revisited and Revised The survival principle is a powerful check on analytical predictions. If a set of predictions, whatever their underlying assumptions, is inconsistent with the survival of the economic entity, the theory or theoretically deduced policy rec¬ ommendation is cast in doubt. But if the conventional wisdom’s framing of the survival principle is incorrect, then it should not serve as a standard by which to judge the rigors of competing economic theories. In particular, if firms that are not maximizing output at any given time can survive in competition with firms that are (effort discretion exists), one can no longer predict with any degree of confidence that nonmaximizing firms will be destroyed in the com¬ petitive process. Both relatively efficient and inefficient firms would survive in long-run equilibrium. To survive, firms must only remain cost competitive. However, cost competitiveness does not necessarily imply output maximiza¬ tion. In other words, the survival principle still holds but in a more general and weaker form. The revised version of the survival principle should read: To survive in the marketplace an economic entity must remain cost com¬ petitive.

Only in a world of perfect or complete x-efficiency would the conven¬ tional interpretation of the survival principle hold. The conventional version is a special case, subsumed under the more general one that focuses on unit costs of production where the possibility of x-inefificiency in production is allowed for. The question then arises as to (i) the role that the realism of the modeling assumptions plays in constructing economic theory inclusive of the survival principle, and (ii) how and the extent to which the revised and more general version of the survival principle affects the modeling of the economy, thereby allowing for theories that generate substantively different and more accurate analytical predictions than those generated by the con¬ ventional and narrower version of the survival principle. That not all assumptions need be realistic is a fundamental premise of model building. This must not be confused with the assertion contained in

58

CHAPTER 3

Friedman’s instrumentalist methodology that behavioral assumptions are unimportant to the efficacy of a theory. As Alan Musgrave (1981) points out in his examination of Friedman’s methodology, there are at least three differ¬ ent types of fundamental assumptions. First, the negligibility assumption specifies that a particular factor has only a negligible or substantively insig¬ nificant effect on the phenomenon being investigated. In this case, one’s hy¬ pothesis or theory is structured as if this factor does not exist. This unrealistic assumption serves as a simplifying assumption. However, the veracity of the negligibility assumption itself must be tested. Second, the domain assump¬ tion specifies the circumstances under which a theory holds. If a factor is analytically nonnegligible and the theory does not hold when this is the case, the domain assumption specifies that the theory is valid only in a world with¬ out this factor or when this factor is of negligible importance. Thus a theory can be built unrealistically assuming the absence of this factor. For example, a theory can be built assuming no effort discretion, as long as one specifies that it holds only under conditions of no effort discretion. The question then arises as to how applicable this theory is; does it apply only under excep¬ tional circumstances? Third, heuristic assumptions assume that a factor is negligible, even if it is not, in order to determine the analytical significance of this factor. In this case, unrealistic assumptions serve the analytical pur¬ pose of attempting to determine the potential causal relationship among vari¬ ables. Friedman and the conventional wisdom assert, as a general principle, that it is unimportant to determine the extent to which a variable is nonnegligible. What counts is the power of the prediction: correlation, not causation. This is also the case with regard to the survival principle. However, in this case a domain assumption is made, at least implicitly, that the analytical predic¬ tions of the theory hold only if inefficient firms cannot survive in a competi¬ tive market. What is critical to my discussion is whether the noimegligibility of certain behavioral assumptions has an analytically meaningful impact on causal and even correlation-type economic analyses. X-Inefficiency, Efficiency Wages, and the Conventional Wisdom The conventional wisdom assumes that economic agents work as hard and as well as they can: the quantity and quality of effort per unit of labor input is fixed at some maximum. Certainly firms comprised of individuals who con¬ sistently deviate from such a maximum would fail the test of survival in the marketplace according to the conventional wisdom. The absence of effort discretion is a critical domain assumption of the conventional wisdom, whereas its existence is a critical domain assumption of x-efficiency theory.

METHODOLOGY AND SURVIVAL PRINCIPLE

59

There is, however, considerable evidence for the existence of effort discre¬ tion as a general as opposed to a special case, where effort discretion is at least partially attributable to the costs of drawing up, signing, monitoring, and enforcing contracts.^ Leibenstein concludes as follows: Firms and economies do not operate on an outer-bound production possi¬ bility surface consistent with their resources. Rather they actually work on a production surface that is well within that outer bound. This means that for a variety of reasons people and organizations normally work neither as hard nor as effectively as they could. In situations where competitive pres¬ sure is light, many people will trade the disutility of greater effort, of search, and control of other peoples’ activities for the utility of feeling less pres¬ sure and better interpersonal relations. But in situations where eompetitive pressures are high, and hence the costs of such trades are also high, they will exchange less of a disutility of effort for the utility of the freedom from pressure, etc. (1966, 413) In this sense, x-efficient production represents the maximum level of output that can be realized by utility-maximizing agents. There is now considerable evidence for the existence of x-inefficiency (Frantz 1997). Although Leibenstein argues that x-efFiciency can be realized under a cooperative system of industrial relations, his focus is on market forces as a, if not the, fundamental determinant of the level of x-inefficiency—the devia¬ tion of labor productivity from its maximum. Thus as the market becomes less competitive or more monopolistic, firms can get away with being rela¬ tively x-inefficient. X-inefficiency implies, ceteris paribus, higher unit pro¬ duction costs, so that some form of protection is required in Leibenstein’s modeling for the x-inefficient firm to survive in the marketplace. The persis¬ tence of protection for the x-inefficient firms is another domain assumption of x-efficiency theory. Average costs are given by the following basic equation assuming, for simplicity, that labor is the only input into the production function:

AC^

(3.1)

where ACis average costs, w is rate of labor compensation, L is labor input, and Q is output. This equation can be rewritten as:

60

CHAPTER 3

Average costs are determined by the labor productivity and the wage rate. Ceteris paribus, the lower the labor productivity (the less x-efFicient is the firm), the higher the average or unit costs of production. A firm protected from market forces can charge higher prices to cover the higher costs generated by the x-inefficient behavior of its economic agents. As long as product markets are highly imperfect over time, x-inefflciency can per¬ sist, and the survival principle cannot predict that the market will generate xefficient firms or x-efficiency in production. However, few mainstream economists accept the realism of the domain assumption of the persistence of protection to the x-inefficient firms as a general case, especially when protection takes the form of a dearth of competitive pressure.^ In this sense, x-inefficiency cannot be expected to be more than a transitory phenomenon. Nevertheless, Leibenstein points to the probable and predictable specific economic losses to society to the extent that protection is afforded to the x-inefficient firm in the absence of some ideal system of industrial relations prevailing in the firm. Given effort discretion and protection, one cannot infer from the survival or even from the prosperity of a firm that it is x-efficient in production. In this case, the realism of the domain assumptions is of fundamental importance. In addition, what economic agents actually do in the firm becomes of utmost importance if one is to better understand and predict a firm’s productivity and level of x-efficiency. X-efficiency theory also suggests new areas of research, such as efforts to estimate the level of x-inefficiency due to protection and efforts to develop means to reduce the level of x-inefficiency. One specific public policy dimension of x-elficiency theory relates to the prediction of con¬ temporary economic wisdom that corporate bigness will be economically effi¬ cient since owners increase firm size to save on various costs inclusive of transaction costs (chapter 14). If, however, x-inefificiency is a byproduct of corporate bigness, due to the slacking of effort for which it allows, the predic¬ tion that corporate bigness is good for society from a cost-saving point of view loses its force. The validity of the contemporary wisdom depends on the real¬ ism of its domain assumptions. The latter remain an empirical question. The opportunity cost of adopting the conventional wisdom’s analytical predictions incorrectly relates to adopting government policies that recommend and facili¬ tate mergers and, more generally, corporate bigness when this can serve to increase the net economic costs to society. In addition, the conventional wis¬ dom precludes research in an area that might reveal economic costs of corpo¬ rate bigness that might exist and that might otherwise remain hidden from the eyes of both scholars and public policy makers (chapter 14). To the extent that the x-inefficient firm can survive even in a competitive environment, many of the fundamental predictions of the conventional wis¬ dom are called into question. In this case, Leibenstein’s domain assumption

METHODOLOGY AND SURVIVAL PRINCIPLE

61

of the persistent protection of x-inefficient firms can be dropped as a neces¬ sary condition for the existence of x-inefficiency, whereas to the extent that such protection actually exists, the case for x-inefficiency would, ceteris paribus, only be enhanced. The argument developed throughout this book, building upon the logic and empirics of x-efficiency theory and applying some of the insights of efficiency wage theory, is that the quantity and qual¬ ity of effort is positively and causally correlated with the conditions of em¬ ployment embodied in a firm’s industrial relations system or work culture. However, in contrast to Leibenstein’s modeling of the economic agent, behaving in a manner that is consistent with x-efficient production is not assumed to be the standard for rational and utility maximizing behavior from the perspective of workers, managers, or owners given the constraints they face and the arguments contained in their respective objective functions. Only under very specific industrial relations conditions would it be optimal or utility maximizing for rational economic agents to perform x-efficiently. This is es¬ pecially true when workers, managers, or owners are characterized by differ¬ ent objective functions and constraints.*® In this model, the wage rate is used, for convenience, as a proxy for a system of industrial relations where the immediate costs of labor, which in¬ corporate wages plus all other costs related to the employment of labor, are positively and causally related to the quantity and quality of effort inputted into the process of production and thereby with labor productivity. The latter is an empirically based assumption rooted in the x-efficiency and efficiency wage literature.** Moreover, to the extent that labor costs and labor produc¬ tivity, through the intermediary of changes in effort inputs, move in step with each other, x-inefficiency in production is not necessarily associated with higher unit production costs and x-efficiency with lower unit production costs. This can be deduced from Equation 2. This point is also illustrated in Figure 3.1, where an array of wage rates is associated with a unique unit production cost. In this case, more x-efficiency (increased productivity) is associated with an offsetting increase in labor costs, whereas more x-inefficiency (de¬ creased productivity) is associated with an offsetting decrease in labor costs. Where this domain assumption holds empirically, firms of varying degrees of efficiency can coexist and survive in a competitive market. Market forces could not force the relatively x-inefficient firms to become more x-efficient. In other words, the terms of the conventional version of the survival prin¬ ciple do not hold. The revised principle simply requires that firms remain cost competitive, a goal that is realizable even for the most x-inefficient firms if labor costs are relatively low enough. The survival principle, given the domain assumptions of this model, does not predict that only the efficient firms survive or, in a word, that x-inefficiency cannot exist.

62

CHAPTER 3

Figure 3.1

Unit costs

If x-inefficiency can exist even under eompetitive conditions, an array of analytical predictions generated by the conventional wisdom no longer nec¬ essarily holds. This relates to analytical predictions that are most closely tied to the anticipated eonsequences upon the economy of increasing immediate labor costs, including analyses of minimum wage laws or increases in mini¬ mum wages, unionization, and the introduction or improvement of labor or safety standards (chapters 5 and 10). The conventional wisdom’s analytical predictions are predicated upon the domain assumption that firms are oper¬ ating x-efificiently or tend toward x-efficiency in production. These predic¬ tions relate to how increases in real wages invariably and negatively impact upon labor employment and growth. The policy recommendations following upon these predictions speak against high wages and improvements in labor standards that are not a direct product of unfettered market forces. If the conventional domain assumption of no effort discretion does not hold and, moreover, effort inputs and labor costs are positively and causally correlated, the conventional wisdom’s predictions need not hold. Higher wages need not result in more supply-side induced unemployment, higher unit costs, or lower rates of return when higher wages generate commensurate increases in labor productivity.Increased labor productivity serves to compensate for any potential negative supply-side effects on the firm that might result

METHODOLOGY AND SURVIVAL PRINCIPLE

63

from higher labor costs. On the other hand, lower labor costs, achieved by reducing or eliminating minimum wages, weakening or eliminating labor unions, or eliminating government-regulated labor standards, need not yield substantive economic benefits to the firm if these actions result in firms be¬ coming increasingly x-inefficient. In effect, the firm’s labor demand curve, in this model, is not invariant to movements in labor costs and is both posi¬ tively and causally correlated with movements in these costs since move¬ ments in labor costs are expected to affect the level of x-inefficiency.'^ A necessary condition for the realization of the harmful effects of higher wages on the economy predicted by the conventional wisdom is that no ef¬ fort discretion exists. The domain assumptions related to a model that allows for effort variability yield different and more ambiguous analytical predic¬ tions than do the domain assumptions embedded in the conventional wis¬ dom. The actual impact of wage changes on the economy critically depends on the elasticity of effort inputs per unit of labor to changes in labor costs, where the measure of elasticity is an empirical question. Moreover, the re¬ formulated survival principle does not guarantee that all firms tend toward xefficiency in production and suggests that the opposite is to be expected in relatively low-wage firms. The alternative domain assmnptions open the door to analytical predictions that differ substantively from those of the conven¬ tional wisdom. In fact, the evidence is at best only ambiguously consistent with the predictions of the conventional wisdom. There is no strong evi¬ dence to suggest that minimum wages or unionization have had a negative impact upon the economy (Card and Krueger 1995; Freeman and Medoflf 1984). Only by modifying the assumptions of the conventional wisdom can the data be reconciled with theory and visa versa. Otherwise data that are inconsistent with the theory tend to be dismissed as inconclusive, incom¬ plete, or error prone. In this example, the opportunity cost of accepting incorrect or unrealistic behavioral assumptions as the foundation of one’s theory is recommending public policy that encourages the development of a low-wage economy as opposed to investigating the possibilities for a relatively high-wage economy even within the context of a relatively competitive market. More specifically, this scenario ignores the possibility of persistent x-inefficiency and the costs that this entails to society in terms of lost output and lower levels of material well-being for employees.*"^ This, of course, represents a potentially signifi¬ cant social cost if an increase in material welfare is desired by employees— if it is utility maximizing from their point of view. Related to this type of opportunity cost is that the conventional wisdom, by definition, assumes away principal—agent or larger correctable industrial relations problems inside the firm that may give rise to persistent x-inefficiency. A further problem is that

64

CHAPTER 3

firms that survive the competitive process may incorrectly be credited with being economically efficient. This precludes the analyst from objectively confronting the possibility that viable firms may be economically inefficient. Conclusion The major concern of this chapter is broadly related to the mainstream defi¬ nition of the survival principle and its implication for economic analysis. In itself the survival principle appears incontestable in the sense that firms must behave in a fashion consistent with their survival on the market. However, the assertion that efficient behavior is necessary to the survival of the firm and that therefore any surviving firm must be efficient is true only under very special circumstances: in a world where there is no effort variability and where x-inefficiency does not exist. It is this particular domain assumption that must be realistic if many of the critical predictions of the conventional wisdom are to hold true. In other words, in contradistinction with instrumen¬ talism, assumptions can be of critical importance to model building. More¬ over, if the conventional wisdom’s particular rendition of the survival principle represents a false abstraction from economic reality, important analytical pil¬ lars of the conventional wisdom are open to question. The reformulation of the survival principle presented in this chapter allows for the simultaneous existence of a whole array of firms ranging from highly x-inefficient to highly x-efficient. In this case, x-efficiency in production need not be a necessary condition for firm survival. Revising and making more realistic the domain assumptions upon which the conventional conceptualization of the survival principle is based allows for hy¬ potheses whose analytical predictions are at least as compatible with the stylized facts as are those of the conventional wisdom, and maybe even more so. There¬ fore, data that appear anomalous or even incorrect through the lenses of the conventional wisdom may appear correct from the new theoretical prescription. Although Friedman’s instrumentalist approach links economic theory to economic reality through the intermediary of testing a theory’s analytical prediction, the tale of the survival principle suggests otherwise. Conventional economic theory tends to subsume the facts. The rhetoric tends to overwhelm the reality. In the absence of a structured theory that provides an alternative analytical framework, contrary facts will be resisted and the discovery of new facts will be avoided. Explanations will be constructed and policies will be designed that are consistent with the theories that appear to work. In a sense, revisions to the contemporary wisdom serve to fulfill one of the fun¬ damental roles of economics as an objective and positive social science, as¬ signed to it by Friedman (1953a, 13—14). Friedman argues that economic

METHODOLOGY AND SURVIVAL PRINCIPLE

65

theory and the facts of economic life should be in a dialectical relationship and that amendments and revisions to economic theory should be made as one challenges both old and new data sets with theory and vice versa. Ultimately, a revamped survival principal provides the basis for revisions to public policy, with significant implications for the efficiency of the economy and thereby for the level of material welfare for society as a whole. It is the loss in economic efficiency and thereby in material welfare that is the funda¬ mental opportunity cost of adopting economic theories that are predicated upon erroneous behavioral assumptions. More specifically, the more flex¬ ible and reality-based survival principle articulated in this chapter reveals additional degrees of freedom potentially available to economic agents and public policy makers with respect to designing work arrangements and labor policy that improve the level of economic well-being of employees in a man¬ ner consistent with economic efficiency in a market-based economy. This speaks directly to the question of economic welfare and economic justice. The contemporary conventional neoclassical view of economic justice has focused on the Pareto Optimality conditions for realizing efficiency and thereby maximizing utility and welfare. The Pareto criterion for welfare im¬ provements maintains, rather non-controversially, that society should pur¬ sue an economic change that harms no one if it at least benefits one individual. This dictum holds irrespective of the distribution of income (Graff 1957). The more flexible interpretation of the survival principle presented here sug¬ gests that there exists an array of economic changes, assumed away by the conventional wisdom, that are welfare enhancing for most members of soci¬ ety and need not cause harm to the others. Such changes are, therefore, con¬ sistent with the Pareto criterion for economic change. In Figure 3.2 in an Edgeworth Box diagram two outputs, Yand X, are pro¬ duced by workers and employers, and Pareto Optimality is given by the tangency between their respective indifference curves (/Q; point b represents a Pareto Optimal position for workers and employers. In a world of x-inefficiency and effort discretion, the 1C for workers, ffj, would be thick or fuzzy (in the terminology of fuzzy logic), bounded by and for example. If one begins with x-inefficiency in production at the IC for employers, E-^, would be tangent to the x-inefficient boundary of the workers’ fuzzy IC. Changes that yield x-efficiency in production allow workers to improve their material welfare and utility without harming those of the employers as workers move to the x-efficient boundary of their IC Wj. In other words, welfare can be improved even without engaging the more controversial albeit important methods of improving the material well-being of some (Davis 1999) through income redistribution (distributive justice) or through constructing “fairer” ex¬ changes on the market (commutative justice), both of which necessarily

66

CHAPTER 3

Figure 3.2 X

Employers

cause harm to others. The conventional interpretation of the survival principle rules out such Pareto-consistent and less controversial changes since in a world of x-eflficiency such changes—^improvements in working conditions, for example—cause economic harm to others and, more specifically, threaten the survival of particular firms and economies. Notes 1. See H.A. Simon (1987) for a concise discussion of the importance of behav¬ ioral economics with a focus on bounded rationality. See Altman (1999b) for a more general discussion. The essence of behavioral economics is the view that the realism of the behavioral assumptions underlying economic theory is of fundamental analyti¬ cal importance. Therefore testing for the empirical validity of a theory’s behavioral assumptions and determining the analytical and policy-related significance of intro¬ ducing more realistic assumptions into standard theories becomes a critical excercise. 2. See, for example, McCloskey (1996, ch. 3), who argues that the current focus by economists on logic and mathematical proofs relating assumptions to predictions serves to remove economists from a concern for issues flowing from economic reality. For similar concerns, see Leontief (1971), Rosenberg (1992), and Blaug (1998). Blaug writes: “Perhaps the real trouble [with modem economics] is our age-old belief, going back to Ricardo, that economics is essentially a deductive science, in which we infer economic behavior on the basis of some assumptions about motivations and some stylized facts about prevailing institutions, suppressing even the temptation to ask whether these are descriptively realistic assumptions and accurately chosen facts. It is high time that econo¬ mists re-examine their long-standing antipathy to induction, fact-grubbing, and fact gath¬ ering before, and not after, we sit down to theorize” (1998, 30). 3. Friedman (1953a, 41) also makes the point that the validity of a theory cannot

METHODOLOGY AND SURVIVAL PRINCIPLE

67

be tested by the realism of the assumptions independent of the power of the predic¬ tions yielded by the theory. This at least suggests a two-part test for the validity of theory that involves both a reality check on the assumptions and a power check on the predictive power of the theory. However, Friedman himself does not see the reality check on the assumptions to be of any analytical significance. 4. In contrast to methodological instrumentalism, there is the institutionalist in¬ strumental approach to value theory whereby technology is an instrument to imple¬ ment values, such as social norms and rules, and the adequacy of the instrument is determined by its capacity to actualize the said values. Failure of the instrument re¬ quires that it, as well as the values, be revisited and revised (Gordon 1980, 43). See also Rutherford (1996, 66-67). 5. See Caldwell (1980) on these points. 6. Friedman (1953a, 19, n.l4) favorably references Alchian (1950) in his discus¬ sion of the survival principle. For a detailed and critical discussion of some of the implications ofthe survival principle for social evolution, see George (1989,66), who argues that inelficient processes can survive well in larger systems that are, on net, efficient. 7. See, for example, Altman (1999b), Becker (1998), Kahneman and Tversky (1979), Shiller (1993), Thaler (1992a), and Tversky and Kahneman (1981) for con¬ trasting views on the impact of different behavioral assumptions for an understanding of economic choice in a realm where the survival principle is apparently of negligible importance. 8. See references in chapter 9, for example. See also Akerlof and Yellen (1986) and Stiglitz (1987). 9. See especially Baumol (1982) on the theory of contestable markets, whereby merely the threat of credible entry into an industry suffices to make even seemingly monopolistic markets competitive. 10. See Altman (1996b, 1998) and Miller (1992) on this point. The game theoretic literature strongly suggests that under conditions of mistrust among economic agents utility-maximizing and rational economic agents will not perform x-efficiently. The appropriate environment must be fostered within the firm and sustained over time for x-efficiency in production to be approached. 11. Unlike the efficiency wage literature, I do not assume that a unique wage rate generates a unique and maximum level of labor productivity, which in turn yields a unique and minimum average cost of production (chapter 1; Akerlof and Yellen 1986). Stiglitz (1987), a leading efficiency wage scholar, however, recognizes that it is pos¬ sible for there to be an array of wage rates consistent with a unique unit cost when effort and therefore productivity change sufficiently to just compensate for changes in the wage rate. There is no evidence to suggest that the latter is not the most appropri¬ ate assumption. 12. The conventional wisdom holds that wage rates can increase without nega¬ tively impacting the economy if, ceteris paribus, labor productivity increases first. The increase in labor productivity is not a product of the wage increase in the sense of its contributing toward greater efficiency in production. In contrast, I am arguing that higher wages and overall improvements in working conditions, inclusive of the sys¬ tem of industrial relations, contribute to offsetting increases in labor productivity. Labor productivity is, thereby, at least partially endogenized. 13. These points are discussed in great detail in this book, especially in chapters 5 and 9, and in Altman (1990, 1998). Apart from affecting the level of x-inefficiency.

68

CHAPTER 3

movements in wage rates can also be expected to affect the rate of technological change, where the relatively high-wage firms are induced to develop and adopt new technology to remain cost competitive. Low wages serve as a substitute for techno¬ logical change in the sense and to the extent that low-wage firms can remain cost competitive in the absence of technological change (chapters 2 and 6). 14. Stiglerhas articulated a fundamental critique of x-inefficiency theory as delin¬ eated by Leibenstein where x-inefficiency is part and parcel of quasi-rational behav¬ ior. Stigler maintains x-inefficiency is not inefficiency at all since “In neoclassical economics, the producer is always at a production frontier, but his frontier might be above or below that of other producers.” Moreover, “The near-universal tradition in modem economic theory is to postulate a maximum possible output from given quan¬ tities of productive inputs [...] and to assert that each firm operates on this production frontier as a simple corollary of profit or utility maximization” (1976, 214, 215). Utility maximizing behavior yields efficiency in production, otherwise economic agents would be forsaking an increase in output that could be obtained at less than cost, where costs and benefits are “measured” in terms of disutility and utility. X-inefficiency is impossible by definition. Whatever economic agents do at any given time is defined as efficient. What x-efficiency theory asserts (at least in my rendition of it) is that changing the constraints faced by economic agents, such as the system of indus¬ trial relations and wage rates, yields a utility-maximizing increase in productivity. Productivity could be higher than it is under a more ideal set of constraints yielding a higher level of material welfare to society. This is not to say that increasing the level of x-efficiency is a free lunch. Agents might have to work harder, but more particu¬ larly, more smart. Also, investments in organizational capital might have to be made in the short term (Tomer 1987). Nevertheless, if an economy can be made more effi¬ cient in a utility-maximizing manner by changing the objective constraints facing the firm’s economic agents, society loses out if it is x-inefficient in production.

4

A Behavioral Theory of Economic Welfare and Economic Justice

Introduction The theoretical frame dominating the conventional wisdom’s perspective on economic welfare and economic justice is best encapsulated by the concept of Pareto Optimality (Bator 1957; Graaf 1957; Lockwood 1987) joined with Milton Friedman’s (1953a) rendition of the survival principle. In a word, it focuses on allocative efficiency, which it assumes is best achieved through market forces in a competitive market and that the invisible hand realizes a good approximation of allocative efficiency in the long run. In addition, in¬ efficient or relatively low-productivity economic organizations are assumed to be driven from the marketplace in the long run by relatively more efficient organizations. Only the efficient—more specifically, the x-efficient—firms survive.' Moreover, questions relating to the distribution of income and its potential redistribution are relegated to the back burner of economic dis¬ course as they are tied to the normative and relatively subjective and unsci¬ entific realm of economics. What counts in the conventional model is an appreciation and understanding of the conditions for achieving allocative efficiency—^and more specifically Pareto Optimality—given a particular dis¬ tribution of income. In this scenario, it is the allocatively efficient economy, assuming that all firms are x-efficient or maximizing productivity per unit of input, that maximizes the real income of society and therefore the welfare or utility that emanates from this income. A fundamental assumption for discussions on economic welfare that use Pareto Optimality as the basic reference point is that the economy is operat69

70

CHAPTER 4

ing both x-efficiently and allocatively efficiently or converging toward this point through the guidance of the invisible hand. Given x-efificiency, assumed a priori, and allocative efficiency, increasing the income of one individual can only reduce the income and related level of utility of another—^a zero sum game is assumed. Under these conditions Pareto Optimality is achieved. As long as alternative distributions of income do not affect efficiency—that is, economic efficiency is determined independently from the distribution of income—^the particular income distribution at hand is not considered to be of economic importance since it is assumed that one cannot and should not make interpersonal comparisons of the utility attached to a particular dis¬ tribution. According to this view, one cannot predict, for example, that redis¬ tributing real income from the rich to the poor generates a net gain in society’s economic welfare since it is assumed that the marginal utility of the real in¬ come accruing to the poor is no more than the marginal utility of the real income flowing to the well-to-do. This critically important assumption, how¬ ever, is not empirically based. Given this assumption, speaking to the distri¬ bution of income becomes a matter for ethical and moral choice and judgment as opposed to one based on sound economic reasoning since improving the material welfare of one individual requires a reduction in the level of mate¬ rial welfare of at least one other individual. Moreover, to the extent that the redistribution of income negatively affects the size of the economic welfare pie, the Pareto criteria for welfare maximization speak directly against in¬ come redistribution, for in this case redistribution makes the economy less efficient in terms of reducing output per unit of input from what it could feasibly be, thereby reducing the level of society’s economic welfare (Feldman 1987; Lockwood 1987; Okun 1975). Only if the winners’ marginal utility of income were sufficiently greater than the losers’ would such redistribution not reduce total economic welfare. This possibility is brushed away by the conventional wisdom through the assumption, often implicit, that the mar¬ ginal utility of real income is equal across all individuals. In our behavioral model of economic welfare maximization is framed in terms of alternative (and I would argue more realistic) behavioral assump¬ tions.^ This is particularly important since as Graaf argues, “It is clear that the interest attaching to a theory of welfare depends almost entirely upon the realism and relevance of its assumptions, factual and ethical, in a particular historical context” (1957, 3). Specifically, I examine the analytical conse¬ quences of introducing x-inefficiency into the Pareto Optimality welfare model. It is shown that if an economy is allocatively efficient and x-inefficient in competitive equilibrium, under reasonable conditions economic wel¬ fare can be increased for at least one individual without reducing the welfare of some other individual, at least in the long run. In effect, Pareto Optimality

ECONOMIC WELFARE AND JUSTICE

71

is not achieved in the long run since with x-inefficiency long-run or dynamic efficiencies remain to be realized. In this revised model competitive forces cannot induce economic efficiency to prevail even in the long run, as pre¬ dicted in Friedman’s survival principle. Inefficient economic entities can sur¬ vive in the marketplace. In this case, to the extent that economic inefficiency is a product of inefficient organization (either within the firm or in terms of the macroeconomic environment), distributional changes are consistent with the Pareto criteria for welfare maximization, and we move from the realm of a zero sum to a positive sum game. These results are also consistent with the work of Adam Smith and Arthur Cecil Pigou, where a dynamic causal rela¬ tionship is posited between the distribution of income and the level of eco¬ nomic welfare, such that an equity—efficiency trade-off need not exist. In the behavioral model of economic welfare, which builds upon the insights of Smith and Pigou and where such a trade-off also need not exist, economic welfare is maximized in a Pareto Optimal sense only under an optimal or ideal organizational or institutional setting that would be characterized by a particular level or range of incomes, inclusive of nonpecuniary benefits, tied to the different groups of individuals engaged directly or indirectly in the process of production. Redistribution of income, in a broad sense, could then be justified with respect to its positive impact on total economic welfare, such that in effect the economy is moved from a Pareto-inefficient to a Paretoefficient state over time as dynamic efficiencies are realized. Further redistri¬ butions of income, to the extent that they are efficiency neutral, would be a function of ethical or moral considerations from the Pareto perspective to the extent that they involve a reduction in the material welfare of at least one individual, assuming that the marginal utility of real income is equal across all individuals. Economic Welfare: The Conventional View The focus of the conventional wisdom is on economic welfare: that compo¬ nent of well-being that is related to the consumption of goods and services. This is not to say that noneconomic variables are unimportant, but only that the economic component is of considerable significance to most individuals. In effect, noneconomic variables are assumed to be exogenous, and analyti¬ cal questions relate to the economic variables of well-being given the state of the noneconomic variables (Graaf 1957, 5-6). Therefore, for welfare or so¬ cial utility to be maximized, output (the economic component) must be maxi¬ mized, given the available inputs. To the extent that output is not maximized in this sense, the economy is not efficient, and making the economy more efficient can increase economic welfare.

72

CHAPTER 4

The conventional wisdom focuses on the efficient allocation of resources in production given relative factor prices. The fundamental assumption is implic¬ itly made that whether or not allocative efficiency is achieved, output is being maximized per unit of input. In other words, it is assumed that x-efficiency is a given, no matter the varied organizational constraints facing firm members, either within or outside of the firm (Leibenstein 1966). The conventional wis¬ dom assumes that firm members are not endowed with some choice over how hard and how well they work. It is simply assumed that they work, a natura, at some maxinium or x-efficiently. The conventional wisdom holds that market forces invariably make firm members choose to behave x-efficiently in the long mn. For firms to survive—that is, to remain profitable and cost competi¬ tive—^firm members must behave x-efficiently. This rendition of the survival principle, clearly articulated in Friedman (1953a), remains at the core of con¬ ventional microeconomic reasoning (chapter 3; Reder 1982). In a world where effort discretion is a possibility, x-efficiency in production presumes that firm members make effort choices involving “cooperation with peers, superiors, and subordinates, in such a way as to maximize their contribution to output” (Leibenstein 1978a, 206). In other words, the work environment may result in firm members choosing to work x-efficiently (chapter 9).^ Effort discretion, a necessary condition for x-efficiency, is facilitated by the existence of incom¬ plete contracts (based on their transaction costs) and different behavioral func¬ tions, especially between employees or the agents of the firm and members of the firm hierarchy or its principals (Akerlof and Yellen 1986; Altman 1996a; Miller 1992; Stiglitz 1987). Once allocative efficiency is realized, no alternative allocation of resources would increase at least one desired output without diminishing any other desired output. Assuming x-efificiency prevails, Pareto Efficiency follows from allocative efficiency, and the potential level of economic welfare in society is maximized. If the economy is Pareto allocatively inefficient, at least one person’s economic welfare can be increased by making the economy more efficient without reducing the economic welfare of the other individu¬ als in society. This statement holds true whether or not the economy is xeflficient. Allocative efficiency is a necessary but not sufficient condition for the realization of Pareto Efficiency. For Pareto Efficiency to be effectively realized, x-efficiency in production must be present. A central question of this chapter is concerned with the significance of x-efficiency for Pareto Ef¬ ficiency and welfare-related public policy questions and whether or not the distribution of income and the organizational structure within the firm and society affect the extent of x-efficiency where x-efficiency is in turn neces¬ sary to the realization of Pareto Efficiency or Optimality. In Figure 4.1, the traditional Pareto Optimality story is illustrated in terms

ECONOMIC WELFARE AND JUSTICE

Figure 4.1 Cars

Figure 4.2

Housing

73

74

CHAPTER 4

of two products and two factor inputs. The well-known condition for Pareto Optimality is that production must take place at the point of tangency be¬ tween the production isoquants of the two outputs. Each point of tangency is Pareto Efficient or Optimal, assuming x-efficiency in production. Which point is chosen is irrelevant to the determination of Pareto Efficiency. Once a par¬ ticular point of tangency is chosen, through the intermediary of the structure of demand, any increase in the production of one output requires a reduction in the production of the other. Joining such points of tangency yields the familiar production possibility frontier given by in Figure 4.2. Once the output mix is chosen through the intermediary of the preferences of individuals and efficiently produced, output must be allocated to the indi¬ viduals in society. This allocation is considered Pareto Optimal if, given the distribution of income, there can be no further reallocation of output between individuals that would increase the economic welfare of one individual with¬ out reducing the economic welfare and related utility of the others. In a twoperson world, this condition is given by the tangency of the two individuals’ indifference curves in Figure 4.2. Only at the points of tangency are all gains from exchange exhausted—^allocative efficiency is obtained. For example, point is suboptimal in the sense that a movement from to or will increase the utility of one person without reducing the utility of the other. The utility space is derived from the production possibility frontier, which, in turn, is constmcted from Figure 4.1. Point s in Figure 4.2 is derived from one of the equilibrium points in the production space in Figure 4.1. Which point of tan¬ gency prevails in the utility space depends on the distribution of income. But all such points are deemed to be Pareto Optimal in that movement from one point of tangency to another increases the utility of one individual only at the expense of the other. The assumption is implicitly made that the redistribution of income can have no positive effect on efficiency by possibly affecting the level of x-efficiency since it is assumed a priori that x-efficiency prevails. It is also assumed that the redistribution of income cannot have a more direct effect on total economic welfare since it is assumed a priori that the marginal utility of real income for all individuals is the same. Although these assumptions do not mle out income redistribution for other reasons, the assumption-driven analytical prediction that income redistribution carmot increase total welfare has prejudiced the conventional wisdom in favor of the status quo distribution, whatever it might be (Baumol 1977, ch. 21; Graaf 1957, ch. 10). This bias is only reinforced when joined with the assumption, flowing from Friedman’s (1953a) methodological argument, that the competitive process yields the op¬ timal distribution of income and that any shift from this assumed optimal dis¬ tribution can only reduce total economic welfare, hitting especially hard those with the smallest share of the economic pie. ^2

^2

ECONOMIC WELFARE AND JUSTICE

75

Pareto Efficiency is a necessary condition for Pareto Optimality in the sense that if output is not efficiently produced, it is possible, by increasing the efficiency of production, to increase the economic welfare or utility of one individual without reducing the utility of the other. This is true irrespec¬ tive of the distribution of income. In other words, even if allocative effi¬ ciency prevails, Pareto Optimality cannot be achieved if production is not also x-efficient. A point such as can be Pareto Optimal only if production is both allocatively and x-efficient. In Figure 4.1, even if the production isoquants of the two outputs are tan¬ gent, production can still be x-inefficient if output per unit of input is not being maximized. If the economy is initially x-inefficient, increased efficiency re¬ sults in more of at least one of the two outputs being produced. If, at point m, housing output is being produced x-inefficiently, x-efficiency in production can be given by an inward shift in the production isoquant from to where the level of output produced at is identical to that produced at In this case, given more cars are now produced in equilibrium than pre¬ viously since factor inputs are released from housing production due to the move from x-inefficiency to x-efficiency in housing production. At the new equilibrium n, for example, total output is greater than at m since more cars are now produced than previously along with the same quantity of housing. Alter¬ natively, if at point m housing is now x-efficiently produced, total output in¬ creases since there are now more houses and no fewer cars produced with the same set of factor inputs. The movement from x-inefficiency to x-efficiency in production shifts the production possibility frontier outward if both outputs were initially pro¬ duced x-inefficiently. In Figure 4.2, if the production possibility frontier represents the initial state of x-inefficiency in production, represents x-efficiency in production, where more cars and housing can be produced with the same quantity of factor inputs. In this case, given the production possibility frontier allocatively efficient point is not Pareto Opti¬ mal since making production more x-efficient allows for increasing the eco¬ nomic welfare or utility of one individual without reducing the utility of the other. For example, moving from point to e, does not entail a reduction in the utility of individual b if this increase in individual a’s economic welfare, generated by an increase in real income measured in terms of cars, follows from or coincides with increasing the level of the economy’s x-efficiency. If the relatively x-efficient economy’s utility space is given by point t along the more x-efficient production possibility frontier, a movement from to e, simply involves increasing individual a’s utility by allocating to her the in¬ creased production. Individual b's utility does not suffer since the point of origin of her utility space shifts outward from 5 to r as the economy becomes

76

CHAPTER 4

more x-efficient, allowing for individual a’s utility to increase without di¬ minishing that of individual b. In effect, income is being redistributed to individual a from b, as individual a is allocated a larger share of a larger real income pie. On the other hand, if the equilibrium point remains at as the economy becomes more x-efficient, individual b's utility increases without impinging on the utility of individual a. Alternatively, with the increased level of x-efficiency in the economy, it would be possible for both individu¬ als’ utility to be increased simultaneously. In a word, as long as the economy is x-inefficient, Pareto Optimality cannot be achieved since by increasing the level of x-efficiency in production it is always possible to increase the utility of at least one individual without reducing the utility of another.

Adam Smith, the Cardinalists, and Economic Welfare It is of more than antiquarian interest to ascertain the analytical perspective of Adam Smith, in his The Wealth of Nations, with regard to economic wel¬ fare and relate this to Pareto Optimality. Adam Smith is considered to be the penultimate free marketeer, whose masterworks are viewed as benchmarks for an appreciation of the free market, for he is the dual author of both maxi¬ mal wealth and economic justice. Adam Smith argues that higher real wages, be they the product of labor power or excess demand on the labor market, not only add to the vitality of the market economy, but are also imperative from the point of view of social justice (Muller 1993, 34, 58, 60, 75). With respect to the linkage between social justice and higher real wages. Smith argues as follows: Is this improvement of the lower ranks of the people to be regarded as the advantage or as an inconveniency to the society? The answer seems at first sight abundantly plain. Servants, labourers and workmen of different kinds, make up the far greater part of every great political society. But what im¬ proves the circumstances of the greater part can never be regarded as an inconveniency to the whole. No society can surely be flourishing and happy, of which the far greater part of the members are poor and miserable. It is but equity, besides, that they who feed, cloath and lodge the whole body of the people, should have such a share of the produce of their own labour as to be themselves tolerably well fed, cloathed and lodged. (Smith 1937, 78) Redistributing real income from the well-to-do to the “servants, labourers and workmen of different kinds” is given a clear ethical basis in The Wealth of Nations. It is the economic circumstance of the majority of the population that must be improved for society to be “flourishing and happy.” Although

ECONOMIC WELFARE AND JUSTICE

77

Smith does not explicitly engage in interpersonal comparisons of utility as do the Cardinalists, well represented in the work of Pigou (1951, 1952), he implicitly assumes that total welfare improves by redistributing income from the well-to-do to the less well off. He is therefore assuming that the marginal utility real income of the latter exceeds that of the former. Smith, however, never specifies the extent to which real income should be redistributed from the well-to-do to the less well off. With respect to the microeconomic implications of higher wages, Adam Smith’s reasoning reverberates with the logic of x-efficiency and efficiency wage theory, pioneered by Harvey Leibenstein (1957, 1966) close to two hundred years following the publication of The Wealth of Nations in 1776. Smith saw a clear causal linkage between wages and the quantity and quality of effort supplied by workers to the production process. There are two key components to this process. One involves the relationship between labor compensation and effort per unit of time supplied to the production process insofar as effort supply is positively affected by the impact that the level of labor compensation has upon the nutritional intake and overall health of workers. Leibenstein (1957, ch. 8) presents evidence supporting the hypoth¬ esis that, especially with respect to workers who are at lower absolute levels of calorie consumption (a proxy for nutrition) and health, higher wages tend to increase the effort supply of labor, thereby increasing labor productivity. Reducing the wage should have the opposite effect (see also Bliss and Stem 1978; Dasgupta and Ray 1986; Stiglitz 1976). The second component re¬ lates to the manner in which the overall work environment, given the level of workers’ nutrition and health, affects the level of effort inputted into the process of production. Leibenstein (1966, 1987) hypothesizes that improve¬ ments to the work environment positively affect effort levels and thereby labor productivity (see also chapter 9; Altman 1998). There is also consider¬ able contemporary empirical work that speaks to a strong and positive rela¬ tionship between productivity and work environment which includes more active employee participation in the functioning of the firm, more labormanagement cooperation associated with a less hierarchical system of man¬ agement, labor compensation related to labor productivity, and a relatively long-term employment relationship with the firm (Alcaly 1997; Appelbaum and Batt 1994; Barney 1995; Becker and Huselid 1998; Gordon 1996; Ichniowski et al. 1996; Kochan, Katz, and McKersie 1986; Levine 1995; Levine and Tyson 1990; Neal and Tromley 1995; Pfeffer 1995). Smith argues, “The liberal reward for labour, as it encourages the propaga¬ tion, so it increases the industry of the common people. The wages of labour are the encouragement of industry, which, like every other human quality, im¬ proves in proportion to the encouragement it receives. A plentiful subsistence

78

CHAPTER 4

increases the bodily strength of the labourer, and comfortable hope of better¬ ing his condition, and of ending his days perhaps in ease and plenty, animates him to exert that strength to the utmost. Where wages are high, accordingly, we shall always find the workmen more active, diligent, and expeditious, than where they are low” (Smith 1937, 81). Increasing wages, in the first instance, represents a redistribution of incom.e in the context of a zero sum game. How¬ ever, once it is assumed that such wage increases positively affect productivity, there need be no losers, insofar as higher wages end up securing the increased output that is'their dynamic byproduct. This scenario is captured in Figure 4.2, where the increase in real income has the effect of shifting the production possibility frontier outward. In this case, doing the right thing from Adam Smith’s perspective is consistent with Pareto Optimality since improving the level of material well-being of workers is not necessarily at the expense of the owners. Such wage increases would not be expected to negatively affect the wealth of nations. However, it is not at all clear that Smith, from the point of view of equity, would not oppose real wage increases even if they resulted in a reduction in the real income of owners. Adam Smith recognized that the wage of the common laborer is a func¬ tion of the constant struggle between workers and masters, where both par¬ ties try to maximize their own particular interests. But in the natural state of the market, the master or the employer always has the advantage over labor since, in a struggle over wages, the master can hold out longer than the work¬ ers. In addition, the masters combine among each other to drive down the wage rate or to keep it from rising. These efforts are often supported, or at least not opposed, by the state. On the other hand, the state typically opposes efforts of workers to combine to raise their wages or simply to keep them from falling. In other words, even when the employer is not part of a larger organization, the free market is biased against workers with respect to the determination of the equilibrium wage. Without effective organization workers face a clear disadvantage on the labor market that can be reversed only by a thriving economy that creates a persistent or dynamic excess demand for labor. It is implicitly assumed here that increases in the demand for labor always exceed increases in the supply. Otherwise, masters will force wages down, even below what Smith considers to be the minimum sustainable level, one that allows the supply and demand for labor to be equilibrated. Smith did not believe that workers could ever produce stable combinations that would have a long-term bearing on the wage rate. As such, only a pros¬ perous and growing economy, characterized by a persistent (i.e., dynamic) excess demand for labor, could produce higher wages for workers (Smith 1937, 66—81). Only in such a prosperous economy might market forces tip the balance on the labor market in favor of workers such that they might

ECONOMIC WELFARE AND JUSTICE

79

enjoy an equitable share of the fruits of their labor. Even in a prosperous environment, however, combinations by employers could countervail mar¬ ket forces, preventing workers from securing increasing real wages. On the other hand, for Adam Smith, efforts by workers to countervail employers so as to secure a share in the wealth of nations are to be commended since the raison d’etre of the market economy is to improve the level of material well¬ being of all individuals, inclusive of “servants, labourers and workmen.” Once again, such an approach to distributive justice does not entail a zero sum game when increasing the real income of the working population serves to increase the real income of society as a whole. Closely related to the arguments articulated by Adam Smith with respect to real wage growth and the relationship between the growth in real wages and efficiency is the work of the Cardinalists, referred to by Cooter and Rappoport (1984) as the Material Welfare School (MWS). The focus of MWS is on material welfare. In point of fact, this focus does not differ from the practical orientation of the contemporary or Ordinalist School of welfare economics that has taken the form of Pareto Optimality (Graaf 1957). A critical distinction between the two approaches has most often been located in the MWS’s assumption that one can engage in interpersonal (more accu¬ rately intergroup) comparisons of utility for public policy purposes and that, moreover, one should assume, based on empirics and common sense, that the marginal utility of real income to the poor is higher than that to the rich. The Ordinalists assume that such comparisons cannot be made scientifically so that any such comparisons are normative or subjective. However, this Ordinalist assumption is not fact based. It is simply an assumption. Indeed, the Ordinalists tend toward the assumption that the marginal utility of real income is equal across all individuals. In addition, the focus of the MWS is on income redistribution and the alleviation of poverty, in contrast with the focus of Pareto Optimality on allocative efficiency given any particular dis¬ tribution of income. Of critical importance for our purposes is the case made by the MWS that increasing the real income of the less well-to-do serves to increase the efficiency of labor (Cooter and Rappoport 1984). This is in line with the case made by Adam Smith in The Wealth of Nations. The essence of the MWS’s perspective on economic welfare is clearly expressed by Pigou (1952) in The Economics of Welfare, first published in 1920. According to Pigou, his own approach to economics is scientific because it speaks directly to “what is and tends to be, not a normative science of what ought to be” (1952). And he is ultimately concerned with how to promote the welfare of society, in effect to maximize it. In this context, Pigou argues that material welfare is measurable in terms of real money income and represents only a proxy for total welfare, albeit an important one. A primary concern to

80

CHAPTER 4

Pigou is the redistribution of income from the well-to-do to the less well off insofar as this serves the objective of promoting the total welfare of society. Pigou argues that as a rule one should redistribute income to the less well-todo unless such an income transfer serves to reduce the total income pro¬ duced in society, for this would have the effect of reducing the total amount of material welfare in society. According to Pigou, such constrained income redistribution will increase the total level of economic welfare. Pigou clearly assumes a diminishing marginal utility of real income so that the more real income an individual has, on average, the lower is the marginal utility of income. Therefore, the marginal utility of real income is greater for the poor than for the rich (Pigou 1951, 299—300; 1952, 89-90; Cooler and Rappoport 1984, 513—514). Pigou writes: “It is evident that any transference of income from a relatively rich man to a relatively poor man of similar temperament, since this enables more intense wants to be satisfied at the expense of less intense wants, must increase the aggregate sum of satis¬ faction. The old law of ‘diminishing utility’ thus leads securely to the propo¬ sition: Any cause which increases the absolute share of real income in the hands of the poor, provided that it does not lead to a contraction in the size of the national dividend from any point of view, will, in general, increase eco¬ nomic welfare” (1952, 89). This conclusion is only reinforced when one considers the proportion of total satisfaction or utility yielded by compara¬ tive levels of income as opposed to absolute levels of income. This propor¬ tion is much smaller for the poor than for the rich. Pigou argues as follows: Now the part played by comparative, as distinguished from absolute, in¬ come is likely to be small for incomes that only suffice to provide the necessaries and primary comforts of life, but to be large with large in¬ comes. In other words, a large proportion of the satisfaction yielded by the incomes of rich people comes from their relative, rather than from their absolute, amount. This part of it will not be destroyed if the incomes of all rich people are diminished together. The loss of economic welfare suffered by the rich when command over resources is transferred from them to the poor will, therefore, be substantially smaller relative to the gain of eco¬ nomic welfare to the poor than a consideration of the law of diminishing utility taken by itself suggests. (1952, 90) Therefore, whether or not the aggregate level of society’s real income is increased by income redistribution, Pigou recommends income redistribu¬ tion as a means of increasing the total level of society’s economic welfare. Pigou’s case can be illustrated in Figure 4.2. Assume that the point of origin for the rich person is 5 and that the initial equilibrium is given by point e^. A movement from e^ to e, represents a clear violation of the principles of

ECONOMIC WELFARE AND JUSTICE

81

contemporary welfare theory or Pareto Optimality in that it results in the increase of one individual’s welfare at the expense of the other individual’s welfare: the poor benefits at the expense of the rich. From Pigou’s perspec¬ tive, as long as the marginal utility of real income is relatively greater for the poor individual, the movement from Cq to e, represents a net gain in total economic welfare to society, where the gain in welfare or utility to the poor exceeds the loss in utility to the rich. On the other hand, a movement from e, to Cq would violate both the conditions for Pareto Optimality and Pigou’s criteria for welfare maximization. This movement not only increases the utility of the rich by reducing that of the poor—a clear violation of Pareto Optimality—but given diminishing returns to real income and a relatively greater marginal utility of real income for the poor, it also reduces the utility of the poor by more than it increases the utility of the rich, thereby generat¬ ing a net loss of total economic welfare to society. In contrast to the MWS, the Ordinalist School argues that one cannot identify a measurable or com¬ parable value with the level of utility associated with a particular indiffer¬ ence curve, and one cannot therefore make objective inferences about the welfare consequences of redistributing income apart from what can be de¬ duced from the Pareto criteria. According to Pigou, the total level of economic welfare would eventually diminish if income redistribution had the effect of shifting the production possibility curve inward, thereby reducing the total level of material welfare available to both the rich and the poor. Nevertheless, like Smith, Pigou caus¬ ally relates income redistribution from the rich to the poor to increases in labor productivity. Pigou argues that increasing the income of the less wellto-do serves to improve their productivity by improving their capacities as workers. In effect, the quality and quantity of effort is increased as one im¬ proves the nutritional levels and health of workers. In addition, effort per unit of labor input increases as one invests in the education and skill upgrad¬ ing of labor through the transfer of income from the well-to-do to the less well-to-do. Pigou argued against the view, preeminent at the time, that an individual’s varied capacities, inclusive of the capacity to do work at various levels of intensity and quality, were predetermined biologically. He made the case for environmental variables, inclusive of economic factors, playing a fundamental role in determined effort levels, given an individual’s biologi¬ cally determined characteristics. Moreover, Pigou argues that environmental factors affecting capacities in one generation will impact on the capacities of future generations. Poor capacities, which negatively affect labor productiv¬ ity, are often positively correlated with a poor environment, and the poor environment with poverty. Alleviating poverty can therefore serve to break the cycle of poverty over the long run (Pigou 1952, ch. 10).

82

CHAPTER 4

Pigou does not consider increasing wages as the key tool in building the capacity of the poor, although he admits to the possibility that increasing wages can motivate employers to increase productivity through organiza¬ tional and technological change and that they can raise the capacity of lowwage labor by improving workers’ nutritional levels. If increasing wages do not cause increasing productivity, the total economic welfare in society will diminish (Pigou 1952, ch. 17). Pigou focuses on the redistribution of income through taxation, which in turn will be invested in the less well-to-do popu¬ lation as the main mechanism for building the capacity of workers. Pigou emphasizes transfers of income in kind to labor as opposed to an income transfer. The former precludes workers’ choosing how to spend income trans¬ fers, for if such a choice exists, Pigou maintains, many would not invest in their human capital formation. Pigou prefers that government invest directly in the education, training, and nurture of workers. He argues that workers do not have the means to optimally invest in their education. Firms do not have the incentive to optimally invest in general on-the-job training of their work¬ ers. Such direct investment would yield a rate of return, in terms of increased labor productivity, that would greatly exceed the return of investing in plant and equipment. Investing in the health and nurture of the sick will also yield high returns by improving the capacity of workers. Pigou pays special atten¬ tion to the education and nurture of children, which he considers to represent a significant investment in the future productive capacity of society (Pigou 1952, 746-749): “Here there is immense scope for profitable investment. It is just when their children are young, and, therefore, in many ways afford the most fruitful soil for investment, that poor families find themselves in the greatest straits, and, therefore, least able to provide adequately for them” (1952,750). Education has to be tied to the nourishment, housing, and medi¬ cal care of the children if it is to build up their productive capacities. A poorly fed, housed, and sickly child could not effectively learn (Pigou 1952, 751— 754). In effect, Pigou speaks to market failures with regard to the investment in human capital formation. Such failures are now an important topic for analyses in the “new endogenous growth” literature (Aghion, Caroli, and Garcia-Penalosa 1999; Osberg 1995). Overall, Pigou argues that the market does not a natura generate an x-eflficient economy in the sense of maximiz¬ ing the productive capacity of society. Real income transfers, if properly executed, serve to move the economy toward x-efficiency in production. However, transfers from the rich can have a negative effect on total output if they impinge on investment as opposed to consumption. Even in this case, the net effect of such income transfers is positive if the increased productive capacity of the poor outweighs the reduced output due to a fall in investment in plant and equipment (Pigou 1952, 742-744).

ECONOMIC WELFARE AND JUSTICE

83

Since the MWS, as expressed in the work of Pigou, maintains, as does Adam Smith, that real income transfers are typically accompanied by in¬ creasing productivity, over time the level of real income of the rich need not fall since the production possibility frontier shifts outward as a consequence of the transfers. In Figure 4.2, a movement from to e, yields no long-term loss of material welfare to individual b if the related real income transfer to individual a causes an outward shift in the production possibility curve from ^xi^xi ^XE^XEdynamic scenario, where the total level of material welfare and utility are increased to the less well-to-do without reducing that of the more well-to-do individuals, is consistent with Pareto Optimality. Also consistent with Pigou’s criteria for increasing the level of material welfare is any economic change that shifts the production possibility frontier outward such that the wealthy become wealthier and the poor’s level of material wel¬ fare remains unchanged. However, such an exogenous shift in the produc¬ tion possibility frontier would yield an even greater increase in the level of economic welfare, according to the MWS, if followed by a transfer of in¬ come to the less well-to-do. Pigouvian Optimality Adam Smith discusses at a very basic level what contemporary economists would fold under the rubric of welfare economics, making the case that an economy where workers are better off makes for not only a more just soci¬ ety, but also a wealthier society. Pigou provides a much more specific dis¬ cussion of economic welfare, and one can infer from his work what would be an optimal or best distribution of income or criteria for what may be re¬ ferred to as Pigouvian Optimality. Pareto Optimality focuses on the condi¬ tions to achieve allocative efficiency while assuming, a natum, x-efficiency in production. But Pareto Optimality cannot speak to which one of the allocatively efficient distributions is optimal; from the Pareto perspective, objectively one should be indifferent among all allocatively efficient distri¬ butions. Nor can Pareto Optimality speak to the distributional questions that relate to an economy that is x-inefficient. The fundamentals of Pigouvian Optimality are based on a different set of behavioral and institutional assumptions from those that underlie Pareto Optimality. Pigou does not explicitly concern himself with allocative effi¬ ciency, but his conditions for welfare maximization are not inconsistent with the Pareto criteria for allocative efficiency. However, as compared with the Pareto Optimality, Pigou identifies the redistribution of real income as an independent variable affecting the extent of x-inefficiency of an economy."* For Pigou, an optimal distribution of income from an x-efficiency per-

84

CHAPTER 4

spective would be one that maximizes the production-related capacities of workers. In this case, not all equilibrium points in Figure 4.2, for example, would be equally efficient. Only the distribution of income that yields x-efficiency in production would be consistent with Pareto Efficiency. It is impor¬ tant to repeat that in this dynamic scenario total economic welfare increases without necessarily reducing the welfare of the well-to-do since, ex post, the increased level of real income to the less well-to-do is financed by increasing the level of the economy’s x-efficiency. Once such a distribution of income is achieved, material welfare is maximized if and only if further redistributions of income will not yield net increases in the level of material welfare. Pigou assumes that once x-efficiency is achieved, up to some point output produced is inelastic to shifts in real income from the well-to-do to the less well-to-do. In this case, such income redistribution will not reduce incentives to invest and thereby reduce the productive capacity of an economy. Rather, the less well-to-do will benefit at the expense of a reduction in the conspicuous consumption of the rich. Under these circumstances, redistributing income from the rich to the poor increases the level of economic welfare, assuming the marginal utility of real income to the rich is less than that to the poor. Further redistributions serve only to lower the level of output, by reducing the incen¬ tives to invest on the part of the well-to-do, and thereby eventually lower the level of economic welfare. A lower level of real output need not necessarily, on the margin, reduce the level of economic welfare. This is true if the net gains in utility to the less well-to-do from redistribution exceed the drop in utility of the well-to-do. At some point, the drop in real output will be large enough to di¬ minish the level of economic welfare for society as a whole. Clearly, from Pigou’s perspective, the level of output produced in society on a per capita basis is not the sole measure of economic welfare or material welfare since a dollar of output to the less well-to-do is worth more than a dollar in the hands of the better-off folks in society. Maximizing material welfare for Pigou means maximizing not only the real output produced in an economy, but also the utility derived from that output. Maximizing both real output and economic welfare (the total utility derived from the consumption of real output) can be achieved only through an appropriate distribution of income since the levels of real output and economic welfare are functions of the distribution of income, ceteris paribus. Technically, this distribution is achieved when the marginal utility of real income is equal across all individuals A Behavioral Theory of Welfare Optimization A behavioral model of welfare optimization and material welfare builds upon the Smithian and Pigouvian perspectives on material welfare. I will refer to

ECONOMIC WELFARE AND JUSTICE

85

this behavioral model as Smithian in recognition of its historical roots and the all-encompassing nature of Smith’s arguments. Of particular importance is that Smith and Pigou speak to what is today referred to as dynamic effi¬ ciencies, which derive from redistributing income to the less well-to-do mem¬ bers of society. A key assumption of the behavioral theory of welfare optimization is that market forces, even in highly competitive markets, need not generate firms that are both allocatively and x-efficient. Even if allocative efficiency is generated, this need not go hand in hand with x-efficiency in production. Moreover, it is assumed that income redistribution, in line with the reasoning of Smith and Pigou, positively affects productivity. In other words, market forces do not necessarily yield a unique distribution of in¬ come that is causally correlated with allocative and x-efficiency. If market forces do not yield such unique and efficient results, it must be tme that xinefficient firms can survive in a competitive market. Neither Smith nor Pigou discusses the market imperative rooted in Friedman’s survival principle in their articulation of the relationship between redistribution and the produc¬ tive capacity of individuals. However, in the behavioral theory of welfare optimization, where it is assumed that x-inefficient firms can survive even in the long run, one must reconcile this assumption with Friedman’s survival principle, which remains the mainstay of contemporary theory. Moreover, one must reconcile the assumption of the long-run survival of x-inefficient firms with a proposition that is derived from Friedman’s survival principle: the Axiom of Modest Greed (McCloskey 1990). This axiom stipulates that utility-maximizing and rational economic agents exhaust all reasonable static and dynamic gains from trade. Why would firms remain x-inefficient if rea¬ sonably large gains are to be had by becoming x-inefficient? Underlying a behavioral theory of welfare optimization that in turn is predi¬ cated upon the existence of x-inefficiency is the assumption that in a world of effort discretion economic agents, and thereby firms, will not perform xefficiently unless the appropriate organizational parameters are in place. As already discussed with regard to the issues raised by Smith and Pigou, the level of x-efficiency can be positively affected by the conditions of work, inclusive of the wage rate, and also, more specifically, by the standard of living of low-income workers. In other words, in the rhetoric of Pigou, as working conditions improve, they will positively affect the productive ca¬ pacity of currently employed workers, as well as that of future generations of workers. In addition, according to Pigou, investing in human capital such as education and on-the-job training should have the same effect. The latter could result from higher labor income. Focusing on the relationship between effort discretion and productivity, Leibenstein (1966, 1978, 1979) argues that only under monopoly or mo-

86

CHAPTER 4

nopolistic market structures or under the protective umbrella of tariffs or subsidies can x-inefificiency in production prevail. This statement follows from his assumption that an x-efficient firm produces at a lower unit cost than does an x-inefficient firm. X-efficiency is achieved by workers, manag¬ ers, and owners working harder and better (chapters 1 and 2). It is assumed that such an improvement in the quantity and quality of effort is obtained at no economic cost. Relatively high-cost producers cannot survive in a com¬ petitive market. Such x-inefficient firms either become x-efficient or perish. Such an analytical prediction is consistent with Friedman’s survival prin¬ ciple except that Leibenstein assumes that inefficient firms can be protected from competitive market forces even over the long run. Friedman and most conventional economist types would not agree, arguing that in the long run market forces force firms into becoming efficient or force the inefficient firms into bankruptcy (chapter 3; Baumol 1982; Reder 1982). Does this then imply that x-inefficiency can exist only under exceptional circumstances, or is it possible for x-inefficient firms to survive even under relatively stringent competitive conditions? The Axiom of Modest Greed predicts that all firms will be x-efficient even under monopolistic and other forms of imperfect product markets, so that such market structures yield only allocative inefficiencies. Leibenstein points to nonallocative inefficiencies that are byproducts of monopolistic and otherwise protected markets, where effort is a discretionary variable, and when economic agents choose not to perform x-efficiently. In Figure 4.2, not only is the economy at 62, consistent with the conventional wisdom’s analytical prediction of allocative ineffi¬ ciency under a regime of imperfect product markets, but it is also at an infe¬ rior production possibility frontier, representing x-inefficiency in production. This assumes that economic agents are utility maximizers as op¬ posed to profit maximizers cum output-per-unit-of-input maximizers and that they trade off higher potential profits against the effort required to become xefficient (Leibenstein 1978). In this “imperfect” product market scenario the level of material welfare is less than it would be in a more competitive envi¬ ronment. Leibenstein assumes that economic agents are quasi-rational in the sense that they choose to forgo opportunities to lower costs—^that is, to be¬ come x-efficient—^when competition in the product market is not stringent. In effect, they choose to pass over $500 bills on the street (McCloskey 1990, 112) in violation of the Axiom of Modest Greed because it is not worth the effort in the absence of a significant measure of market pressure. In the behavioral model of the firm discussed in this book (see especially chapters 1 and 2; Altman 1998), the calculating, forward-looking individuals or rational economic agents need not choose to perform x-efficiently in a highly competitive market, and such x-inefficient behavior is not in violation of the

ECONOMIC WELFARE AND JUSTICE

87

Ajciom of Modest Greed. Assume that all economic agents are rational utility maximizers, with different objective functions for agents and principals in face of different objective constraints. Also assume a competitive product market such that principals are subject to the binding constraint that unit costs must be competitive—relatively high-cost producers are not protected by a monopo¬ listic market structure, tariffs, or subsidies. Output is a product of inputs such as capital, labor, land, and technology, as well as the work culture or the indus¬ trial relations system in the firm. Ceteris paribus, a more effective work cul¬ ture generates a higher level and quality of output (chapter 9; Alcaly 1997; Appelbaum and Batt 1994; Barney 1995; Becker and Huselid 1998; Gordon 1996; Ichniowski et al. 1996; Kochan, Katz, and McKersie 1986; Levine 1995; Levine and Tyson 1990; Neal and Tromley 1995; Pfeffer 1995). This analyti¬ cal prediction also relates to Leibenstein’s (1978, 206) argument that x-efificiency requires a cooperative work culture. Based on the stylized facts of the firm, also assume that an x-efficient work culture requires that the firm invest in organizational capital (Tomer 1987). This includes improved screening and training of employees and managers and redesigning and reconfiguring the shop floor for a more participatory work environment. Members of the firm hierarchy would also have to work harder and better. They would at least have to reinvent their approach to human resource management. Moreover, not all layers of the firm hierarchy benefit from the superior, more cooperative work culture. For example, the ratio of management to employees tends to be much lower in the superior work culture, thereby threatening the position of many middle managers. In addition, principals may witness a loss of power and prestige, to which they attach utility and a heavy weight in their objective function. In this sense, achieving x-efficiency is not a free ride. It clearly in¬ volves opportunity costs (both material and nonmaterial) that differ across the spectrum of the firm’s economic agents (chapter 9; Alcaly 1997; Appelbaum and Batt 1994; Gordon 1996; Ichniowski et al. 1996; Klein 1984; Levine and Tyson 1990). Of fundamental importance to the behavioral theory of the firm is the analytical prediction that, ceteris paribus, in the long run, by changing the work culture of the firm, factor inputs become more productive, shifting the production possibility frontier outward. The consequence of modeling x-eflficiency as a function of investing in an xefficient work culture is that shifting to a more x-efficient system of produc¬ tion affects both productivity and costs, with implications for Friedman’s survival principle and the Axiom of Modest Greed. This can be illustrated by comparing an x-efficient to an x-inefficient firm; for simplicity, one assumes that the only difference between the two types of firms is the application of different work cultures, where the latter incorporates the different rates of labor compensation and different stocks of organizational capital.

88

CHAPTER 4

Average costs are given by the following basic equation assuming, for simplicity, that labor and organizational capital are the only inputs into the production function:

AC =

OCi + wL

(4.1)

Q where y4C is average costs; is the value of organizational capital per unit of labor input; w is rate of labor compensation; L is labor input; and Q is output. This equation can be rewritten as:

AC =

qc,

w (4.2)

Q L

L

Average costs are determined by the per unit organizational capital costs plus the per unit labor costs, which are expressed as a value of organizational capital per unit of labor input deflated by labor productivity and the wage rate deflated by labor productivity, respectively. Assuming that OC^ and w increase with the adoption of more x-efficient work cultures, it is clear from Equation 4.2 that the higher productivity x-efficient firm might be character¬ ized by the same average costs as the lower productivity x-inefficient firm if productivity and corresponding costs rise in an offsetting fashion. It would then be possible for the x-inefficient firm to compete on the basis of low rates of labor compensation and a smaller investment in organizational capi¬ tal and the x-efficient high-wage firm to compete on the basis of higher lev¬ els of productivity. Under these circumstances there would be no economic reason for there not to exist in a competitive environment a multiplicity of firms producing at different levels of x-efficiency. This assumes that there is no unique wage rate or work culture that minimizes unit production costs, at least over a significant range of wage rates or work cultures. In other words, there is some linearity in the relationship among work cultures, effort, and productivity (Stiglitz 1987). This point is illustrated in Figure 4.3, where a unique average cost, C*, is associated with an array of work cultures, which, in turn, incorporates an array of wage rates to 0IT*. In contrast, in a world where x-efficiency is given a nature, any wage increase is accompanied by an increase in average cost, ceteris paribus. Given technology, after the wage rate or the overall cost of the work culture reaches a particular level (OW*), further increases in wages yield higher unit costs as productivity improve¬ ments cum increases in the level of x-efficiency can no longer keep pace

ECONOMIC WELFARE AND JUSTICE

89

Figure 4.3

with the wage increases. In this case, the lower-wage firms would have a competitive advantage over the high-wage firms. On the other hand, only when the productivity differentials, controlled for the quality of output, in favor of the relatively more x-efficient firms exceed the cost differentials favoring the x-inefficient firms, can one expect the competitive process to result in the dominance of the high-productivity, relatively x-efficient firms. To the extent that technological change is induced by changes in the cost of labor, an additional degree of freedom is introduced to the capacity of highwage firms to survive in a competitive market by increasing the firms’ pro¬ ductivity to compensate for the increasing cost of labor and, by extension, increasing society’s level of material welfare (chapters 2 and 8; Altman 1998).^ The behavioral model of the firm demonstrates that in a world with xinefficiency, where a unique average cost corresponds to an array of work cultures, both x-inefficient and x-efficient firms can coexist in competitive equilibrium. In this case, Friedman’s survival principle does not hold, and there exists no market imperative forcing economic agents to perform xefficiently. In a world where x-inefficiency is the rule, the survival principle must be revised as follows: “To survive in the marketplace an economic en¬ tity must remain cost competitive.” Moreover, where both x-efficient and xinefficient firms are cost competitive, there is no market incentive for firm managers and owners (principals) to build a more x-efficient firm. This is

90

CHAPTER 4

particularly true when workers, through higher wages and improved condi¬ tions of work, capture the increased productivity generated by a higher level of x-efficiency. In this case, the persistence of x-inefficiency in a competitive market would be consistent with the Axiom of Modest Greed since the prin¬ cipals, the decision makers of the firm, would not be passing over free $500 bills. This cash on the street would cost $500 to pick up. The beneficiaries of the cash would be the agents or workers. Only if one of the arguments in the principals’ utility function is the utility of the agents, whose utility is posi¬ tively affected by higher levels of x-efficiency, would principals be expected to proactively work toward the realization of more x-efficiency in produc¬ tion. Otherwise, the utility of the majority of firm members—^Adam Smith’s “servants, labourers and workmen of different kinds”—^remains an external¬ ity to the principals. Market forces, even highly competitive market forces, need not cause this important externality to be internalized by the principals, when x-efficient firms remain cost competitive with x-inefificient firms. This behavioral model of the firm (BMP) can serve as the theoretical ba¬ sis for a behavioral or Smithian theory of welfare optimization. The key distinction between the behavioral theory of welfare optimization and the conventional one is a different set of behavioral assumptions, which yield a different set of analytical predictions, in particular that x-inefficiency is a stylized fact and that x-inefficiency can be reduced by improving the work conditions of labor and, more generally, by improving the work culture of the firm. The behavioral model shows that given its underlying assumptions, changes in the work culture that represent an actual redeployment of re¬ sources to workers are consistent with the survival of the firm in a competi¬ tive market. Moreover, there need not be a unique distribution of income or resources within the firm that is consistent with its survival since an array of distributions is consistent with equilibrium in a world where x-inefficiency exists. In other words, in this scenario, income redistribution need not result in reducing the competitiveness of a firm as predicted by Friedman’s sur¬ vival principle. However, there might be a particular or unique distribution of resources that is consistent with x-efficiency in production. And this dis¬ tribution would be related to the work culture that allows and facilitates the realization of x-efficiency in the firm. In addition, the realization of this unique distribution need not entail a reduction in the level of material well-being among any firm members since this distribution is both purchased by and a cause of an increase in the productivity of the firm. Finally, the BMF de¬ tailed above demonstrates that the distribution of income consistent with xefficiency in production need not be pursued by rational economic agents when the decision makers of the firm do not benefit from improvements in efficiency and when x-inefficiency in production is consistent with the long-

ECONOMIC WELFARE AND JUSTICE

91

run survival of the firm. Therefore, an economy’s failure to move toward its x-efficient production possibility frontier as its ultimate competitive equilib¬ rium is consistent with the Axiom of Modest Greed and with the survival principle as modified for a world of x-inefficiency. In a world where x-inefficiency prevails, an economy’s level of material welfare can increase by improving upon the nutritional and health levels of the poor—that is, their physically related productive capacity—^as is argued by Smith and Pigou and later by Leibenstein. Such improvements generate dynamic efficiencies such that the level of material well-being of the rela¬ tively well-to-do need not diminish. Moreover, the competitiveness of af¬ fected firms need not deteriorate so long as productivity increases neutralize any increase in the costs of labor. This argument also applies where improve¬ ments to the productive capacity of labor are a product of income-tax-based transfers, as recommended by Pigou. Similarly, improvements in the work culture of the firm, which typically serve to improve the level of material well-being of the workers, also generate dynamic efficiencies such that the level of material well-being of the relatively well-to-do need not diminish and the competitiveness of the firm need not deteriorate. In other words, given the existence of x-inefficiency and the recognition of unrealized dy¬ namic efficiencies in the economy, it is possible to make at least some people better off without making some other people worse off In this case, income redistribution does not violate a basic principle of Pareto Optimality, at least in a dynamic sense. Moreover, it is clear that under conditions of x-ineffi¬ ciency there is an additional degree of freedom available to redistribute in¬ come away from the relatively well-to-do without negatively impacting upon the competitiveness of firms and, by extension, economies. And redistribu¬ tion here translates not into taking from one group to give to another, but rather into increasing the share of one group relative to another by increasing the level of output. In this fashion, the aggregate level of society’s material welfare is increased if those who are receiving an increase in their real in¬ come attach a positive utility to this increase. This is consistent with the conventional wisdom, which implicitly assumes that all individuals attach the same marginal utility to a change in real income. Society’s material wel¬ fare is maximized, under these conditions, when any marginal redistribution has the effect of reducing the size of the economic pie. This occurs when the economy becomes x-efficient. This scenario is similar to what was discussed with respect to Pigouvian Optimality except now his and also Adam Smith’s analyses are situated within a behavioral theoretical framework and are shown to be consistent with the Axiom of Modest Greed and a revised version of the survival principle. Figure 4.4 illustrates the connection between income redistribution and eco-

92

CHAPTER 4

nomic welfare. As with the Pigouvian or Smithian world, redistributing income from the well-to-do to the less well-to-do increases real income by reducing the extent of x-inefficiency. Increases in real income yield increases in the material or economic welfare in society. The extent of this increase depends upon whether or not one assumes that the marginal utility of real income is greater for the less well-to-do than for the well-to-do. If the marginal utility of real income is equal across all individuals in society, the increases in economic welfare equal the increases in real income. In this diagram, the increases in welfare are depicted as greater than'increases in real income since it is assumed here, in accordance with Pigou, that the marginal utility of real income is greater for the less well-to-do individuals. These increases in welfare might be exaggerated to the extent that such a differential in the marginal utility of real income is not valid. Whether it is or not must be an empirical question, and one should not simply accept as a given the conventional wisdom’s implicit assumption that the marginal utility of real income is equal across all individuals or groups of individuals. In Panel I of Figure 4.4 the relationship between the distribution of income and real output is mapped, ceteris paribus. One begins with a skewed distribu¬ tion of income at that yields an x-inefficient level of real output at Q^. This, in turn, yields in Panel II, where the relationship between output and welfare is mapped, a level of material welfare to society of W^. In Panel IV the income distribution-welfare relationship is drawn. This relationship is derived from Panels I and II. Increasing the share of income going to the poor to yields increas¬ ing levels of real output up to Qxeq^ or the x-efficient level of output. This is associated with an increase in welfare to in Panels II and IV. The increase in welfare is greater than the increase in real output given the assumption that the marginal utility of real income is greater to the poor. In Panel I, further shifts of income to the poor from to have no effect on output; output is per¬ fectly inelastic at This follows from the behavioral assumptions of Pigou. However, in Panel IV, there is a positive slope to the distribution-welfare curve since the redistribution of income to the less well-to-do has the effect of increas¬ ing the net level of material welfare given the assumption of a differential mar¬ ginal utility of real income in favor of the poor. The level of economic welfare increases to given Pigou’s behavioral assumptions. In Panel II, this rela¬ tionship is expressed as a particular level of output being associated with an array of levels of welfare that, in turn, are a product of different income distribu¬ tions. In Panel I, further income redistributions to the poor, such as D^* and D^**, negatively affect the level of output, eventually reducing the level of eco¬ nomic welfare. Initially, this fall in output need not yield a drop in welfare due to the compensating effect that the redistribution might have upon the aggregate level of economic welfare through its positive effect on the welfare of the less well-to-do, (Panels II and IV). Eventually further redistribution can be expected

ECONOMIC WELFARE AND JUSTICE

Figure 4.4

93

Income Distribution and Material Welfare Q

to reduce the level of society’s economic welfare. However, redistributions that reduce the level of real output by reducing the efficiency of the economy are not sustainable if they are not consistent with the survival of the firm, even if they might otherwise positively impact upon the aggregate level of economic wel¬ fare. In this diagram, economic welfare is maximized at either or depending upon the realism of Pigou’s behavioral assumption about the differ¬ ential marginal utility of real income between the less well-to-do and the well-todo. The optimal level of material welfare is here directly tied to the distribution of income, which contributes to the realization of the maximum level of eco¬ nomic welfare. Conclusion A behavioral or Smithian theory of material welfare allows for different and more realistic behavioral assumptions than does conventional Pareto

94

CHAPTER 4

Optimality. It does not contravene the well accepted insight in Pareto Optimality that allocative efficiency is a necessary condition for maximizing economic welfare. In addition, it does not necessarily recommend policy that would increase the economic welfare of one individual or group at the expense of another. Rather, the behavioral theory suggests that the conventional theory of economic welfare makes assumptions about economic or x-efficiency and the relationship between income distribution and the level of real output that are unreasonably unrealistic, yielding public policy recommendations that are unreasonably biased against income redistribution. The different perspectives on economic welfare are outlined and illus¬ trated in Table 4.1 and Figure 4.5. In Figure 4.5, the real incomes of two individuals are mapped out such that it is theoretically possible for each indi¬ vidual to control and consume total income, and this total income is the same for each individual. The three income curves, 1,2, and 3, indicate higher levels of real income, and the slope of each real income curve is one. A simple Pareto-consistent world is one where income redistribution results in the loss of income for at least one individual. This is illustrated by a move¬ ment from C to D in Figure 4.5, wherein individual A gains FA income while individual B loses KL income. The Chicago approach to Pareto Optimality suggests that income redistribution has a dynamic effect on real output, causing the level of real output to fall over time. In this case, there is a shift in the income curve from 2 to 1, to point V. Not only does income redistribution involve a zero sum game, but it also has a negative effect on total output. A world that is consistent with the behavioral model of economic welfare is one where income redistribution has a dynamic and positive effect on real output through its positive impaet on the efficiency of the economy in a world where x-inefFiciency exists. Redistributing income from point CioD yields an outward shift in the income curve to 3, such that the economy might end up at point E, wherein individual B does not suffer a loss of income as a eonsequence of income redistribution. This is in stark contrast with the Chi¬ cago worldview, where income redistribution is a negative zero sum game in a world where x-inefficiency does not and eannot exist, as determined by market forces and the Axiom of Modest Greed. The focus on redistributional issues in the Smithian or behavioral theory of material welfare is most closely related to income redistribution as medi¬ ated through changes in the work culture of the firm, as well as on in-kind transfers such as education, health services, and housing, which most di¬ rectly impact upon the level of material (and psychological) well-being of workers. Such transfers, it is argued, serve to increase the productivity of labor. To the extent that the underlying behavioral assumptions of the Smithian theory are empirically valid, income redistribution is self-financing up to

ECONOMIC WELFARE AND JUSTICE

95

Table 4.1

Taxonomy of Potential Welfare Choices Model type

Effect

Figure 4.5 example

Pareto-consistent

Zero sum

C to D

Pareto-consistent Chicago approach Time dependent Behavioral-consistent (Smilh-Pigou) Time dependent

Zero sum negative

C to D to V

Win-win positive

C to D to E

some point and therefore does not involve a zero sum game, as it must in the welfare world of Pareto Optimality. In the Smithian theory, income redistri¬ bution is good for growth and is not inconsistent with the basic conditions of Pareto Optimality in a world where x-ineflficiency exists. In this scenario, income redistribution is not imbued with the largely negative subtext that flows from the conventional modeling of economic welfare. This has impor¬ tant implications for public policy. If income redistribution can be shown to be good for growth, as well as welfare-improving for the vast majority of a population, with few or no negative welfare implications for the well-to-do, the case against income redistribution policies loses its theoretical and em¬ pirical bearing. The focus of debate should then shift to the determination of the types of redistributions that are most productive and welfare-enhancing and the extent to which redistributions are economically viable.^ At a more general level, it is important to test the proposition that a substantial or eco¬ nomically significant equity—efficiency trade-off exists. The empirical literature on income distribution does not directly speak to the issues raised in the economic welfare literature. Nor can analyses based upon macroeconomic indicators provide definitive proof supporting or re¬ futing one of the important analytical predictions underlying the behavioral or Smithian theory—that is, that an equity-efficiency trade-off is at best a weak one so that increasing income equality need not be purchased with less efficiency. Nevertheless, this literature casts some empirical light on this important theoretical question. Of course, the conventional wisdom hypoth¬ esizes that a substantial trade-off between equity and efficiency is the mle. There is a wide array of macro-oriented empirical work relating economic growth and the level of real per capita output or gross national product (GNP) to income inequality. For these analyses to support the conventional wisdom s view of the equity-efficiency trade-off, one should find that per capita GNP is strongly and positively correlated with income inequality. The higher the

96

CHAPTER 4

Figure 4.5 A

B

level of per capita GNP, the higher should be the level of income inequality. In this case, the level of per capita GNP is an indirect proxy for efficiency. Much of the empirical literature is in response to Kuznets’s (1955) pioneer¬ ing -work, where he argued for an inverse-U relationship between per capita GNP and income inequality. In this case, income inequality increases up to some maximum as an economy’s per capita GNP rises. Only as the economy becomes increasingly wealthy does income inequality diminish. One would expect to find no equity-efficiency trade-off only at very high levels of per capita output. However, the evidence supporting Kuznets’s hypothesis is mixed at best. Todaro summarizes this literature: “Few development economists would argue that the Kuznets process of increasing then declining inequality is inevi¬ table. There are now enough case studies and specific examples of countries

ECONOMIC WELFARE AND JUSTICE

97

like Taiwan, South Korea, China, Costa Rica, Sri Lanka, Hong Kong, and others to demonstrate that higher income levels can be accompanied by fall¬ ing and not rising inequality. It all depends ... on the nature and character of the development process. Those who argue for the inevitability of the Kuznets process—especially those political leaders in countries with large and grow¬ ing inequalities—more often than not are simply searching for a convenient smokescreen behind which to mask their goals of economic aggrandizement or cover policy failures” (1989, 165—166). Moreover, in a comprehensive survey of the most recent empirical literature, Aghion, Caroli, and GarciaPenalosa find that “The picture they draw [recent studies] is impressively unambiguous, since they all suggest that greater inequality reduces the rate of growth.” They add: “The traditional view in economic theory, then, is that there is a fundamental trade-offhQtsNQQn productive efficiency (and/or growth) and social justice.... Overall, the view that inequality is necessary for accu¬ mulation and that redistribution harms growth is at odds with the empirical evidence” (1999,1617,1620). ^ In fact, there does not appear to be a system¬ atic equity-efficiency trade-off. These observations find further substantiation in a simple cross-sectional analysis of countries for which the World Bank has both real per capita GNP estimates and income inequality estimates. Across the entire real income range for 1998 there are seventy-nine such countries that also provide esti¬ mates for at least one year in the 1990s for income inequality as measured by the income share of the highest 20 percent relative to that of the lowest 20 percent. The GNP for all nations is provided in terms of purchasing-power parity equivalencies (Table 4.2; Figure 4.6). In Table 4.2 countries are ranked by their level of per capita income, with the highest income country ranked number one. Data are also provided on income inequality, with the most egalitarian country ranked number one, and on per capita GNP. In Figure 4.6 the relationship between income inequality and per capita GNP is plotted. These data clearly reveal, albeit only cross-sectionally, that there exists a negative, nonlinear relationship between increasing per capita GNP and re¬ ducing the level of income inequality. In addition, the Spearman rank corre¬ lation coefficient, which measures the linear correlation between the per capita GNP and the income inequality ranking, is 0.24. Increasing the GNP per capita ranking tends to positively correlate with increasing the income in¬ equality ranking, improving a country’s level of income equality. Generally speaking, countries with the highest levels of per capita GDP are those with relatively low levels of income inequality. No high-income economy is characterized by the higher levels of income inequality that characterize many of the low GDP per capita economies. On the other hand, low per capita GDP is associated with a wide range of income inequalities. These estimates

98

c 0 CC

fO nJ

Q.

Z O

Q.

CO

cj)0'>-c\jco'>4-incDr^co(T>oi-c\j Ti-mLnininLotninminiocDcDCD

0 0 0 0 0 0 CO CM LO CM (J3 CO 00 CM

0 0 0 0 0 0 0 0 0 0 0 0 0 0 CD

if ^ o o "2 O CO

c

■55 2 CO

CO ly ^ o > w p

C2. O

ro CD) O

Q. (T3 k0

a.

2 O

O Q.

Income Distribution and GNP per Person

6) “cB c 0 *0 CC 0

O "g

Q.

C/) CC 5 0 k. 3 ■D o)-S C 0 X

0 3

0

0 _>

0 E O) 0 0 0 c 0 0 C 0 c 0 ’3 0 0 0 z 0 0 CO c > O) 0 0

'c 0

c

0 S C W 0

O CL

lO CD

I^COCDOf-CMCO-^tncDt^COCDO

OOOOOO CM 03 IT) CO CD CO^ CD^ CM O 00^ Tf ctT cd^ rt co~ co’* CM CM CM CM CM CM

oooooooooooooo TfCMCMi-rJ-r^OCOOOTt-.-COCncD) i^_ co_ CD_ ix)_ CD irM_ CM_ T- Tt co_ co_ o_ cq tcm" cm" T-" o' o' CD o' o' Oi CO |C co' cm' c6

O 03 CM O m CD m CM CM

i-cor^oincocx)coTt-»tcMT-CT>r^ CMt-t-CM -^cm cococmcdco

0

C\J CO

CC

>» ■D g 0 Q

1!Q.

^

-a 3

CO CD

n c > o " ?< CC3 > -

C

0

0

a. (D (D

5

CO CD

O

c 0

C

o

CC

3

CMCMCMCMCMCMCMCM1-1-1-T-1-

s«

« 9^ ^ 0

q3 § c

Q.^ (D

E o >< o c\j -Q CD o c 4- "O W (D o 0^ -g C\J •§)> ¥ "O

Tj- CT> IT) CM CD CD C73 iri CO LO CO CO

CMcqcDT-lCDCDi-^CJDCDTtCMTfTl-CM coiniriTfiricouoiricocbcDirit^cD E

(/)

C/D

0 CO?

C/) -2 _

>^ 0 T3 0 0 "D 0 N > 0 .ti C C C $ O 0

E

g ig)

.2

CD

3

3 CO Z O Q CQ

w


.0 tsl t >■ 05 o «

FT3 5 C C

g

~ nr

C0r= CO :=

CO

JO

CO 1= CO CO O -Q o _ ‘c E 5'«c=’-n™ CO 05 CO-;;:.^! Ci^'C « W i ^ - COW

—, ^

5o^ocooo2o§oDCo > O ^ (/) Q. 0. O CO I-I-Q

c C/) cc 4—* Sr? cc c CO

_

2^ QCC :=

’C Q

Ig

—'w—»C0-CC^O^ ^

(DCLCCQ-CL^^a LU < _i CD

3 Co

Figure 4.6

Real per Capita GNP and Income Inequality In the 1990s 100

ECONOMIC WELFARE AND JUSTICE

101

and data speak against the conventional worldview that more income equality must be purchased by less efficiency. Indeed, highest levels of per capita GNP are strongly and positively correlated with low levels of income inequality, which is consistent with the Smithian theory of economic welfare. The Ameri¬ can exception is an excellent illustration of this point. The U.S. achievement of the world’s highest per capita GNP, at over $29,000, is linked with an income inequality level of over 9, which earns the United States a ranking of 50 in terms of income equality. Other nations with such a high level of income in¬ equality include China, Thailand, Ecuador, Burkina Faso, and Kenya, which produced per capita GNP of between $1,000 and $6,000 in 1998. A relatively high level of income inequality is clearly not a sufficient condition for achiev¬ ing a high level of per capita output. On the other hand, Switzerland realized an income inequality of 5.9 in conjunction with a per capita GNP of almost $27,000, and Norway’s $24,000 was matched with a level of income inequal¬ ity of only 3.5. Apparently, a high level of income inequality is not a necessary condition for realizing a high level of per capita GNP. Overall, these and other highly aggregate estimates do not suggest an equity-efficiency trade-off. Moreover, they suggest that more equity in the distribution of income is correlated with a higher level of per capita GNP. This is consistent with the self-financing nature of increasing income equal¬ ity predicted by the Smithian theory of economic welfare. If anything, these estimates speak to the potentially positive causal role that increasing income equality plays in per capita GNP growth. This is consistent with the worldview underlying the Smithian theory that changes in the distribution of income are not simply a matter for ethical consideration but have objective positive eco¬ nomic effects on the economy. Finally, the aggregate estimates that pertain to the relationship between income inequality and per capita output suggest that society has much greater scope over the distribution of income than is predicted by the conventional wisdom since it appears that increasing in¬ come equality does not negatively affect the level of per capita output.

Notes 1. Leibenstein (1966) points out that microeconomic theory is largely concerned with allocative inefficiency, which in reality is of a trivial dimension, to the exclusion of other types of inefficiencies, which Leibenstein broadly groups under the province of x-inefficiency. In fact, the conventional wisdom assumes that x-inefficiency cannot exist (Altman 1999b). 2. Herbert Simon writes that behavioral economics is “a commitment to empirical testing of the neoclassical assumptions of human behavior and to modifying eco¬ nomic theory on the basis of what is found in the testing process.” In addition, “be¬ havioral economics is concerned with drawing out the implications, for the operation

102

CHAPTER 4

of the economic system and its institutions and for public policy, of departures of actual behavior from the neoclassical assumptions” (Simon 1987, 221). 3. See Frantz (1997) for empirical evidence on the existence of x-inefificiency. 4. See Thorbecke (1990) on the question of institutional x-inefficiency. See also North (1990) for a discussion of the raison d’etre of the persistence of inefficient institutions based upon a transaction cost analysis. 5. In this scenario, technical change is induced by changes in the cost of labor. To remain cost competitive firms must develop or adopt technologies that serve to reduce unit costs to competitive levels. Such technical change results in an inward shift of the firms’ production isoquant and an outward shift of the economies production possibil¬ ity frontier. 6. See Bowles and Gintis (1998) and Gordon (1996, 1998) for a detailed discus¬ sion on the significance of different types of distribution for productivity. Bowles and Gintis focus on the redistribution of assets as the means for enhancing productivity and redistributing income, whereas Gordon focuses on reorganizing the work culture of the firm. 7. Aghion, Caroli, and Garcia-Penalosa (1999) also discuss some underlying theo¬ retical conditions under which more equity yields more real per capita output. They focus on the impact that imperfect capital markets and limited liability have upon suboptimal investments in the economy, inclusive of human capital. In such a world, increasing income equality yields higher levels of per capita output.

5

The Economics of Exogenous Increases in Wage Rates in a Behavioral/X-Efficiency Model of the Firm

Introduction Standard neoclassical theory predicts that higher real wage rates result in less employment and a lower rate of employment growth than would otherwise exist with lower real wage rates.' Thus exogenous interventions in the labor market that serve to increase real wage rates, such as effective minimum wage legislation and effective unionization, are said to have deleterious effects on the economy. Underlying these dominant theoretical propositions and their concomitant policy prescriptions, which deciy minimum wage legislation and the existence of trade unions, is the assumption that given a prevailing factor input mix, higher wage rates do not affect labor productivity.^ But for this to hold true, on average, those firms affected by minimum wage legislation and unions must be maximizing output per unit of labor input, or as Harvey Leibenstein puts it, firms must be operating x-efhciently. As noted, Leibenstein’s general x-efficiency theory suggests, however, that as a rule, labor productiv¬ ity in the firm is not maximized; therefore, the firm generally tends to produce x-inefificiently.^ The argument presented in this chapter holds that to the extent that xinefficiency exists, higher wages “shock” a firm into producing more x-efficiently and, in effect, cause an upward shift in the firm’s marginal product of labor curve. In contrast, lower wage rates serve to shelter a firm from com¬ petitive pressures, allowing the low-wage firm to generate the same rate of 103

104

CHAPTER 5

profit or unit costs as a high-wage firm does but at a lower level of x-efficiency. Under these specified conditions, higher wage rates that result from minimum wage legislation, unionization, or any other variation of nonmarketinduced increases in wage rates need not cause a reduction in either employ¬ ment or employment growth. Theoretically, it is feasible that such wage shocks to the economic system will generate even more employment and higher rates of employment growth than would be possible in a relatively low-wage world. Therefore, introducing the possibility of x-inefficiency in one’s mod¬ eling of the firm allows one to raise the possibility that policy prescriptions flowing from standard theory are inappropriate and misleading and can, in and of themselves, have harmful effects on the economy. Finally, the existing empirical literature is still ambiguous as to the net ef¬ fects of exogenous increases in real wages on either employment or productiv¬ ity."* It remains unclear whether or not exogenous increases in real wages actually reduce employment from what it would be otherwise. On the other hand, the proposition that such wage increases positively affect productivity is less am¬ biguous. There is also a large and growing body of literature that strongly suggests that x-inefficiency in the firm exists (Frantz 1988, 1997). The model developed here, by incorporating the existence of x-inefficiency, allows for such empirical results and suggests avenues for future research. The Existence of X-Inefficiency Underlying the arguments presented in this chapter is an operational defini¬ tion of x-inefficiency and a presentation of the condition necessary for the existence of x-inefificiency. The existence of x-inefficiency is placed in the context of standard economic definitions of economic efficiency and con¬ strained optimization. Further, it is argued that the existence of x-inefificiency is consistent with exercises in constrained optimization on the part of the economic agents within the firm. I assume here that individuals do the best they can to maximize their utility given the constraints they face. These con¬ straints can be of a pecuniary or psychological nature, or even of a techno¬ logical sort, wherein it is recognized that individuals are limited in their capacity to process information.^ It should be noted that assuming non¬ optimizing behavior, which is what Leibenstein is wont to do, would only strengthen my results.^ What is x-efficiency? Leibenstein’s (1978) clearest formulation of this concept can be found in his response to Stigler’s (1976) critique of the xefficiency approach to microeconomics. Leibenstein argues consistently (1966, 1973, 1978, 1983) that x-inefificiency exists when the firm is produc¬ ing above “minimum” average production costs or above the technically ef-

EXOGENOUS INCREASES IN WAGE RATES

105

ficient production isoquant. Minimum costs are achieved at “the cost level that would result if firm members attempted to interpret their job in such a way that they made effort choices which involved cooperation with peers, superiors, and subordinates, in such a way as to maximize their contribution to output” (1978, 206). This represents an ideal of economic efficiency, one that takes place within the ideal institutional framework of the firm where economic agents operate. Only in an environment of trust and cooperation will economic agents provide maximum levels of effort per unit of time and thus a maximum amount of output per unit of labor.^ As one deviates away from such an ideal environment, economic agents choose to provide less effort to the firm (chapters 2 and 9).* It is important to emphasize here that x-eflficiency refers to some maxi¬ mum output per unit of labor and that Leibenstein associates maximum labor productivity with a unique minimum cost of production. However, the con¬ nection is drawn between labor productivity and unit costs because Leibenstein assumes that effort levels vary for a given real wage rate. Effort levels are a function of variations of the work environment independent of variations in the wage rate. However, it is quite possible that increasing labor productivity goes hand in hand with increasing wage rates, such that a range of wage rates exists that corresponds to a unique unit cost of production. X-inefficiency theory, therefore, need not imply that the x-inefficient firm is produc¬ ing at a higher cost than the x-efficient firm or that the x-efficient firm produces at the lowest possible unit cost. Although this is a possibility, it is also quite possible, when one allows for variations in wage rates, that a unique mini¬ mum unit cost corresponds to different levels of x-inefficiency.^ Therefore, there need not be any connection between the level of x-inefficiency and unit costs. To reiterate the discussion in previous chapters, this argument can be expressed by the following equation: wL AC =-

Q

w

Q

(5.1)

L

where AC is average cost, w is the wage rate, L is labor inputs, and Q is output. Assuming, for simplicity, one factor input, average costs need not change as the level of x-efficiency (measured in terms of labor productivity) changes as long as the wage rate changes in a sufficient proportion. The same holds true for the relationship between average cost and movements in the wage rate. Figure 5.1 illustrates the same point, where up to wage W* there is a unique average cost C* and a range of levels of x-efficiency match¬ ing the array of wage rates. Moreover, profits are protected as long as pro-

106

CHAPTER 5

Figure 5.1

Unit costs

ductivity increases, following upon wage increases, sufficient to maintain unit costs constant. It is important to note that in the real world, where there is more than one factor input, the percentage change in labor productivity required to cover an increase in the wage rate or related labor costs would be less than when labor is the only factor input. The required percentage in¬ crease in labor productivity diminishes as the share of labor costs to total costs diminishes. This relationship can be expressed as follows:

dQ , WxL , —=( ) X (-), Q W WxL + NLC

(5.2)

where dQ is the change in output, dWis the change in the wage rate, and NLC is nonlabor costs. For example, if output is at 100 units, labor units are at 10, and the wage rate per unit is $2.00. This yields an average of 20 cents when NLC is zero. In this case a 10 percent increase in the wage to $2.20 requires a 10 percent increase in output to 110 units to keep average costs constant. On the other hand, if in this scenario labor costs comprise only 20 percent of total costs, the required wage increase to keep average cost constant is 20 percent of 10 percent, or 2 percent. Clearly, the extent to which labor productivity need

EXOGENOUS INCREASES IN WAGE RATES

107

be increased to compensate for increases in labor costs is negatively related to the labor intensity of production. For output not to be maximized per unit of labor input—that is, for x-ineffi¬ ciency to exist—requires the assumption of some degree of effort discretion on the part of economic agents.’*^ A simplifying assumption made here is that there are two groups of economic agents in the firms: workers, or agents, and members of the firm hierarchy, or principals. Effort discretion is a product of incomplete contracts between agents and principals with respect to effort in¬ puts. These contracts, in turn, are a product of monitoring and enforcement costs involved in supervising the operations of a firm and the search costs involved in acquiring detailed information on the cost-efficient mode of activ¬ ity of the firm’s economic agents. Thus x-efficiency theory assumes that com¬ plete contracts are typically too costly to construct and enforce relative to the benefits that can be expected from them. Such relative costs could be mini¬ mized in an ideal labor-management environment and would increase as the firm deviates from the ideal. Therefore, one of the increased costs that charac¬ terizes the x-inefficient firm is the transaction cost associated with realizing any particular level of labor productivity and quality of output." Unless an ideal or cooperative work culture prevails, it would not be possible to induce economic agents to work as hard and as well as they might. Even if workers prefer x-efficient behavior given the existence of an xefficient environment, the firm hierarchy might have a different set of prefer¬ ences. There is strong evidence that a more cooperative work culture yields a more productive firm, ceteris paribus (Alcaly 1997; Appelbaum and Batt 1994; Gordon 1996; Ichniowski et al. 1996; Klein 1984; Levine and Tyson 1990). This type of work environment is characterized by significant benefits to work¬ ers, including higher real wages, improved working conditions, and a more egalitarian work culture. However, there is no evidence of symmetrical ben¬ efits flowing to the firm hierarchy. Indeed, the firm must incur costs to make the firm more cooperative. As already discussed, such pecimiary costs might be covered by the improved productivity of the firm, as indeed they must, for the firm to remain cost competitive. In addition, members of the firm hierarchy might incur nonpecuniary costs, such as an increase in their effort inputs and a reduction in their status within the firm relative to workers. Unless the firm hierarchy’s objective function includes a variable for work satisfaction, x-efficiency in production is not consistent with utility maximization on the part of the firm hierarchy. This would be even more so when members of the firm hierarchy possess a subjective distaste for high-priced labor or, more specifi¬ cally, high-priced unionized labor (see Kochan, Katz, and McKersie 1986; Bluestone and Harrison 1988). Given that the firm hierarchy is the deci¬ sion maker within the firm, left to its own wiles and if it is utility maximiz-

108

CHAPTER 5

ing, it will not choose to be x-efficient if the x-inefficient firm is cost competi¬ tive. As already discussed, in term of unit costs x-efficiency in production can be one equilibrium solution to the firm productivity problem, even under con¬ ditions of perfect product market competition. Under these circumstances and given one is in such an x-efficient equilibrium, pressure is required to push the firm into a more x-eflficient equilibrium. Such pressure can take the form of unions and minimum wages. X-efficiency achieved in the firm under these conditions is consistent with constrained utility-maximizing behavior on the part of thd firm hierarchy. However, this equilibrium cannot be a stable one unless the preferences of the firm hierarchy change in favor of the relatively xefificient organizational stmcture. Such pressure need not necessarily result in more x-efficiency within the firm if the firm hierarchy can pass on the costs incurred through unions or minimum wages by raising product prices. Product market imperfections allow the firm hierarchy to raise prices in order to compensate for increased unit costs and to keep the rate of profit constant. Leibenstein (1979, 489490) also argues that the firm hierarchy, when faced with increased pres¬ sures, can attempt to invest in sheltering activities to escape these pressures. Sheltering activities are substitutes for cost-reducing activities, such as those that yield more x-efficiency in production. Once shelters are removed and product markets become more competitive and, moreover, pressure on the firm hierarchy to become more x-efficient is reduced, organizational entropy tends to manifests itself in the typical firm, where organizational entropy refers to the latent tendency within the firm for its economic agents to return to the effort choices that maximize their utility in the absence of external pressures. In this case, if the objective function or preferences of the firm hierarchy remain antagonistic to an x-efficient organizational structure, as shelters and pressure are reduced, it is then consistent with the firm hierarchy’s constrained utility-maximization behavior to revert to the more x-inefficient organizational form. Wage Differentials, X-Inefficiency, and the Shock Effect Refer back to our discussion of the possible multiple equilibria wherein one unique unit cost for producing a particular product is consistent with an array of wage rates or set of labor costs since wage rates and other labor costs medi¬ ated through the work culture of the firm positively affect the level of x-effi¬ ciency. In this scenario, if wages fall in one firm relative to another, the firm hierarchy of the low-wage firm can allow its labor productivity to fall, up to a point, as organizational entropy sets in, and still realize the same unit produc¬ tion costs as the firm characterized by the higher wage rates. On the other

EXOGENOUS INCREASES IN WAGE RATES

109

hand, the firm with higher wage rates must be more productive than the firm with lower wage rates if it is to compete effectively with the lower-wage firm. In effect, higher wage rates and other labor costs serve to pressure the firm hierarchy into reorganizing production so that the utility-maximizing effort input by the firm’s economic agents increases toward competitive lev¬ els, always assuming, of course, that one begins with the existence of xinefficiency. On the other hand, lower wage rates serve as a substitute for other forms of sheltering activity. By lowering wage rates, the firm hierarchy can revert to its preferred utility-maximizing form of organizational structure, even if it is an x-inefficient one. The increased level of x-inefficiency need not cause unit costs to rise or profits to fall as long as the productivity does not drop in disproportional fashion (see Equations 5.1 and 5.2). In such a scenario, a fall in the wage rate is clearly a cause of an increase in x-inefficiency. Firms that are pressured into becoming more x-efficient because of in¬ creased wage rates are experiencing upward shifts in their marginal revenue product (MRP) curves. In such a world, one cannot logically deduce that higher wage rates necessarily result in unemployment or lower profit. Given technology, much depends on the extent to which x-inefficiency exists and the extent to which effort per unit of labor time increases as a result of the increased wage-induced pressure. Where x-inefficiency exists, higher wage rates, by forcing economic agents to contribute more effort per unit of time to keep average costs from increasing, can result in economic agents offset¬ ting, in whole or in part, the negative employment or growth effects that standard theory predicts must follow from higher wage rates in both the short and the long run. For unlike the extension of x-efficiency theory presented above, standard theory critically assumes that the MRP curve is fixed inde¬ pendently of the wage rate. George Stigler, in his critique of minimum wage legislation, has put forth with great clarity the conditions under which minimum wage legislation would generate negative economic effects. These conditions also hold for union interference with the market mechanism; “Unless inefficient workers’ pro¬ ductivity rises ... the minimum wage reduces aggregate output, perhaps raises the earnings of those previously a trifle below the minimum, and re¬ duces the earnings of those substantially below the minimum. These are un¬ doubtedly the main allocational effects of minimum wage in a competitive industry” (1946, 358-359). Moreover, Stigler argues, unless the productiv¬ ity of labor is increased, minimum wage legislation causes less efficient workers to be discharged. At best, these newly unemployed workers will find work at lower returns in the unregulated sectors of the economy. Stigler points out that one possible method of increasing labor productiv¬ ity is for workers to work harder and for entrepreneurs to be shocked out of

no

CHAPTER 5

their lethargy into adopting more productive and profitable techniques. Of interest to this chapter is whether entrepreneurs can be shocked into provid¬ ing more effort and more efficient management. Stigler dismisses these pos¬ sibilities. He assumes that workers can work no harder and that entrepreneurs are doing their best given the hardy competitive environment in which they operate. Richard Lester (1946, 1947) offers a critique of the Stigler argu¬ ment, grounding his analysis in controversial empirical evidence. In essence, Lester argues that firms tend to respond to wage increases by becoming more productive. Nevertheless, to this day, Stigler’s fundamental assumption about the behavior of “economic man” dominates the economics profession. What is critical to the Stigler-type argument is the assumption that x-inefficiency cannot exist and that therefore economic actors are working as hard and as well as they can. An x-efficient supply of effort is always being pro¬ vided to the production process. The argument presented in this chapter hinges on the assumption that where conflict within the firm predominates and where economic agents’ objective function includes nonpecuniary arguments, xinefficiency will exist when pressure is slack. Under these assumptions, higher wage rates can serve to reduce x-inefficiency. Thus exogenous increases in the wage rate can be a determinate cause of higher labor productivity. This is not to say that higher wage rates cannot cause unemployment or slower growth. Rather, given a world of x-inefficiency, the likelihood of negative repercussions emanating from higher wage rates is diminished. This holds true if one assumes a zero growth economy, which is the focus of this chap¬ ter, and also if one allows for the impact that increased wage rates might have on the rate of employment growth over time. The former argument can be illustrated using classic labor demand analysis. In the short run, the demand for labor is given by the firm’s MRP curve [(product price) 3 (marginal labor product)], such as MRP^ in Figure 5.2. Since the firm is expected to hire labor up to the point where the marginal cost of labor equals its marginal revenue, standard theory predicts that given diminishing returns to factor inputs, increasing wage rates will yield a fall in labor demand. Standard theory further assumes that the marginal product of labor curve is fixed independently of movements in the wage rate. The ex¬ tension to x-efficiency theory presented in this chapter suggests that this need not always be the case. If x-inefficiency exists, increasing wage rates, such as from to OfVj in Figure 5.2, can induce economic agents to reduce xinefficiency and thereby shift the MRP curve outward, from to This can be referred to as the x-efficiency effect of increasing wage rates. For employment not to fall in the short run, the MRP of labor must in¬ crease in the same proportion as the wage rate. Since firms will not hire labor when the marginal cost of labor exceeds the value of the average product, the

EXOGENOUS INCREASES IN WAGE RATES

111

Figure 5.2

Units of Labor

relevant portion of the MRP curve for labor is the portion that lies below the average revenue product curve. Therefore, as long as the percentage increase in the average product equals the percentage increase in the wage rate, em¬ ployment will not diminish. This equality will always be realized by firms attempting to keep average costs from rising if labor costs are the only costs and, of course, assuming a sufficient degree of x-inefficiency. If other costs are present, the percentage increase in average costs will be less than the percentage increase in the wage rate by the ratio of labor costs to total costs (see Equation 5.2). Whether the objective of maintaining average costs is consistent with completely offsetting standard theory’s hypothesized nega¬ tive employment effects of increasing wage rates, therefore, depends on the above cost ratio and the ratio of marginal to average product corresponding to the prewage increase level of employment. Nevertheless, increasing labor productivity always offsets, at least partially, the potential negative impact of wage increases upon employment. And, as already mentioned, the most current empirical literature suggests little if any job loss following upon in¬ creases in minimum wages (note 4; Card and Krueger 1995). The same argument can be made with respect to the long run. Here some

112

CHAPTER 5

flexibility is introduced in the firm’s choice of factor inputs (although tech¬ nology is assumed to be constant) and product prices are assumed to change in response to changes in output induced by changes in wage rates. In the long run, standard theory introduces shift factors that serve to diminish the demand for labor as wages increase, to an even greater extent than would be expected in the short run. The four shift factors are the substitution, output, profit-maximizing, and output-price effect. The substitution effect causes capital to replace labor as labor becomes relatively more expensive (holding output to its initial level). This results in more of the cooperating factor, capital, being made available to labor. This, in turn, shifts the MRP curve outward and generates an increase in the demand for labor. The output effect is concerned with the impact of changing factor prices on the ability of the firm to hire factors of production from a given level of expenditure. As wage rates rise, a given budget can purchase less labor and less of its cooperating factor, capital. This causes the MRP curve to shift inward, resulting in a decrease in the demand for labor. The profit-maximizing effect refers to the impact that the increased wage rate has on the marginal cost of production. Since marginal costs increase as wage rates rise for every unit of labor em¬ ployed, less output is produced; thus less labor and its cooperating factor, capital, are employed. This also shifts the MRP curve inward and thereby further reduces the demand for labor. Standard theory assumes that the latter two effects outweigh the first and thereby serve to reduce labor demand as wage rates rise unless, of course, labor is an inferior factor. The net effect of the above three shift factors is illustrated by an inward movement of MRP^ to MRP^q in Figure 5.2, yield¬ ing a long-run labor demand curve of ab. At an industry level, the reduced demand for labor yields, ceteris paribus, a fall in output and thereby a rise in product price. This serves to shift outward, somewhat, the firm’s MRP curve. But it is assumed that the net effect of increasing wage rates remains that of reducing the demand for labor. (See Ferguson 1969, chs. 6 and 8; Ferguson and Gould 1975, ch. 13.) The introduction of the x-efficiency effect, a factor not considered by stan¬ dard theory, serves to offset any long-run reduction in labor demand by shift¬ ing the MRP curve of labor outward. In this case, the long-run effect of increasing wage rates can be illustrated first by an outward shift of the MRP curve from MRP^ to MRP^, which is the x-efficiency effect. The combined substitution, output, and profit-maximizing effects yield an inward shift of the MRP curve to MRP^q. This yields a long-run labor demand curve of cb. Whether labor demand actually falls as a consequence of an increase in the wage rate depends on the extent to which x-inefficiency exists and the extent to which it is reduced as the wage rate increases. However, the impact of the x-efficiency

EXOGENOUS INOREASESIN WAGE RATES

113

effect is somewhat reduced to the extent that the increased output generated by the x-efficiency effect causes the output price to fall. If the wage shock is such that it significantly and positively affects industrywide output, one would ex¬ pect the product price to fall. Thus the MRP curve would have shifted some¬ what inward from MRP^q in Figure 5.2. Nevertheless, as long as the elasticity of product price to output [{dPIP)l{dlQIQ)\ is less than 1, where P is price and Q is output, the x-efficiency effect will cause the MRP curve to shift outward and at least partially offset the fall in labor demand caused by the increased wage rate.’^ Moreover, there is no reason to believe that the product demand is static; it might very well increase over time. Higher wage rates also need not reduce the rate of growth of labor em¬ ployment. As in the no growth scenario, much depends on the extent to which x-inefficiency exists and the extent to which higher wage rates shock the firm’s economic agents into reducing the level of x-inefficiency. The rate of employment growth, assuming no technological change, can be written as:

(5.3)

where n is the rate of growth in labor employment; s* is the propensity to save from total income; k* is the required capital to labor ratio; 7is the value of total output; and L is the number of homogeneous workers employed at equivalent hours per day. The term 5* is assumed to be given by the value of profits relative to total income. Thus s* is, in part, determined by the wage rate. Ceteris paribus, any increase in the wage rate will reduce s* and thereby s*Y, or total savings. This is so since increasing the wage rate reduces profits and thereby savings, given these assumptions. This, in turn, will cause a fall in the rate of growth in labor employment. However, if wage increases result in increases in labor productivity such that profits remain unchanged and thus total savings (s*Y) remain what they were prior to the wage increase, the wage increase will not result in a reduction in the rate of growth of labor employment. In this case Y/L increases sufficiently to offset increases to w. At the very minimum, increasing labor productivity serves partially to offset the negative impact that higher wages would otherwise have on the growth process. Therefore, from a theoretical perspective, one cannot predict that employment growth will be negatively affected by rising wage rates when xinefficiency exists. It must be stressed that the argument that higher wage rates need not have adverse effects on labor employment or its growth rate and that higher wage rates might even positively affect these variables is most potent when the

114

CHAPTER 5

firm is limited in its alternative responses to increased wage rates. The firm hierarchy can respond to increased wage rates by eventually relocating the process of production to a relatively low-wage region. As the costs to the firm in shifting to a low-wage region diminish, the effectiveness of increas¬ ing wage rates to reduce x-inefficiency diminishes. However, lower produc¬ tivity in a low-wage region can result in a low-wage region’s firms being characterized by similar or even higher unit costs than those of firms in a high-wage region. Nevertheless, to the extent that members of the firm hier¬ archy accrue a higher utility in employing low-wage as opposed to highwage labor, where both types of labor yield the same unit costs, new firms will be located in the low-wage region, as opposed to such firms locating in the high-wage region, where x-inefficiency must be reduced if the firm is to survive. Quite possibly, then, the “invisible hand,” when left to its own machi¬ nations, can result in pressures on society that eventually lead to the low wage, or what Bluestone and Harrison (1990, 368—369) refer to as the “low road” to profitability or, more generally, the low road to economic develop¬ ment (see also chapter 6). Conclusion: X-Efficiency and Neoclassical Microeconomics It is critical at this point to recognize a key distinction between the neoclassical and behavioral cum x-efficiency approaches to microeconomics as it relates to the question of output maximization. Stigler (1976) argues, in his attack on xefficiency theory, that the behavior of economic agents within the firm can be fairly modeled as an exercise in constrained utility maximization. The utility function contains a variety of arguments, not solely that of maximizing output since to increase output requires expenditures on the enforcement of contracts between people who would have to engage in unpleasant tasks. Efforts to in¬ crease output would take place to a utility-maximizing degree: “Output and util¬ ity would be larger if resources were not necessary to the enforcement of contracts” (Stigler 1976,213—214). As long as economic agents are engaged in constrained utility maximization, firms cannot be economically inefficient since an increase in output requires the sacrifice of economic resources or the sacrifice of other desired goals. All producers are, therefore, always operating at the production frontier. Only different producers are at different frontiers (Stigler 1976, 215). Therefore, two firms identical with respect to their technical system of produc¬ tion, one with higher unit production costs than the other, are equally efficient. In terms of the argument presented in this chapter, these two firms producing at the same unit costs but operating at different levels of labor productivity are also equally efficient. There can be no x-inefficiency. Producers are always doing their best: they are all maximizing utility given their constraints.

EXOGENOUS INCREASES IN WAGE RATES

115

A key concern in this chapter is to explain why productivity can differ between otherwise identical firms, even if it can be argued that all economic agents of such firms are maximizing utility. Moreover, it is critical to under¬ stand if and how it would be possible for labor productivity to increase in firms whose economic agents contribute a utility-maximizing level of labor productivity and effort input. For Leibenstein, firms whose labor productiv¬ ity is less than potential are x-inefficient, even if the economic agents of such firms are maximizing utility. “Potential” is, of course, defined by what is achievable by the “ideal” firm or, more practically, by the level of productiv¬ ity that is realized by the firm or firms that approach the ideal. Following the logic of standard theory, since all outcomes are consistent with utility maximization, a wage shock that yields less x-inefficiency (higher labor productivity) is also, by definition, consistent with utility maximiza¬ tion. However, a reduction in x-inefficiency does require more effort input or a greater quality of the same by some or even by all of the firm’s economic agents. Moreover, less x-inefficiency also typically requires improvements in employees’ working conditions. This is consistent with Stigler’s conten¬ tion (1976,214, 216) that there can be no such thing as a free ride, that there is a cost to increasing output. However, what is important here is that it is possible for higher wage rates to cause a reduction in the equilibrium, utilitymaximizing level of x-inefficiency and, as a result, an increase in the equilib¬ rium level of labor productivity. The reduction in x-inefficiency takes place due to the pressure that the wage shocks places on the firm’s economic agents, especially members of the firm hierarchy. In other words, the increased pressure changes the moti¬ vational structure of the firm, which, in turn, ultimately results in increasing the size of the economic pie produced by the firm. This process contributes toward increasing the material wealth (goods or services) in the market economy. On the other hand, it remains an open question whether or not reductions in x-inefficiency induced by wage increases ultimately raise the utility of all firm members. Indeed, there might be both winners and losers. Who those winners and losers are and what their distribution is are both questions that x-efficiency researchers must investigate. Nevertheless, the argument presented in this chapter suggests that when xinefficiency exists, increasing wage rates need not result in welfare losses as¬ sociated with a reduction in employment or with a fall in the growth of employment. Indeed, by reducing x-inefficiency, higher wage rates can con¬ tribute to increasing the material wealth of society, which, by shifting the MRP curve upward, prevents, in part or entirely, the decline in employment and employment growth that follow from a rise in wage rates in a world without xinefficiency. Which world is the more appropriate one for economic models to

116

CHAPTER 5

replicate should, of course, be left to empirical research to deterrmne. The most current empirical research suggests that economic theory must allow for the possibility that exogenous wage increases do not have negative effects on the economy and might even have significant positive effects. Notes 1. For the classic arguments on the effects of minimum wage legislation and those of tradounions, see Stigler (1946) and Friedman (1951) respectively. See Lewis (1954) for the quintessential statement on the negative impact of higher wage rates on the process of economic growth and development. Altman (1988) offers a critique of the Lewis argument. 2. Recently, another approach to the labor market, most eloquently expressed by Akerlof (1982,1984) and Akerlof and Yellen (1986), has been developed to explain why nominal wage rates tend to be sticky downward over the business cycle. Referring to it in the literature as efficiency wage theory, its proponents assume that effort per unit of labor time is variable and a function of the wage rate. The profit-maximizing firm is expected to choose the wage rate that minimizes labor costs per efficiency unit (see also Solow 1986). A further fall in the wage rate is expected to yield a fall in labor productiv¬ ity such that unit production costs increase. Thus the wage rate will not fall further even in the face of unemployment. The argument presented here is consistent with the logic of efficiency wage theory in so far as x-efficiency theory also assumes that effort is variable per unit of labor time. However, I do not assume that effort input per unit of time is necessarily a function of the wage rate in that all that is required to minimize cost is for the firm to choose an appropriate wage rate given the effort supply function of labor. In other words, x-efficiency theory does not assume that the firm hierarchy chooses or even can choose wage rates that will necessarily maximize profits or even minimize costs. Therefore, the prevailing wage rate need not be the efficiency wage and a particu¬ lar efficiency wage can be consistent with the existence of x-inefficiency (see chapter 1). See Katz (1986) for a critical review of the efficiency wage literature. 3. See Frantz (1988, 1997) for an excellent summary of x-efficiency theory, as well as of the debates and empirical work surrounding this paradigm. 4. See Pencavel and Hartsog (1984,209,210,215,217) on the relationship between changes in the wage ratio in the union relative to the nonunion sector of the American economy and the relative changes in employment in these sectors. The authors conclude that “when we come to the effect of unionism on relative man-hours worked, we are not at all satisfied that the analysis of these data unambiguously points to a negative effect” (217). The empirical literature on minimum wage legislation does suggest a small but negative effect on the employment of teens. Results for the overall macroeconomic effect on employment of minimum wage legislation are ambiguous. See also Wellington (1991), Brown (1988), Eccles and Freeman (1982), and Brown, Gilroy, and Kohen (1982). In addition, one of the most comprehensive empirical studies to date on minimum wages suggests no negative effect on employment (Card and Krueger 1995). There is much evi¬ dence to suggest that unions both increase wage rates and contribute toward increased labor productivity, although they may also yield lower firm profits (see Brown and Medoff 1978; Freeman and Medoff 1984; Marshall 1987,121-131,149-154). See chapter 6 and Altman (1988) for a discussion of how higher wage rates can increase the rate of economic

EXOGENOUS INCREASES IN WAGE RATES

117

growth by motivating an increase in labor productivity. See Holzer (1990) for estimates on the positive relationship between increasing wage rates and productivity. For an opposite view see, for example, Bemmels (1987). Bemmels uses sample data for 1982 and con¬ cludes that unions have a negative impact on productivity, to a large extent, through their negative effect on the sample firms’ industrial relations. Kochan, Katz, and McKersie (1986) place the relationship among unions, high wages, labor productivity, and profits, in the context of the firm’s industrial relations system and the preferences of the firm hierar¬ chy (see also chapter 9). American management prefers nonunion shops, low-wage labor, and minimal labor participation in the management’s of the firm. To avoid unions, man¬ agement tends to invest in new plants, that are nonunion; therefore, new technologies are embodied in new plants, as opposed to the older, increasingly antiquated union shops (Kochan, Katz, and McKersie 1986,14,34,55,65,66-79,176,180). To keep plants from becoming unionized, given the presence of a union threat, management provides workers with a relatively cooperative, high-tmst framework of industrial relations, and this con¬ tributes to higher productivity in many nonunion plants. The viable threat of unionization spurs firms to become, in effect, more x-efFicient (Kochan, Katz, and McKersie 1986,88103). Thus an important factor in making nonunion plants more productive than union plants, when this is the case, is management’s choosing to adopt more productive tech¬ nologies and management techniques in nonunion plants (Kochan, Katz, and McKersie 1986, 104—108). However, such productive procedures can be and are adopted in union plants if the management cannot avoid the process of unionization (Kochan, Katz, and McKersie 1986,158—176). To the extent that unions diminish in strength and the threat of unionization diminishes, management will tend to compete on the basis of low wages with little labor-management cooperation (Kochan, Katz, and McKersie 1986,228—251). Simply put, the firm’s productivity is related to its technology and its industrial relations system. Union shops can be as productive as nonunion shops. Much depends on the choice of management and, of course, the cooperation of labor. To the extent that nonunion shops are more productive than union shops, this is, to a large extent, due to the threat of union¬ ization providing management with enough incentive to do what it would otherwise pre¬ fer not to do: establish a system of industrial relations based on labor-management coop¬ eration. Without unions, such pressure would collapse, and management would revert to competition based on low wages. See chapter 9 and Bluestone and Harrison (1988) for an elaboration on this theme. 5. The latter is the now classic argument of Simon (1959, 1978). See also Cyert with Hedrick (1988) and Cyert with Pottinger (1988). 6. By maximizing behavior, 1 am assuming only that economic agents are doing their best given the constraints they face. These constraints include transaction and search costs plus the costs in utility that can be a byproduct of the environment of a particular set of labor-mianagement relationships. From Leibenstein, it appears that nonmaximizing behavior is behavior that involves the use of habits, conventions, or incomplete information through bounded rationality. It can be argued, however, that in a world where transaction and search costs are positive, the use of habits and conven¬ tions reduces these costs and is therefore consistent with maximizing behavior. In such a world, incomplete information can be viewed as a product of the costs involved in acquiring more information on the margin. It should be noted that Leibenstein does not consider dropping the maximization assumption as a necessary precondition to the ex¬ istence of x-inefficiency. Defining what is being maximized or minimized is critical for Leibenstein, as is specifying the conditions under which maximization or minimization will or will not occur (1978a; 1985; 1987, appendix).

118

CHAPTER 5

7. It is not clear if this is Leibenstein’s preferred definition ofx-efficiency. Never¬ theless, this definition is adopted in this chapter as the most viable from a method¬ ological standpoint. In many instances, Leibenstein suggests that pressure is neces¬ sary in order to induce individuals into providing more effort per unit of time than they would otherwise. However, this type of pressure is required in a world where economic agents mistrust one another, where labor-management relations are an¬ tagonistic. In this case, a latent Prisoner’s Dilemma solution to the productivity prob¬ lem always lurks menacingly in the background. See, for example, Leibenstein’s most recent contribution (1987, chs. 3—5, 7—8, 10, 12, 14). 8. Weitzman and Kruse (1990,100,101,105,108,114,123—126,133—134) sum¬ marize the literature on profit sharing and conclude that strong evidence exists on the positive relationship between profit sharing and labor productivity given an appropri¬ ate industrial relations system based on trust and cooperation, also Kochan, Katz, and McKersie (1986) and Frantz (1988). Finally, Ehrenberg and Milkovich (1987) argue that the evidence on a positive relationship between the type and level of compensa¬ tion to labor and management is not unambiguous. Compensation polieies are part of a larger resource package and cannot be examined in a vacuum. On theoretical grounds, Rozen (1990) argues that building an industrial relations system based on trust and cooperation is crucial to avoiding what would approach a Prisoner’s Dilemma solu¬ tion to the productivity problem (see also chapter 9). 9. Efficiency wage theory, on the other hand, assumes that labor productivity is a function of the wage rate and that there is a unique wage rate that, when chosen by the firm, minimizes the cost of production through its direct effect on effort supply per unit of labor input. See note 2 above for more details. 10. When Leibenstein refers to “effort discretion” or “effort input,” he refers to more than the pace of effort per unit of time. He understands effort input to have four dimensions: (1) activity, (2) pace, (3) quality, and (4) time duration and sequence. In this chapter, the focus is on the pace component of effort input. In a survey conducted on a randomly selected sample of American jobholders, Yankelovich and Immerwahr (1983) find that effort discretion exists in the American labor market and that such discretion is increasing in the new categories of jobs coming on-line. For example, 75 percent of those surveyed (1983,2) said that they could have been significantly more effective on the job than they actually were. Yankelovich and Immerwahr find that workers would work much harder and much more carefully if there were more and better positive incentives in the workplace. 11. Hashimoto (1990), using modem Japan as a case in point, provides us with impor¬ tant insights on the role transaction costs can play in affecting changes in labor productiv¬ ity. He argues that the low transaction-cost environment that has developed in many Japa¬ nese firms has made it profitable for firms to invest in the employment relationship. 12. However, Wessels (1985) argues that what is referred to here as the x-eflficiency effect must result in even more unemployment than standard theory would predict. Wessels assumes that if labor productivity increases proportionally to wage increases, labor demand must fall in proportion to the increase in labor productivity. Thus, in effect, the firm is constrained to sell what it sold prior to the increase in labor productivity. Therefore, a critical assumption made by Wessels is that the demand for labor is constrained by the firm’s demand function, which, in turn, takes on a particu¬ lar form. This extreme form of a demand function, however, need not characterize firms facing increasing wage rates.

6

A Behavioral Model of Endogenous Economic Growth

Introduction Conventional economic theory tends to support the view that there is only one sustainable wage path to economic development. In the long run, one expects economies to converge to like levels of real per capita gross domes¬ tic product (GDP) with like mean factor prices, of which the wage rate is, of course, one. There can be no such thing as a high-wage path to economic growth and development since convergence occurs through the process of interregional and international trade and factor mobility and is facilitated by the unfettered working of the marketplace. A caveat to this argument is re¬ lated to the work of Paul Romer (1986, 1990, 1994), who, following the earlier work of Arrow (1962) and Young (1928), introduces the notion of increasing returns and positive externalities into the basic neoclassical model pioneered by Solow (1956, 1957; see also Abramovitz 1952) by incorporat¬ ing knowledge, in the form of learning by doing or research and develop¬ ment, in the production function. Gene Grossman and Elhanen Helpman (1994) add endogenous innovations to the basic Romer framework (see also Aghion and Howitt 1998).’ Regions that begin first have a first-mover ad¬ vantage over those that are followers in the investment and development process. Economic growth then takes on a sustained and cumulative form. In this case, convergence need not take place through the free market forces of supply and demand. Off the mainstream, Gunnar MyrdaTs (1963) work on cumulative causa¬ tion argues in favor of persistent divergence of per capita GDP and factor 119

120

CHAPTER 6

prices as a consequence of the unfettered working of the market economy. The law of supply and demand results in the rich getting richer and the poor getting poorer. Once a region or country is ahead of the economic pack, it becomes profitable for investment to be biased toward the leaders. The advantaged areas have, in effect, a first-mover advantage. Only government intervention can break what becomes a vicious cycle. Nicholas Kaldor (1957) also develops an endogenous theory of economic growth whereby the rate of technical change is a function of the growth in capital stock. But like the neoelassical'production function, Kaldor’s enhanced production function is characterized by diminishing returns. He also constructs a model, based upon the work of Myrdal and P. J. Verdoon, wherein exports, exports of manufac¬ tured products in particular, cause differential levels of per capita output and growth rates through the process of cumulative causation since it is assumed that manufacturing is the most dynamic sector in an economy and is charac¬ terized by increasing returns (Kaldor 1970).^ However, like the basic neo¬ classical growth model, the alternative growth models do not specify why one region should be ahead of another to begin with, nor why and when regions should begin the process of convergence or what variables actually initiate the process of divergence. The causal process in all of these models takes on a random form or, alternatively, becomes a function of events exog¬ enous to the model. The empirical evidence is heavily weighted in favor of the argument that convergence has not taken place internationally over time (Altman 1999c; Baumol 1986; Baumol and Wolff 1988; De Long 1988; Dowrick and Gemmell 1991; Pritchett 1997). This is not to say that no convergence whatsoever has taken place. Over the last one hundred—odd years the presently developed economies have converged. Some less developed countries are in the pro¬ cess of converging. Nevertheless, the majority of the less developed coun¬ tries are characterized by laggard economic performances. In point of fact, over the long haul, low-income and high-income economies have persisted side by side. The coexistence of these two types of economies is an impor¬ tant stylized fact and should be explained by growth theory. A model of endogenous growth is developed here that complements both orthodox and nonorthodox growth theories. The main concern of this chap¬ ter is a causal one. I attempt to provide an explanation for sustainable “equi¬ librium” differences in real per capita GDP, as well as for either transitory or “permanent” differences in growth rates that ultimately cause differences in real per capita GDP. I therefore go beyond a discussion of the immediate sources of growth, such as the traditional factor inputs, research and devel¬ opment, and human capital, to an analysis of the motivational structure that might help explain some of these important differences.^ Using the more

A BEHAVIORAL MODEL OF GROWTH

121

realistic behavioral assumptions embedded in efficiency wage and x-efficiency theory, I argue that there are at least two sustainable paths to eco¬ nomic growth. At one extreme is the low-wage path, and at the other is a high-wage path. The latter is associated with a higher level of equilibrium per capita real GDP and the former with a lower level. Convergence in real per capita output need not take place through the workings of the free mar¬ ket. Moreover, competitive pressures (or market forces in general) need not result in the convergence of wage rates. In a word, what I show here is that both a low-wage and a high-wage economy can be consistent with a com¬ petitive economic regime in the long run. But a high-wage economy poten¬ tially yields a higher level of material well-being for all its members. I further maintain that in the short run and under certain conditions in the longer run, the high-wage economy may also be characterized by a higher rate of growth. High wages serve to pressure firms to become more efficient and innova¬ tive, whereas low wages allow firms to engage in less efficient economic behavior, all the while remaining competitive and profitable. For this reason, high wages are viewed as a potential boon to the process of economic growth and development, whereas low wages are viewed as a potential drag. This is contrary to the view typically held by both neoclassical and nonconventional economists wherein high wages are considered to be anathema to the growth process. In the model presented here, economic agents can choose among different available growth paths. There is no one road to growth. The choice made, however, affects the long-mn equilibrium level of real per capita GDP and the rate of economic growth. It is evident that owners and managers of business enterprises, the decision makers, when free to choose, are to an increasing extent choosing the low-wage path to economic growth and prof¬ itability (Bluestone and Harrison 1990; Perelman 1993). In my model, this choice would have predictable negative consequences for the level of mate¬ rial well-being for the majority of a region’s population. To the extent that the purpose of growth is to allow for the maximization of the consumption basket of the typical consumer, this “preferred” path to growth generates inferior or suboptimal welfare outcomes. Only if the low-wage path is deemed the only path to growth can it and all that it entails be considered in any way optimal or welfare maximizing. The Conventional Growth Model and Revisions Conventional neoclassical growth theory does not rule out the possibility of nonconvergence. In particular, to the extent that technical change or innova¬ tions are not transferable interregionally or internationally, there is no reason for per capita GDP to converge. The same holds true if there are significant

122

CHAPTER 6

and persistent barriers to trade. In fact, the classic neoclassical growth model as developed by Solow is not concerned with the question of international or interregional differences in the level or rate of growth in real per capita GDP. Instead, it is largely concerned with establishing the necessary conditions for the long-run sustainability and stability of a given rate of growth. Solow’s model was developed in response to Harrod’s (1939) work on modem growth theory, which suggests the likelihood of capitalist growth being character¬ ized by considerable long-mn instability. In SoloW’s analytical framework the long-mn equilibrium level of per capita output is determined by an exogenously given propensity to save and rate of technical change. The long-mn equilibrium growth rate is given by an exogenously determined labor force growth rate. Long-mn stability is achieved through variations in the capital to output ratio. In the basic model, real output is a function of capital and labor inputs, where the production function is characterized by constant returns to scale while the individual inputs are subject to diminishing returns. Perfect competition is also assumed, as is full employment in the long mn, with Keynesian macroeconomic policy in play in the background. In addition, savings are identically equal to in¬ vestment so that the gro'wth in capital stock is identically equal to savings. It is further assumed that production is x-efficient. Therefore, all economic agents are working as hard and as well as they can. Given these simplifying assumptions, the fundamental neoclassical growth equation is:

k-s(—)-nk,

L

(6.1)

where Y=j{k). The capital to labor ratio is given by k, s is the propensity to save; L is the labor force; n is the rate of growth in the labor force; and the dot above a variable symbolizes its rate of change. Since full employment is assumed to exist in the long mn, n is identical to the growth in employment. The rate of growth in the capital to labor ratio is a positive function of sav¬ ings and is negatively related to the amount of savings required to outfit the growing labor force with the standard capital stock, nk. The capital to labor ratio is critically important to the model in so far as it determines the level of output per worker, subject to diminishing returns and absent technical change. Output per worker, in turn, determines the level of per capita output or GDP. This is evident from the following equation, where P is the population:

y

L ~

(|)xi P

(6.2)

A BEHAVIORAL MODEL OF GROWTH

123

or

(6.2a) The extent to which per capita output is less than per worker output is given by the labor force participation rate (assuming full employment). The growth in per capita GDP is positively related to the growth in labor productivity and employment and negatively related to the rate of growth in population. The rate of growth in per capita GDP is less than the rate of growth in labor productivity to the extent that the population is growing at a faster rate than the labor force: YY--

(6.2b)

For balanced or equilibrium growth it is required that:

(6.3)

Or in terms of output per worker:

y=(-)x/:.

L

(

s

From Equation 6.3, the rate of growth in the labor force n is given by:

(6.4) k

This can be rewritten as: s

(6.4a)

where {K/Y) is the capital to output ratio. The rate of growth in the labor force is the long-run neoclassical equilibrium growth rate for the economy.

124

CHAPTER 6

A rate of growth in real output different from n generates the necessary mar¬ ket forces, through adjustments in the capital—output ratio, to bring the growth in output in tandem with the growth in the labor force. This argument is illustrated in Figure 6.1. A critical point that flows from the Solow model is that if, for example, the propensity to save increases, this does not entail a sustained increase in the rate of growth is output per worker or per capita. Rather, the increase in the rate of growth is transitory, sufficient to drive the economy to a higher level of output per worker, from e* toe**, a function of capital per worker increasing from k* to k**, as the production function shifts upward from sj[ky to sfik)". In equilibrium, the economy remrns to the growth rate given by n. If one introduces technical change into the argument, this too has the ef¬ fect of temporarily increasing the growth rate as the production function is shifted upward, only to return to the equilibrium growth rate given by n. Nevertheless, as is true for an increase in the propensity to save, the economy subject to technical change ends up with a higher level of output per worker. It is important to note that, for Solow, technical change is a shorthand classi¬ fication for any factor that shifts the production function. The term “techni¬ cal change” embodies any and all variables, such as learning by doing, human capital formation, innovations, research and development, and improved ef¬ ficiency, that contribute to enhancing the productivity of factor inputs. Un¬ less all regions and nations are characterized by an identical propensity to save and rates of technical change, which is what conventional wisdom sug¬ gests to be true, convergence in per capita GDP is not expected. With the introduction of technical change the Solow growth equation is transformed into: n+m -

s

(6.5) where m is the rate of technical change. The coeffieient m becomes a perma¬ nent feature of the growth equation insofar as technical change is a perma¬ nent and persistent feature of the economy. More generally, with technical change the growth in real GDP is a function of the growth in factor inputs plus the rate of technical change, and the growth in labor productivity be¬ comes a function of the growth in the capital to labor ratio (which is subject to diminishing returns) and the rate of technical change. In the Solow growth model, this argument takes the following form, where is the technical change shift parameter, K is total capital stock, and L is employment. Con¬ stant returns to scale are, once again, assumed: (6.6)

A BEHAVIORAL MODEL OF GROWTH

125

Figure 6.1

The key to sustained growth in the Solow model is, therefore, technical change writ large. Increased savings can, in the long run, affect only the level of per capita output. And there is a limit to which a society can increase the average propensity to save. Moreover, to the extent that the assumption of diminishing returns to capital is reasonable, there is a limit to the increase in output that can be evoked by increases to the capital stock. Since technical change is so much of a catchall phrase in the neoclassical model and it is assumed to be exogenously determined, this model has been criticized for leav¬ ing most of the growth process unexplained. But as Solow argues, the assump¬ tion of exogenous technical change makes his model more methodologically flexible: “To say that the rate of technological progress is exogenous is not to say that it is either constant, or utterly erratic, or always mysterious. One could expect the rate of technological progress to increase or decrease from time to time. Such an event has no explanation within the model, and may have no apparent explanation at all. Or else it might be reasonably understandable in some reasonable but after-the-fact way, only not as a systematic part of the model” (1994,48). This, of course, leaves us with no clear causal mechanism to explain the process of technical change and economic growth.

126

CHAPTER 6

To address this potential problem, the new growth theory, centered on the work of Paul Romer (1986, 1990; also Aghion and Howitt 1998; Grossman and Helpman 1994; Helpman 1992; Lucas 1988), has evolved wherein growth is explained endogenously. In the new growth models increasing returns to scale are assumed to characterize the aggregate production func¬ tion. In particular, increasing returns are assumed to characterize invest¬ ment in knowledge as distinct from physical capital. Therefore, increasing investment in human capital or research and development has the potential effect of increasing labor productivity without bound (Solow 1994,50). Thus the production function becomes convex so that the rate of change in labor productivity increases as investment in the knowledge sector(s) of the economy increases. As Grossman and Helpman put it with respect to their version of the endogenous growth model: “The endogenous learning here— like the exogenous technological progress of the neoclassical model—pre¬ vents the marginal product of capital from falling to the point where investment ceases to be profitable. Innovation sustains both capital accumu¬ lation and growth” (1994, 35). With investments in knowledge, sustained and persistent growth is possible above and beyond the rate of growth in the labor force n, and technological change, given by m or becomes endog¬ enous to the model. Once the growth process is triggered through invest¬ ments in knowledge, it becomes self-sustaining. Moreover, since investment in knowledge is assumed to be characterized by positive externalities to the economic agent or the firm, it is quite possible for an economy to underinvest in knowledge and for the free market to fail in generating the “optimal” level of investment and therefore the “optimal” rate of economic growth and level of per capita output. The new growth theory, however, does not provide us with a framework within which to understand why convergence does not take place or, in¬ deed, why divergence may take place over time or why leading economies might eventually become laggards—Britain being a classic case in point. One is left with an understanding that persistent differences in per capita GDP among economies and persistent differences in rates in per capita economic growth are a product of a variety of ad hoc factors, among which are different types of economic policy or historical accident. For this rea¬ son, the new growth theory leaves us no further ahead of the analytical game than Solow’s theory of economic growth. Moreover, empirical work on the new growth theory does not provide strong support for the hypoth¬ esis that investment in knowledge is a fundamental cause for sustained economic growth and differentials in growth and per capita output among economies (Fagerberg 1994; Mankiw, Romer, and Weil 1992; Pack 1994; Parente and Prescott 2000; Solow 1994).

A BEHAVIORAL MODEL OF GROWTH

127

High and Low Wages and the Path and Pattern of Economic Growth A critical assumption of all standard models of economic growth is that wage rates play no positive role in the growth process. At most, relatively high wage rates, by inducing higher capital to labor ratios, yield a relatively higher level of per capita output, although at the price of higher unit costs. But economic agents are assumed to be x-efFicient in behavior. For this reason, economic agents cannot be made to work harder or better, and all firms are assumed to be cost minimizers in the sense of maximizing output per unit of input. Once the assumption of x-efficient behavior is dropped, however, the predictions of the conventional neoclassical growth model are dramatically alfected. In a word, by only marginally modifying the simplifying assump¬ tions of the Solow model, a model of economic growth is developed that can help explain persistent differences in per capita GDP and rates of economic growth, as well the movement of regions from leaders to laggards in the growth process.'^ There is now a significant literature suggesting that typically economic agents do not work as hard or as well as they might. In other words, effort is a variable in the process of production (chapter 9; Akerlof and Yellen 1986, 1990; Altman 1992; Blinder, ed. 1990; Frantz 1988,1997; Leibenstein 1966). If effort is variable, then so is labor productivity, and if effort is not maxi¬ mized, neither is productivity. Effort and therefore productivity become dis¬ cretionary variables. Effort variability itself is a product of incomplete contracts (both informal and formal) due to transaction costs in drawing up, monitoring, and enforcing them. Moreover, the behavior of the firm hierar¬ chy is more in line with utility maximization as opposed to profit maximiza¬ tion so that if the marginal utility costs of increasing the hierarchy’s own effort or that of its workers exceed the marginal utility benefits, effort per unit of time will not be maximized. Nevertheless, the bottom line for the firm hierarchy remains the competitiveness of the firm, at least in the short run. An important tenet of the behavioral model presented in this book is that in the absence of an ideal work environment, x-efficient behavior will be the exception to the rule. X-efficiency is defined as that level of output that can be achieved in the ideal, relatively cooperative work environment. It is pos¬ sible for x-inefficient firms to survive even with perfect product markets if wage rates differ among firms and if the different wage rates and the work cultures, for which they may be a proxy, affect labor productivity by affect¬ ing the effort inputted into the process of production. The rate of labor com¬ pensation is therefore positively related to effort per unit of time and thereby to the level of labor productivity. Under these circumstances, a low-wage

128

CHAPTER 6

regime goes hand in hand with a low-productivity regime and a high-wage regime goes hand in hand with a high-productivity regime, and the two wage regimes need not necessarily yield any differences in unit costs or even in rates of return. This argument can be expressed as:

(6.7) L

where AC is average cost, w is the wage rate, and (Q/L) is output per unit of labor input. For simplicity it is assumed that labor is the only factor input. The main point here, however, is that^C remains unchanged in the face of changes in the wage rate (w) as long as labor productivity (Q/L) changes sufficiently. Under these conditions, one can posmlate a given level of average and marginal costs and rate of return to capital that correspond to an array of rates of labor compensation. There need not be any unique wage rate or labor compensation package that minimizes costs or maximizes the level of profits (chapters 1 and 2; Altman 1998; Akerlof and Yellen 1990; Stiglitz 1987).^ In this case, high-wage firms can compete with low-wage firms by becoming more productive, and output is a function of effort per unit of labor input s, as well as of labor, capital, and technical change.

(6.8) In this scenario, low-wage—low-productivity and high-wage-high-produc¬ tivity firms and, more generally, low-wage-Tow productivity and high-wagehigh-productivity economies can exist simultaneously and persist over time since there is no mechanism in place within the structure of the free market (perfect product market competition) to drive the wage rates and the levels of labor productivity toward convergence: low-wage x-inefficient firms do not have a competitive advantage over high-wage firms that can ultimately force convergence to take place. Indeed, low wages serve to shelter x-inefficient firms from competitive pressures. High wages are compensated for by higher productivity. On the other hand, if relatively high rates of labor compensa¬ tion are not followed by sufficient increases in labor productivity, the highwage firms will, ceteris paribus, fall by the wayside in the process of interfirm or international competition. One can expect this to occur once, given tech¬ nology, firms become x-efficient and effort becomes perfectly inelastic to increases in wages or improvements in overall working conditions. It is theo¬ retically possible for wage rates to converge in the long run if perfect mobil-

A BEHAVIORAL MODEL OF GROWTH

129

ity exists in the labor market. I assume that labor market mobility is far from perfect. In fact, labor does not move freely, nor does it costlessly flow from one region or country to another. Apart from this, many economies, espe¬ cially the high-wage economies, have institutional mechanisms in place (the right of workers to organize and a social welfare safety net) that can main¬ tain wages at relatively high levels even in the face of significant labor mo¬ bility. As long as relatively high rates of labor compensation are accompanied by relatively high rates of productivity, the product market cannot generate the forces necessary to bring about the convergence in per capita GDP. The notion that relatively high wages can positively affect labor produc¬ tivity and thereby cause long-run international differentials in real per capita GDP is not a new one. Such a scenario was suggested by H.J. Habbakkuk (1962) in his classic comparative study of nineteenth-century American and British economic development. He argues that America’s advantage lay in the relatively high rates of labor compensation that characterized its indus¬ trial sector. Habbakkuk maintains that in order to remain competitive and profitable, American firms were forced to make more efficient use of their factor inputs (to be more x-efficient), and American entrepreneurs were in¬ duced into becoming more creative and innovative in developing new tech¬ nology, as well as being more apt to adopt known technology. Thus in Habbakkuk’s interpretation of American-British comparative economic de¬ velopment, America’s relatively high-priced labor was the primary cause of two key events. First, it affected the x-efficiency of production. Second, it affected both the extent and the rate of technological progress. All of this would have the effect of shifting the production function upward, such as from sf(k)' to sf(k) " in Figure 6.1, yielding a higher level of per capita output in British as compared to American firms. The critical adjustment to the conventional neoclassical growth model suggested in this chapter is that given the possibility of x-inefficiency, rela¬ tive differences in wage rates in like industries affect the differences in xinefficiency among industries and economies and the extent and pace of technological progress. This, in turn, affects equilibrium differences in real GDP per capita among economies, as well as, at a minimum, the short-run growth rate in per capita real GDP. Output becomes a function not only of capital and labor, but also of relative rates of labor compensation and other labor costs, which, in turn, affects the shift or technical change parameter through its effect on x-inefficiency and technological progress:

Y = M^Y(K,L). Wj

(6.9)

130

CHAPTER 6

Output is now only an indirect function of effort since effort per unit of time is assumed to be the product of relative rates of labor compensation among economies {W/Wp. To simplify the discussion, I assume that the output of two (or more) economies is identical. In this model, an increase in the relative price of labor has the immediate and short-term effect of increasing the cost of production and reducing the rate and amount of profit. If the firm is to survive, however, ceteris paribus, an increase in the relative rate of labor compensation must be accompanied by a sufficient increase in labor productivity. This wage effect on productiv¬ ity can be broken down into three components. The x-efficiency effect, the technological progress effect, and the savings effect will be discussed in turn. 1. The x-efficiency effiect. High-wage firms are pressured into becoming more x-efficient. The decisionmakers of the firm cannot be expected to in¬ crease x-efficiency on their own if the benefits of the increase accrue largely to the employees and if the process incurs increased disutility to members of the firm hierarchy in terms of increased effort, loss of status, and even loss of jobs to some (chapter 9). Increasing the level of x-efficiency serves to shift the production function outward. On the other hand, the relatively low-wage firms remain relatively x-inefficient as a result of low wages. 2. The technological progress effiect. High-wage firms are forced to adopt already available technologies that are new to these firms. These firms may also find it worthwhile to develop new technologies (to innovate). The new technologies might also require the investment in on-the-job training and formal education for employees and the firm. Technological progress shifts the production function upward. Low-wage firms need not adopt new tech¬ nology or innovate if they can effectively compete on the basis of low wages. Moreover, the low-wage regime may result in labor being too x-inefficient for the new technology to be cost-effective. This point is illustrated in Figure 6.2. Assume two economies, / and J, producing identical outputs and using identical technologies at identical factor prices. The initial equilibrium for both economies is at e^ at isoquant Qy The wage rate increases in economy / such that the isocost curve ac pivots to ab and is tangent to isoquant at e^, where Qq represents a smaller amount of output than . In the new equi¬ librium, the high-wage economy is more capital intensive than under a lowwage regime, and unit costs have also increased. For to be produced by the high wage economy /, expenditures would have to increase, shifting Ts isocost curve to dc. The new equilibrium is found at e,. The old, higher level of output can now be produced, but only at a higher cost than in the lowwage economy. For the high-wage economy to survive, assuming away for the moment the possibility of affecting the degree of x-inefficiency, techni¬ cal change must take place. For example, if g^embodies the new technology

A BEHAVIORAL MODEL OF GROWTH

131

Figure 6.2

such that output at Qq equals output at , unit costs in the high-wage economy would be equal to unit costs in the low-wage economy. The low-wage economy, however, need not adopt the new technology to compete. If, how¬ ever, the new technology adopted by the high-wage economy is represented by isoquant Q, where the level of output Q equals that at Qj, the low-wage economy would be pressured to adopt new technology since otherwise unit costs in the high-wage economy would be lower than in the low-wage economy. Otherwise there is good economic reason for the low-wage economy to avoid technical change. The low-wage regime may not be capable of ex¬ ploiting the new technology as x-efficiently as the high-wage economy. For this reason, the low-wage firm shifting to the new technology given by Qq need not reduce average costs (Leibenstein 1973). Therefore, shifting to the new technology may not make the low-wage economy any more competi¬ tive. Moreover, adopting the new technology may involve both net financial

132

CHAPTER 6

and utility-related costs to the firm hierarchy, at least in the short run. These costs might include improving the working conditions of labor and more time and effort and increased stress on the part of the firm hierarchy. Also, if the time horizons of the firm hierarchy are not especially long, technological change will not be utility maximizing and will be avoided. The low-wage economy can also respond to the competitive challenge presented by techni¬ cal change by forcing wage rates downward. Low wages can serve as a sub¬ stitute for technological change. The effectiveness of the latter strategy depends, of bourse, on the impact that a fall in labor compensation has on the level of x-inefficiency. For these reasons, technological progress is not inevi¬ table. The relative rate of labor compensation can play an important role in determining the timing, extent, and rate of technical change. Only when the new technology overwhelms the other technologies in use in terms of the unit cost at which output can be produced at all firms, no matter their relative labor costs, will firms engage in technological progress (chapter 2).^ 3. The savings effect. Under pressure to adopt new technology or new forms of organization by higher rates of labor compensation, high-wage firms may be forced to increase the propensity to save so as to increase the rate of investment. This may result in an increase in the economy’s propensity to save. Much depends on the extent to which increasing wages serve to pull down the economywide or average propensity to save (with workers being characterized by a relatively low propensity to save). To the extent that the saving rate increases, the production function shifts upward. These three wage effects serve to increase the level of real per capita GDP in the high-wage economy. In the short run, at least, the rate of per capita growth is also augmented as the high-wage economy adjusts to the new higher level of per capita output. It is quite possible that in the long mn the growth rate in output will converge toward the growth rate of the labor force. How¬ ever, growth may be positively affected over the longer run to the extent that relatively higher and increasing rates of labor compensation affect the rate of improvement in x-efficiency and the rate of technological progress writ large. But clearly the wage effects can be expected, at a minimum, to increase real per capita output and the short-run per capita growth rate. Moreover, the consequences of higher rates of labor compensation should be stable insofar as they are consistent with Solow’s stability conditions for economic growth. In addition, the low-wage economies have no competitive advantage over the high-wage economies. In this sense, both the high- and low-wage path to economic growth and development are sustainable in the long run. The high-wage regime might be more unstable than a low-wage regime, however, insofar as corporate leaders might have a preference for the latter and have the capacity to undermine the former by political means or by relo-

A BEHAVIORAL MODEL OF GROWTH

133

eating to low-wage regions. This is particularly true if an economy approaches the neoclassical ideal of perfect competition, with the perfect mobility of capital and the absence of tariffs. Footloose industries would be more prone to this particular problem. High-wage industries tied to a region or country for geographical, economic (transportation costs, economies of scale), or political reasons would be more stable. A high-wage regime would also be difficult to sustain if the necessary infra¬ structure to maintain a high level of productivity were not established. This includes the educational, research and development, health, legal, and trans¬ portation and communication inffastmeture, which Moses Abramovitz (1986) refers to as a nation’s “social capabilities.” Social capabilities represent neces¬ sary conditions for sustained growth and development (see also Olson 1996). High wages may induce governments to engineer the social capabilities neces¬ sary. A low-wage regime may have the opposite effect on government. On the other hand, adequate social capabilities are not sufficient to induce high rates of economic growth and high levels of real per capita output. On a microeco¬ nomic level, high wages are necessary to induce firms to become more xefficient and to adopt more productive technology.^ The key point this chapter seeks to make is that one can develop a reason¬ able endogenous theory of economic growth by introducing wage rates as a causal determinant of the extent and rate of x-inefficiency and technical change into the Solow growth model. Thus the origins of important and stable dif¬ ferences in per capita GDP and growth rates become more comprehensible. Moreover, it is now possible to more clearly explain how an economy can revert from leader to laggard in the growth and development process. Whether or not convergence takes place depends on the state of the labor market and the extent to which an economy can sustain relatively high rates of labor compensation. This is not to say that other factors, both endogenous and exogenous, are not important to explaining this process. However, by focus¬ ing on the relative rate of labor compensation, one is isolating an easily iden¬ tifiable and measurable as well as significant economic variable as a possible supply-side determinant of the growth process. This variable has typically been ignored in the literature as an important determinant of x-efficiency and technical change. As with all hypotheses, the one presented in this chapter must bear the test of empirical scrutiny. Conclusion: The Low Wage Economy and Market Failure To the extent that both high- and low-wage economic regimes are sustain¬ able over time, the low-wage regime can be regarded as a product of a mar¬ ket failure. The objective of economic growth is not growth itself; rather it is

134

CHAPTER 6

to maximize a social welfare function. One might argue that an important argument in such a welfare function is the basket of commodities consumed by members of an economy over a specified time horizon. A low-wage-4owproductivity economy cannot serve as the basis for maximizing the consump¬ tion basket of the typical consumer of such an economy. Apart from this, such an economy produces a level of output that is below its potential where this potential is given by what is produced in the high-wage-fiigh-productivity economy. But the behavior of economic agents in both economies can be consistent with utility-maximizing behavior. And this is the rub. The lowwage economy might be preferred by the utility-maximizing members of the firm hierarchy and corporate leaders in general, as well as political leaders— that is, those individuals who ultimately determine the wage regime firms follow given the institutional, social, political, and economic constraints they face. What is of some importance from the perspective of social welfare is whether the utility function or preferences of the decision makers take into account the preferences of the larger population. In other words, do exter¬ nalities exist when choosing the wage regime that an economy is to follow? If a high-wage economic regime is adopted, this can be expected to in¬ crease the utility of workers in terms of the higher income generated. These benefits are typically not part of the preference function of corporate leaders. In this case, by maximizing their own utility, the corporate leadership will tend to “underinvest” in the development of a high-wage economy. But why might a low-wage regime be favored by corporate decisionmakers when both regimes yield competitive and profitable economic structures? It all depends on their utility function. If utility is negatively related to more work and effort on the part of the firm hierarchy and these are required to develop a high-wage economy, the marginal costs of developing a more productive economy tend to outweigh the marginal benefits. Such benefits might be zero, given our assumptions. If creating a high-wage-Tiigh-productivity economy also involves investments that can only be covered over time, this will further encourage a preference for a low-wage economy. This is espe¬ cially true if members of the firm hierarchy are present-oriented, if they have a high rate of time preference. Since choosing a low-wage regime ignores the benefits accruing to consumers in general, such a choice generates a market failure. The preference for a low-wage regime on the part of members of the firm hierarchy cannot always be realized, however. They are not always free to choose. The state of the labor market acts as a powerful constraint on the be¬ havior of corporate leaders. Labor market conditions, which are influenced by supply and demand considerations as well as by institutional factors, can pre¬ vent the low-wage route to growth and development. But once the labor mar-

A BEHAVIORAL MODEL OF GROWTH

135

ket becomes “flexible,” the low-wage regime becomes the chosen path of cor¬ porate leaders unless their utility functions undergo a significant transforma¬ tion. As Bluestone and Harrison remark, an environment has emerged “in which it was simply easier for many firms to attempt to contain their own labor costs than to seek enhanced profits through investments in expensive new plant and equipment. For the most part, firms abandoned revenue-enhancing strategies to boost profits and turned sharply towards tactics that emphasized cost reduc¬ tion instead” (1990, 369). This comment on the American economy is appli¬ cable to an increasing number of developed high-wage economies. To the extent that the choice of a low-wage route to growth represents a market failure, it is important to develop institutions that prevent such a fail¬ ure. At a minimum, labor market institutions need to be developed that mini¬ mize the probability of the low-wage path when the high-wage path is a competitive and profitable option. Moreover, an institutional environment (social capabilities) must be developed that contributes toward making the high-wage path competitive and profitable. This is particularly important since, according to the behavioral model of economic growth presented here, the low-wage path is viable and sustainable over time in spite of its being suboptimal from a social welfare perspective. Market forces, left to their own devices, cannot easily push an economy from a low- to a high-wage path. As a result, there are significant long-run regional and international differences in per capita output. Notes 1. Many of these arguments, often referred to as the new growth theory, are pre¬ dated by Abramovitz (1952), albeit in a less formal presentation. Moreover, Schumpeter’s (1974) classic study on technological change, first published in 1934, has also informed much of the new growth theory literature, as well as the recent work on evolutionary growth (see Nelson 1995, 1998). 2. See Hodgson (1989) for a critical assessment of Kaldor’s theory of cumulative causation. Kaldor’s (1970) rendition of “Verdoon’s Law” is that there exists a strong positive relationship between the rate of growth of productivity and efficiency and the rate of growth in the scale of economic activities. Thus there is a strong positive correlation between the rate of growth of exports and the rate of growth of productiv¬ ity and efficiency. 3. Nelson (1998) points out that both the basic Solow growth framework and the new growth theories fail to take us beyond a discussion of the immediate sources of economic growth. He argues that variables that underlie the immediate sources of growth hold the secret to a better understanding of the persistence of growth and per capita income differentials across nations. Nelson identifies the organizations that adopt technologies, the institutional framework in which they operate, and the pro¬ cess of technological change as key underlying variables that affect the process of economic growth.

136

CHAPTER 6

4. Unlike the Solow model, the model developed in this chapter does not assume x-efficiency in production or perfect labor market mobility. Also, as v^'ith the con¬ ventional growth models the demand side is not dealt with here, which is not to say that it is unimportant. Rather, I focus on the supply-side variables that may affect the growth process. See You (1994) for an alternative wage-led growth model that incorporates the demand side. 5. In the traditional efficiency wage literature there exists a unique wage that maximizes effort per unit of time and thereby minimizes the cost of labor per unit of time. This unique wage is chosen by the profit- maximizing firm (Akerlof and Yellen 1986). 6. The argument presented here clearly fits in with the notion of induced techno¬ logical change, wherein changes in relative factor prices affect the direction of tech¬ nological change (Hayami and Ruttan 1971; Hicks 1932; Ruttan 1997). 7. Reviewing the history of technological change. Nelson and Wright (1992) find that the globalization of the world economy has resulted in common technolo¬ gies across nations, easing the international transfer of technology when the appro¬ priate social capabilities are available in the target nations. Fagerberg (1994) finds that the most important determinant of technological change, given the existence of a sufficiently strong social capability, are the “intentional activities of private firms.” The model presented in this chapter suggests that in a world of accessible technolo¬ gies and adequate social capabilities differentials in wage regimes can have a sig¬ nificant impact on the “intentional activities of private firms.”

7

Interfirm, Interregional, and International Differences in Labor Productivity Why Convergence Need Not Take Place Introduction Economists have traditionally attempted to explain labor productivity differ¬ ences among firms, regions, and nations by focusing upon differences in capital—labor ratios, differences in the quality of labor and capital, and dif¬ ferences in the quality and quantity of investments in human capital.^ Rich¬ ard R. Nelson (1968,1984) broke with this tradition by arguing that differences in the technologies employed by firms at a given time should be incorpo¬ rated into the analysis (see also chapter 6; Altman 1998; Parente and Prescott 2000). Liebenstein (1980; 1987, ch. 4) argues that productivity differences among firms in the same industry and among countries at a similar stage of development can be explained to a large extent by the differences in motiva¬ tional systems embodied.^ However, since differences result, according to Leibenstein’s schema, in differences in unit costs, the firms cannot survive in competitive markets (Leibenstein 1966, 408-410; 1973a, 772—775; 1979, 488,490,492; 1983b, 841). Of course, the conventional wisdom argues that such productivity differences should disappear over the long haul as long as competitive conditions prevail. Therefore, there should be convergence in productivity across firms, regions, and nations in the long run as market forces have the opportunity to effectively discipline economic agents into becom¬ ing as efficient as possible. Relatively unproductive economies simply can¬ not compete in the open market. But as pointed out in chapter 6, such convergence has simply not taken place in the world economy, even among relatively open economies (Altman 1999c; Baumol 1986; Baumol and Wolff 1988; De Long 1988; Dowrick and Gemmell 1991; Pritchett 1997). In this chapter, I outline the conditions under which it becomes possible 137

138

CHAPTER 7

for differentials in productivity among identical firms producing identical products to be sustained in the competitive long run. The basic finding is that one can help explain persistent productivity differentials by introducing wage rates and related labor costs as independent variables, where relative produc¬ tivity differentials among identical firms are the dependent variable.^ In the model discussed here, wage differentials affect productivity differentials through their impact upon the degree of x-inefficiency in the firm. And un¬ like in the traditional x-efficiency literature, I show that relatively high wages need not cause increases in unit costs or reductions in profits. Wage differen¬ tials can be sustained over time when they are causally related to differences in productivity.'* This argument is directly linked to and overlaps with the behavioral model of economic growth developed in chapter 6. Related to the theme of sustainable productivity differentials is the ques¬ tion of sustainable differentials of wages for identical types of workers pro¬ ducing identical products in different firms and for identical types of workers employed by firms producing different products, such as plumbers, electri¬ cians, computer programs, or secretaries. This is related to the “law of one price.” In reasonably competitive markets the act of buying and selling should force the price of identical goods and labor inputs to converge in the long mn, absent significant transactions and transportation costs. Why pay more for something when you can get it for less? (See, for example. Thaler 1992a, 36-37.) Here too convergence does not appear to have taken place (Krueger and Summers 1988; Thaler 1992a). The behavioral explanation for sustain¬ able differentials in productivity also provides at least one possible explana¬ tion for the lack of convergence in wages, such as compensating differentials, quality of labor differences, and efficiency wages. Indeed, the differentials in productivity and wages are causally related, and one could not exist with¬ out the other. Wage Differentials and X-Inefficiency X-inefficiency can be defined as a difference in productivity between what can be achieved when utility economic agents supply effort in an ideal work environment and what is supplied in a suboptimal work environment. Only in the ideal, relatively cooperative work environment will effort inputs be in some sense maximized, thereby maximizing output per unit of labor (chap¬ ter 9). But for x-inefficiency to exist, effort must be a discretionary variable. The conventional wisdom, however, assumes that this is not the case and that effort inputs must be maximized, for otherwise the low effort inputs would increase unit costs and make the affected firms uncompetitive (chapter 3). But in both the x-efficiency and efficiency wage literature a basic character-

DIFFERENCES IN LABOR PRODUCTIVITY

139

characteristic of the real world of labor markets, the transaction costs in draw¬ ing up, signing, monitoring, and enforcing contracts, results in incomplete contracts and therefore in effort being a discretionary variable. If one as¬ sumes for simplicity two groups of economic agents within the firm, work¬ ers (agents) and the firm hierarchy (principals) or the decision makers in the firm, the marginal cost of producing complete contracts exceeds the mar¬ ginal benefits to the firm decision makers (Frantz 1997; Akerlof and Yellen, eds. 1986; Leibenstein 1979; Stiglitz 1987). In the traditional x-efficiency literature, x-inefficiency in production re¬ sults in higher unit costs; therefore differentials in x-inefficiency yield dif¬ ferentials in unit costs that simply cannot be maintained in a competitive environment. As can be seen from the following Equation (7.1), assuming for simplicity labor as the only factor input, decreasing the level of x-efficiency and thereby reducing labor productivity (Q/L), increases average cost (AC), ceteris paribus. Of course, average cost is here a positive function of the wage rate (w) and a negative function of the level of labor productivity.

L

However, this line of reasoning pays no heed to the known positive relation¬ ship between the work environment of the firm, inclusive of the wage rate and working conditions, and labor productivity (chapter 9; Alcaly 1997; Appelbaum and Batt 1994; Barney 1995; Becker and Huselid 1998; Gordon 1996; Ichniowski et al. 1996; Kochan, Katz, and McKersie 1986; Levine 1995; Levine and Tyson 1990; Neal and Tromley 1995; Pfeffer 1995). In this case, the level of x-efficiency is positively related to the work environ¬ ment of the firm and therefore to the economic costs involved in making the firm more x-efficient. Under these circumstances, relatively low wages, a proxy for the firm’s work environment, both contribute to the existence of xinefficiency and shelter the relatively x-inefficient firm from the relatively x-efificient firm. On the other hand, relatively high wages contribute to re¬ ducing the level of x-inefficiency by inducing increased productivity. This increased productivity, in turn, serves to offset the impact on unit cost that in¬ creasing labor costs would otherwise have. In other words, higher labor costs serve to pressure the firm into reducing the level of x-inefficiency. Otherwise, the firm would lose its competitive position. For this reason, even in the face of perfectly competitive product markets, x-inefficiency can be of significance, and differentials in productivity among identical firms can persist over time in competitive product market equilibrium. In this example, it is possible for

140

CHAPTER 7

there to be a range of wage rates associated with a unique unit cost and matching an equilibrium product price. Corresponding to this, there is a range of productivity levels cum x-inefficiency levels associated with the same unique unit cost, as differences in productivity levels serve to offset any dif¬ ferences in labor cost among firms. This argument is, in part, illustrated in Figure 7.1. When each wage rate is matched by a correspondingly higher level of labor productivity, an array of wage rates is associated with a unique average cost (see Equation 7.1). Once the level of labor productivity can no longer compensate for the higher wage rate, past point fV*, the average unit cost increases. This occurs, given technol¬ ogy, when effort inputs enter into the realm of diminishing returns. Further to this argument, as illustrated in Figure 7.2, for unit costs to remain constant as the wage rate increases, labor productivity must be a linear function of the wage rate by way of the wage rate’s impact on effort supply; otherwise unit costs will either increase or decrease as the wage rate goes up. In Figure 7.2 a linear relationship between labor productivity holds up to fV*, corresponding to B level of productivity. Thereafter, diminishing returns set in. From this point onward, average costs must increase as labor costs rise. It should be noted that this particular modeling of the firm differs from the conventional efficiency wage approach, which maintains that although effort is a variable input, there is one unique wage rate that minimizes the average cost. This is the efficiency wage. The relationship between wages and produc¬ tivity, and therefore wages and effort, is nonlinear (Solow 1986; Stiglitz 1987). Wage rates that deviate from the efficiency wage increase unit costs. Rational firm decision makers therefore choose the efficiency wage. In this scenario, there is only one unique wage associated with a unique level of productivity. If all firms are identical, all firms should be equally efficient, and there should be no interfirm differentials in productivity or wages. Firms where the efficiency wage does not prevail would not be cost competitive. This particular result is consistent with the conventional wisdom. The behavioral model, on the other hand, is consistent with the existence of such differentials because no effi¬ ciency wage is assumed a priori. Firms with varying degrees of x-efficiency can therefore exist simultaneously in competitive equilibrium. A numerical example might further clarify the above discussion of the be¬ havioral model. Compare one firm A to other firms B in a particular product market where all firms are identical and produce identical products at the same unit cost, which equals the equilibrium market price. The only factor input is labor. The “law of one price” mles here. The initial wage rate, $10 per hour, is the same for A and B, as is labor productivity at 100, at a price of $1.00 per unit. This yields an average cost of 10 cents, from Equation 7.1. If the wage rate increases by 10 percent to $10.10 in A, average product must increase by

DIFFERENCES IN LABOR PRODUCTIVITY

141

Figure 7.1

Unit costs

Figure 7.2

Wage

142

CHAPTER 7

10 percent to 110 for the average cost to remain unchanged at 10 cents. No change in productivity yields an average cost of 11 cents. Thus higher wages need not result in an uncompetitively high average cost. The end result of this exercise is the existence of both relatively high- and low-productivity firms. If one introduces a normal profit into this example, to the extent that the produc¬ tivity increase simply serves to offset wage increases, profits do not fall as a consequence of wage increases. However, in this case, those who gamer the profits, typically members of the firm hierarchy, derive no material benefit from increasing or decreasing the level of x-efficiency. Nevertheless, in face of increasing labor costs, members of the firm hierarchy would find it advanta¬ geous to increase the level of x-efficiency so as to maintain unit costs and profits at competitive levels. On the other hand, the low-wage B firms face no pressure to reduce x-inefficiency as their unit costs and profits remain at a competitive level given their relatively low wage rates. Thus the low wage rates serve to shelter the relatively x-inefficient imits of production. In this example, differences in wage rates are associated with differ¬ ences in labor productivity. This causal relationship can hold, however, only if x-inefficiency exists. Clearly, if there is no x-inefficiency to begin with, the relative increase in wage rates can result only in lower profits and eventually higher unit prices in the high-wage firm, which, under competi¬ tive circumstances, must result in the economic demise of that firm. Even given the existence of x-inefficiency, the ability of the firm hierarchy to respond to increasing wages is constrained by the extent of the x-ineffi¬ ciency. A further qualification to this argument is that wage increases pres¬ sure the firm hierarchy into reducing x-inefficiency only when the firm hierarchy is no longer able to reduce profits further in order to compensate for the increased costs of production. Reduced profits can serve as an alter¬ native to increasing labor productivity if one can assume that the firm hierar¬ chy can accept a lower rate of profit or lower total profits. Thus to the extent that the disutility of increasing labor productivity exceeds that of a reduction in profits, the firm hierarchy can be expected to refrain from reducing xinefficiency. Nevertheless, there is a limit beyond which profits cannot fall. At the extreme the limit would be zero. But, more realistically, the limit lies above zero due to considerations such as the desired income of owners/man¬ agers, a lower-bound price of shares to discourage takeover bids, and re¬ quired internal funds for investment. When this limit is reached, x-inefficiency must be reduced. Ultimately, therefore, there remains a positive relationship between relative wage increases and relative increases in labor productivity. The main point of this argument is that given the existence of x-inefficiency and a causal relationship between labor costs and effort inputs, labor produc¬ tivity differences can exist across otherwise identical firms in competitive

DIFFERENCES IN LABOR PRODUCTIVITY

143

product market equilibrium. The dual of this statement is that under these circumstances, identical workers can earn different wage rates and related pecuniary benefits. If one introduces non wage shelters from competitive pressure into the argu¬ ment, the results presented here should not change. If one starts from an equi¬ librium position where wages per unit of output are the same in all firms and where firm A is assumed to be somewhat sheltered from competitive pressures by product differentiation, for example, x-inefficiency in firm A can exceed that of the B firms in equilibrium, and thus unit costs in firm A can be some¬ what higher that in the B firms. Further increases in wages in firm A will, however, cause its unit costs to rise ceteris paribus, such that its market share will be lost to the B firms. Thus the firm A hierarchy will be pressured to reduce the level of x-inefficiency so as to bring the unit costs of the firm in line with the existing state of competitive pressures. Therefore, given the particular structure of shelters in a particular product market, relative wage increases or the existence of wage differences among firms can be a cause of changing differentials and the existence of differentials in x-inefficiency. Nevertheless, it is important to recognize that to the extent that firms, regions, or nations can institute or increase nonwage shelters, this relieves the pressure on the firm hierarchy of the relatively high-wage firms to reduce x-inefficiency. In this sense nonwage shelters and low wages serve as substitutes. Both reduce the pressure on x-inefficient firms to become more x-efficient. Given nonwage shelters, higher prices can compensate for increasing unit costs as opposed to increases in labor productivity. Thus up to some point (determined by the extent and significance of nonwage shelters) relatively high-wage firms can be as x-inefficient as relatively low-wage firms. And to the extent that nonwage shelters can be instituted as relative wages increase, the increase in wages need not cause reductions in x-inefficiency. In this case, wage differentials might exist in identical firms, but these differentials will not correspond to or be a cause of productivity differentials. At this point, it is important to note that governments have a greater capacity to institute nonwage shelters than firm hierarchies. Governments can create and raise tariffs and grant subsidies. Thus they have a greater potential to neutral¬ ize the pressures to reduce x-inefficiency placed on the shoulders of the firm hierarchy by the increasing wages. Moreover, governments can institute poli¬ cies that keep wages from rising. This, of course, would also reduce the pressure on firm hierarchies to make their firms more x-efficient. By intro¬ ducing nonwage shelters and keeping wages low, governments would be contributing to the persistence of relatively high levels of x-inefficiency. An important question raised in this type of modeling of the firm is why the firm hierarchy does not choose to organize the firm such that x-efficiency

144

CHAPTER 7

in production is achieved. If this were the choice of utility-maximizing mem¬ bers of the firm hierarchy, identical firms would end up converging in terms of levels of productivity and wage rates. As already discussed, increasing the level of x-efficiency requires that a firm increase wages and improve work¬ ing conditions. In addition, the structure of the firm must be reorganized. This involves flattening its hierarchical structure and typically requires more time and effort inputs on the part of the firm hierarchy, at least in the short run (see chapter 9). Therefore, improving the level of x-efficiency in the firm not only need not result in material gains to members of the firm hierarchy, but may also impose effort costs on them and reduce their power and pres¬ tige within the firm. Increasing the level of x-efFiciency may therefore not be consistent with the utility-maximizing preferences of members of the firm hierarchy. This is especially the case when low levels of x-efficiency are consistently maintaining unit costs and profits at competitive levels. Even if total profits are lower in the x-inefficient firm, this would pose no problem to the firm hierarchy if the disutility of lower profits is less than the utility gained from the reduction in time, effort, and stress.^ Such preferences on the part of the firm hierarchy yield no convergence in terms of productivity or wages to x-efficient levels. Only if the preference function of the firm hierarchy contains an argument from the workers’ preference function where the latter’s utility increases in the high-wage environment would the firm hierarchy on its own volition choose to increase the level of x-efficiency. In this case, interfirm, interregional, and international productivity and wage differentials can be a product of differences in preferences across firm hier¬ archies. In addition, organizations that facilitate the realization of workers’ preferences for a high-wage firm, such as unions, and government policies that have the same affect, such as minimum wage legislation, can also affect the distribution of x-efficient compared to x-inefficient firms. On the other hand, paying lower wages to workers, in the above modeling of the firm, yields no competitive benefits to the firm hierarchy. Shifting to a low-wage firm structure also incurs short-term costs, especially in terms of a reduction in the level of x-efficiency and the increased effort and stress incurred by members of the firm hierarchy when restructuring the firm. This might deter the firm hierarchy from moving to a low-wage firm stmcture, even if they have the opportunity to do so. Given the distribution of preferences across firm hierarchies and the constraints faced by firm hierarchies with regard to choices of firm structures, there is no good reason to expect convergence of x-efficiency levels or wage rates. The thrust of the argument presented thus far is that following from the assumption of identical firms and the absence of product market shelters, labor productivity across firms will be the same given the same degree of x-

DIFFERENCES IN LABOR PRODUCTIVITY

145

inefficiency across firms. The introduction of wage differentials among oth¬ erwise identical firms results in interfirm differences in x-inefficiency. In this manner interfirm differences in labor productivity are introduced. One would expect labor productivity to increase or to be greater in firms where wages are relatively high. The negative impact of wage increases or wage differences upon x-inefficiency is cushioned by the willingness of the firm hierarchy to absorb higher unit costs by lowering profits and by the ability of the higher-wage firm or governments to introduce or increase nonwage shel¬ ters. However, the following proposition should hold true: given the a priori existence of x-inefficiency, ceteris paribus, relative increases in wages should result in relative increases in labor productivity, and wage differentials should result in differentials in labor productivity in favor of the high-wage firms. Alternatively, relative decreases in wages result in relative decreases in labor productivity by inducing increases in the relative levels of x-inefficiency (firm members reduce their effort supplies to levels that yield more utility). Thus relative decreases in wages result in productivity differentials moving in fa¬ vor of the high-wage firms. The argument made for one firm relative to another can be generalized to firms across regions and nations. At the most basic level, simply assume that all firms are everywhere identical (inclusive of the production of the same output) except with respect to labor compensation. However, matters be¬ come complicated once one assumes that all firms producing the same out¬ put are identical but that each region produces a different array of products, with the production process of each product represented by a different pro¬ duction function. In this case, there would be differences in labor productiv¬ ity among regions as a result of the different array of products produced in the different regions and therefore interregional differences in production processes. Nevertheless, one would still expect that interfirm differences in wages in firms producing the same output would be characterized by differ¬ ences in labor productivity and that relative increases in wages among such firms would result in changes in relative labor productivity due to the effect of wages on x-inefficiency. Therefore, conceptually one must distinguish between differences and changes in labor productivity due to the impact of the wage variable on identical firms located in different regions or nations and those that are a function of interregional and international differences or variations in product mix. Given the relative interregional distribution of prod¬ uct mixes, one would expect a relative increase in wages among regions to result in a relative increase in labor productivity in the high-wage regions, as long as within the regions being compared, some of the outputs being pro¬ duced are the same. A related question is with regard to the expected impact of uniform wage

146

CHAPTER 7

increases in one region relative to another when the different regions pro¬ duce different outputs. At the most general conceptional level wage increases can be expected to reduce the level of x-efficiency since otherwise profits must fall. In this scenario, increasing wages are sustainable as long as xinefficiency exists and productivity can be increased sufficiently to match the increase in wages. In this case, identical workers can earn different wage rates in different industries, which would then be characterized by different levels of x-efficiency. As in the previous examples, there is no incentive here for the firm hierarchy to improve the firm’s level of x-efficiency. Such im¬ provements are motivated here by increasing wages rates. These can be con¬ sidered equilibrium wage differentials in the sense that they are consistent with competitive unit costs and profits. The firm hierarchy might prefer the lower-wage regime. But cutting wages will not serve to reduce unit costs or increase profits because of the negative effect such cuts would have on effort inputs. Needless to say, it appears that such pay differentials have persisted over time (Krueger and Summers 1988; Thaler 1992a).^ The argument thus far presented has assumed that only identical firms can produce the same output. I have, therefore, abstracted from the impact that wage differences or a relative change in wages might have upon the capitallabor ratio adopted in firms producing the same output but experiencing dif¬ ferent relative factor prices. Thus I have abstracted from the productivity differences between the high- and low-wage firms that are due to different capital—labor ratios as opposed to different levels of x-inefficiency.^ The ob¬ ject of this exercise in simplification was to isolate the relationship between wages and x-inefficiency. Nevertheless, even when relative wage increases significantly affect a firm’s capital—labor ratio and therefore, its labor pro¬ ductivity, the same wage increases can also be expected to affect labor pro¬ ductivity through pressuring the firm hierarchy to reduce the firm’s x-inefficiency. Therefore, labor productivity will be affected both by changes in a firm’s capital—labor ratio and by changes in its level of x-inefficiency. If a firm simply adjusts its capital—labor ratio in face of a relative increase in the price of labor, unit costs will increase. To maintain unit costs at a com¬ petitive level the firm hierarchy of the high-cost (high-wage) firm must man¬ age a lower level of x-inefficiency (Figure 2.6). Thus to maintain unit costs at competitive levels, the high-wage firm must experience reductions in xinefficiency, unless, of course, one assumes the introduction of a new tech¬ nology particular to the high-wage firm (chapter 2). Simple changes in factor proportions will not suffice. Therefore, my argument that higher wages con¬ tribute to relative reductions in x-inefficiency is not affected by the introduc¬ tion of factor substitution. My argument is more straightforward for firms characterized by isoquants

DIFFERENCES IN LABOR PRODUCTIVITY

147

Figure 7.3

similar to XI in Figure 7.3. This isoquant contains only two feasible input combinations. Thus increases in wages relative to the price of capital need not result in an increased capital-Tabor ratio. For example, an increase in the price of labor, resulting in a movement from isocost line 1 to isocost line 2, does not affect the capital-Tabor ratio, indicated by point b of isoquant XI. In this case, the increased cost of labor results in increased unit costs. The increase in costs is illustrated by isocost line 1 pivoting inward to isocost line 2'—where both isocost lines represent the same level of expenditure—^as factor prices change. Thus isocost line 2 represents increasing expenditures to produce a given level of output as factor prices increase. However, if the high-wage firm is operating x-inefficiently, unit costs can be reduced by reducing the level of x-inefFiciency. Indeed, only at the more x-efficient point c (where the output at point c is equal to the output at XI) will unit costs be the same for the high-wage firm and for the low-wage firm that is operating relatively x-inefficiently at point b on isoquant XI. If, however, the firm operates x-inefficiently at point b due to overstaffing, point b' can be considered as a relatively x-efficient factor mix. This implies a higher capital-labor ratio for the high-wage firm due to the reduction in x-inefficiency brought on by higher relative wages. In this case, the higher capital-labor ratio is not a cause of increased labor productivity; rather it is the increased labor productivity stemming from the reduction in xinefficiency that is the cause of the higher capital-labor ratio in the high-wage firm. If one assumes that the mix of capital and labor given by point b is the

148

CHAPTER 7

only technically optimal one given relative factor prices, then the reduction in x-inefficiency simply results in an increase in total factor productivity (includ¬ ing labor productivity) in the high-wage firms, with the same combination of factor inputs in the high- and low-wage firms. Conclusion An important result of this chapter is that when x-inefficiency exists even in the face of perfectly competitive product markets, wage differentials causing productivity differentials can be expected to persist over time among identi¬ cal firms. These differentials, in a world of x-inefficiency, can be consistent with a unique average cost and therefore with a competitive product market. The relatively high-wage firm need not be relatively uncompetitive. In these circumstances, wage differentials can be a cause of productivity differentials that cannot be accounted for by traditional economic variables. By focusing on the relationship between relative wages and x-inefficiency, we pay close attention to the possibility that labor productivity can be enhanced by mak¬ ing firms more efficient apart from making operations more capital intensive or introducing new technology. Increasing wages in one firm relative to an¬ other, by increasing relative unit costs and prices and/or reducing profits pressures firm members—the firm hierarchy in particular but also workers— to provide a higher level and quality of effort into the process of production so as to reduce the x-inefficiency. On the other hand, relatively low wages reduce the pressure on the firm hierarchy to keep firms relatively x-efficient. I have not attempted to present a case that wage differentials are the key to understanding all or even most labor productivity differentials through their impact upon x-inefficiency. I have attempted only to indicate the conditions under which wage differentials can affect differentials in labor productivity by affecting differentials in x-inefficiency. The overall significance of the connection among wages, x-inefficiency, and labor productivity must, how¬ ever, remain an empirical question. However, it must be emphasized that relatively high wages should not be viewed in isolation as a means of forcing or encouraging increases in the level of x-efificiency. Rather, in a relatively cooperative firm environment higher wages are part of a package of incen¬ tives designed to induce x-efficiency through forms of cooperation between labor and the firm hierarchy (chapter 9). Notes 1. The conventional analysis also attempts to incorporate variables such as unions, tax policy, and government regulation. For an excellent compilation of the traditional approach, see Kendrick, eds. (1984). Refer to Salter (1969) for an analysis of differ-

DIFFERENCES IN LABOR PRODUCTIVITY

149

ences in labor productivity due to the different vintages of best-practice equipment used among firms. 2. That different management techniques can significantly affect labor productiv¬ ity is clearly demonstrated by Kilby (1962, 303-10; 1971, 29-35). One should note that Leibenstein (1966) gives great weight to the former article in his initial formula¬ tion of x-efficiency theory. 3. For an empirically oriented examination of this question, see Altman (1988). See also Shen (1984, 99, 104), who hypothesizes that in developing countries changes in input productivity are dominated by changes in x-inefficiency, whereas in developed countries they are dominated by changes in factor proportions. He argues that a change in the wage level, as one moves from developing to developed countries, is one variable that can explain changing productivity between these two groups. In this chapter, the connection between wages and x-inefficiency is made explicit and conceptionally de¬ veloped in terms of the more general problem of relative wage differences. 4. Leibenstein writes: “It is frequently stated that if the cost of some input, say labor, increases, the cost of the output must increase accordingly. But this does not necessarily follow. If there is a rise in the cost of some input but at the same time the pressure on management to be more effective increases, resulting in more effective effort choices, then this may engender a reduction of x-inefficiency and a decrease in costs” (1981, 104). Leibenstein never further developed this important insight. See also Freeman and Medoff (1984, 12, 15, 95, 101, 105, 163—70, 178—79). Elsewhere, Leibenstein and others argue that wages can affect productivity by affecting the ca¬ loric intake of workers. This argument applies to individuals earning subsistence wages. For a summary of other arguments relating labor productivity to wages, see Akerlof and Yellen (1986; Akerlof and Yellen, eds. 1986; Stiglitz 1987). 5. Leibenstein writes: “It is important to consider that there exists a set of effort/ wage-cost combinations that imply the same profit level. Clearly the firm seeking to maximize profits would be quite happy with a lower, rather than higher, effort/wagecost combination, so as to avoid some of the resistance to the high effort levels ob¬ tained by monitoring and sanctions. In other words, at very high effort levels there is a disutility to the hierarchy to obtaining that effort, and it is easier to live with a lower effort/wage-cost combination” (1987,104). 6. Krueger and Summers (1988) and Thaler (1992a) suggest that efficiency wage theory best explains wage differentials among identical workers employed in differ¬ ent industries. Efficiency wage theory assumes that wages are independently increased by profit-maximizing members of the firm hierarchy so as to minimize average costs by, for example, minimizing turnover and shirking. This raises the question as to why the firm hierarchies in all industries do not raise wage rates to the same extent. Of course, if this were the case, wage differentials among identical workers would not exist. In contrast, the behavioral model presented here allows for profit-maximizing members of the firm hierarchy to choose or be forced to accept, because of unions, for example, from an array of wage rates. In this model, each wage rate, up to the point of diminishing returns to effort, is consistent with competitive average costs and profits. 7. Although Sutcliffe (1971, ch. 5) finds that different capital-labor ratios exist for the same industry, he argues that these are typically few in number and may very well reflect the different products being produced within a specific industrial product clas¬ sification. This suggests that the production isoquant for firms producing similar out¬ put are far from smooth. Rather, the isoquant would at best be characterized by a few alternative capital—labor ratios.

8

The Economics of Profitable Inefficiency and Market Failure A Behavioral Model of Path Dependency Introduction The conventional wisdom’s understanding of long-run equilibrium paths of growth and development, as well as of equilibrium product market develop¬ ment, has been challenged by the pioneering theoretical research of Paul David (1985) and Brian Arthur (1989, 1990) on path dependency.' They argue that in a world of increasing returns to scale there may be a multiplic¬ ity of possible equilibrium solutions to identical economic problems, and the dominant solution can be suboptimal. The prevalent economic outcome in terms of product type, industry, or labor market institutions, for example, can itself be a product of some seemingly inconsequential and random event that, through the process of increasing returns, ultimately gives this outcome a first-mover advantage over other possible outcomes even if the latter hap¬ pen to be more economically efficient. Path dependency theory has been critiqued on various levels. In particu¬ lar, the view that it is possible for an inefficient outcome to persist has been challenged as implausible, if it is at all reasonable to assume that economic agents eventually respond to economic opportunities afforded to them by suboptimal economic outcomes by adopting the relatively more efficient available and known solutions to particular economic problems. In this chapter, a model of path dependency is developed that is grounded in behavioral economics. The argument presented is that it is possible and reasonable to expect there to be a multiplicity of equilibrium solutions to identical economic problems and for the dominant solution to be subopti¬ mal, even under the assumption of diminishing or constant returns to scale. In a world where x-inefficiency is both possible and probable and where productivity and working conditions/labor relations, as well as institutional and cultural parameters, are intimately related, inefficient or suboptimal eco150

PROFITABLE INEFFICIENCY AND MARKET FAILURE

151

nomic outcomes might be the dominant long-run equilibrium outcome to particular economic problems. In contrast to David’s and Arthur’s modeling of path dependency, in the behavioral model presented below, suboptimal outcomes need not provide economic opportunities for economic agents to exploit in the context of their particular objective functions. It is the existence of such opportunities that go unexploited in equilibrium that constitutes the Achilles’ heel of path de¬ pendency theory from the perspective of the conventional wisdom. The be¬ havioral approach to path dependency is therefore able to help address many of the important questions tackled by David and Arthur, such as the macro question of the persistence of relative underdevelopment and the micro ques¬ tion of the survival on the market of inefficiently produced products, while it is not subject to the substantive critiques levied at path dependency theory by the conventional wisdom.^

Path Dependency Theory: Standard Approaches and Criticisms The fundamental argument in the traditional path dependency literature is that the free market typically generates suboptimal long-run equilibrium so¬ lutions to a variety of economic problems, and the probability of suboptimal equilibrium outcomes increases where increasing returns (positive feedbacks) prevail. Increasing returns need not be firm-specific (local). They might very well be industrywide or even economywide (global)—of the variety discussed in some detail by Young (1928)—^thereby bringing into play the role of posi¬ tive externalities. The economic problem might be which product type or standard should be adopted in an economy or which path of development an economy should follow. This argument is couched in a discussion of there being possible multiple equilibrium solutions to identical economic prob¬ lems, with suboptimal solutions among a larger set of solutions. A random shock to an economic system, be it large or small, will have a determining impact on which equilibrium solution becomes dominant, where the domi¬ nant solution can be the suboptimal one. Whichever solution is in effect cho¬ sen by the random event, it might be locked in or become a permanent or stable equilibrium. It is even possible for efficient and inefficient (subopti¬ mal) solutions to prevail simultaneously in the world of path dependency delineated by David and Arthur. For this reason, one cannot expect the free market to force the economy to converge to unique equilibrium* solutions to economic problems. And it follows that in this case it becomes impossible to predict which solution to a particular problem will be adopted, for the cho¬ sen solution ultimately depends on random indeterminate events taken at some critical juncture in the past. More specifically, one cannot predict that

152

CHAPTER 8

the eventual stable equilibrium solution will be the optimal one, even under conditions of competitive markets. That there might be a multiplicity of equilibrium solutions to a variety of economic problems would not necessarily be controversial from the per¬ spective of the conventional economic wisdom. Neoclassical economic theory clearly predicts that a variety of products can exist to meet the needs of utilitymaximizing consumers characterized by different preferences and incomes. It is even possible for chance events to result in the dominance of particular products or systems as long as these are consistent with consumer prefer¬ ences. But in this world, in equilibrium no product or system would be suboptimal or inefficient. In the David-Arthur world of path dependency, however, in equilibrium prevailing products and economic systems might very well be suboptimal or inefficient. David and Arthur argue that increasing returns produce a first-mover advantage to products or economic systems that are chosen first, an advantage that increases over time. It is assumed that productivity and related costs are time-dependent so that newcomers to a product market or to the development process would face a competitive disadvantage, and this would preclude them from beginning the process of catch-up. This would hold true even if the new¬ comers’ productivity would eventually rise above or at least equal that of the first mover and this fact were known to the newcomers.^ In the conventional path dependency story, the first-mover advantage, re¬ inforced by increasing returns and externalities, becomes a permanent road¬ block to newcomers. The obvious economic superiority of alternative equilibrium solutions that are known to economic agents does not generate the economic forces expected by the conventional wisdom to challenge and eventually displace the inefficient economic regimes—it becomes too costly for private economic agents to adopt the superior economic regimes. This argument is illustrated in Figure 8.1, where, ceteris paribus, average cost is assumed to be a negative function of historical time. Curves 1 and 2 illustrate a scenario where increasing returns to time eventually level off. Curve 1 represents the initial and suboptimal economic regime. Curve 2 represents the optimal, least-cost economic regime. At time r, the average cost of eco¬ nomic regime 1 is OC. If the economic regime represented by curve 2 were in place for the same period of time (Tq/,), its costs would be even lower, at OD. However, if economic regime 2 comes on line at only after economic regime 1 has been in place for y,, regime 2 faces an initial competitive disadvantage of5C. This disadvantage would be eliminated at time t* if the suboptimal regime is given by curve 1. The persistence of inefficiency is given by the dominance over time of the suboptimal economic regime represented by curve 1. This persistence would

PROFITABLE INEFFICIENCY AND MARKET FAILURE

153

Figure 8.1

represent a market failure. And what is clearly suggested by the David-Arthur paradigm is the likelihood of a free market dominated by such market fail¬ ures in spite of the clear and manifest superiority of existing and known alternatives, such as represented by curve 2 as compared to curve 1. Once locked in, the inefficient solutions to economic problems cannot be displaced by market forces alone. Society becomes a prisoner to the inefficiencies es¬ tablished through first-mover advantage or increasing returns, while exter¬ nalities and the absence of perfect future markets play a crucial role in locking in the suboptimal equilibria in a world of increasing returns. Presumably, in a world of no externalities and perfect futures markets economic agents would be able to take advantage of the economic opportunities afforded by alterna¬ tive and more efficient products or systems. That the market cannot eventually displace the type of inefficiencies elabo¬ rated upon in the conventional path dependency story, given the incentives to economic agents to do so in terms of the unexhausted gains from trade, has been subject to severe criticism. The most poignant of these critiques have been enunciated by Liebowitz and Margolis (1990, 1994). They have system-

154

CHAPTER 8

atically challenged the empirics underlying David’s and Arthur’s theoretical cases supporting inefficient equilibria. Moreover, they have raised questions with regard to the theory of path dependency. Most generally, they argue that differences in efficiency among standards or products should generate eco¬ nomic opportunities that would be typically taken advantage of, ultimately causing the elimination of inefficient standards even in a world of imperfect futures markets. One should not simply assume that economic agents will not exploit known potential gains from trade or that the marginal benefits from shifting to a superior standard rarely if ever outweigh the marginal costs. Liebowitz and Margolis argue that the David-Arthur modeling of the world ignores the role of entrepreneurship, which involves risk taking in the expec¬ tation of profits in an uncertain future. They maintain that it is the expecta¬ tion of future, albeit uncertain, profits by risk-taking entrepreneurs that has driven the adoption and development of superior products and systems that have occurred over time. In Figure 8.1, the profit opportunities are given by the difference in cost between regimes 1 and 2 after regime 1 has been in place for more than of time. After this point, regime 2 captures the eco¬ nomic rents determined by this difference so long as price is determined by the marginal suboptimal economic regime. After an uncertain and unpredict¬ able time a new equilibrium price is established consistent with the costs of the optimal economic regime. This is the classic Schumpeterian process of technical change (Schumpeter 1974, ch. 4). For the optimal economic re¬ gime to be chosen, the anticipated losses, given by B*FH, must be exceeded by the anticipated but uncertain rents. Liebowitz and Margolis conclude that “A transition to a standard or technology that offers benefits greater than costs will constitute a profit opportunity for entrepreneurial activities that can arrange the transition and appropriate some of the benefits. . . . Eeonomies do, in fact, move from one state to another. This is not to say that mis¬ takes are never made, in markets or elsewhere. But we do have overwhelming evidence that markets do make transitions to superior products and stan¬ dards—from horses and buggies to automobiles, from typewriters to com¬ puters, from mail to fax” (1994, 146). Central to the Liebowitz and Margolis critique is that the predictions of the conventional path dependency theory are faulty largely because David and Arthur assume that economic agents typically will not or cannot take advan¬ tage of known economic opportunities and that this results in the prevalence of market failure in any given time.'^ It is assumed that known gains from trade exist in the form of superior products and standards that economic agents fail to take advantage of even over the long term. But is this particular assumption critical to the hypothesis embedded in path dependency theory that suboptimal economic systems and products can persist over time after being adopted for

PROFITABLE INEFFICIENCY AND MARKET FAILURE

155

whatever reason? Is it possible for superior standards or products to exist with¬ out there being the erstwhile gains from trade to attract economic agents to move toward optimal equilibrium solutions? A Behavioral Approach to Path Dependency The David—Arthur configuration of path dependency theory assumes, along with conventional economic wisdom, that effort is not a discretionary vari¬ able and that the quantity and quality of effort per unit of labor input are maximized at any given time. Since effort inputs into the production process are assumed to be maximized, variations in productivity are independent of variations of effort inputs, and differentials in labor productivity are in turn independent of differentials in effort input. If effort discretion exists, labor productivity is affected by the quantity and quality of effort inputted into the process of production per unit of time. In this case, the traditional production function, where output is a function of labor, capital, and technology, is aug¬ mented by the quantity and quality of effort inputted. What would be the implications for path dependency theory if ones assumes no externalities and constant returns or constant cost industries while at the same time as¬ suming the existence of effort variability—that effort discretion exists—due, for example, to both informal and formal contracts being incomplete as a consequence of the transaction costs involved (Akerlof and Yellen 1990; Altman 1996a, 1997; Miller 1992; Stiglitz 1987)? The introduction of effort discretion into the modeling of the economic agent and path dependency theory allows for the existence and persistence of multiple suboptimal equilibria and allows for the possibility and helps to ex¬ plain the existence and development of inefficient economic regimes even un¬ der severe competitive pressures. In a nutshell, in a world of effort discretion private economic incentives need not exist for economic agents to adopt supe¬ rior economic regimes and for the inferior, suboptimal regimes to be displaced. Under these circumstances, market failures can result that are consistent with the constrained utility maximization of rational optimizing economic agents, thereby reducing a society’s level of material welfare from what it might other¬ wise be. Assuming no externalities, constant returns and optimizing economic agents bias my results in favor of the conventional wisdom and allow for the isolation of effort discretion as a causal variable in the modeling of path depen¬ dency. Introducing externalities, increasing returns, and nonoptimizing eco¬ nomic agents, which characterize the David—Arthur world, would only strengthen the case in favor of path dependency since, as discussed above, these characteristics provide some protection to the suboptimal economic re¬ gimes. In other words, the case is made that even in a theoretical world favored

156

CHAPTER 8

by the conventional economic wisdom persistent inefficiencies that are path dependent are possible and might even be pervasive. Given effort discretion, the quantity and quality of effort supplied on the job can be affected by important variables such as the organization of the firm, inclusive of the structure and level of wage rates; working conditions; the state of competitive pressures; and an individual’s, community’s, or society’s work culture (chapter 9). Under these conditions, workers would not automatically or mechanically maximize their effort inputs into the pro¬ duction process, nor could members of the firm hierarchy easily or mechani¬ cally induce workers into maximizing effort per unit of time and thereby output per unit of labor. Indeed, they themselves do not necessarily maxi¬ mize the quantity and quality of effort that they supply to the firm. Their behavior would be more in line with utility maximization as opposed to profit maximization, where their objective includes arguments other than profit. Furthermore, in a realistically modeled world of effort variability, there is no rational reason to expect utility-maximizing workers to choose to work as hard and as well as they can in an economy characterized by noncoopera¬ tive, if not outright antagonistic, industrial relations, where they are treated poorly and unfairly. In addition, utility-maximizing members of the firm hi¬ erarchy may prefer to live with or even develop more antagonistic industrial relations if such an environment is consistent with their objective functions. This holds true even if it means that labor productivity and even total factor productivity are lower than they would otherwise be.^ Why would “rational” members of the firm hierarchy not pursue policies designed to maximize productivity? Higher productivity that is a product of more and higher quality of effort inputs requires a more highly paid labor force and often the investment by members of the firm hierarchy of more time, effort, and money to reorganize the effort inputs to facilitate higher levels of productivity. These represent investments in organizational capital (Tomer 1987) that are largely the start-up costs of establishing a different organizational environment. Moreover, it is also possible that members of the firm hierarchy would suffer a lower income if a more cooperative indus¬ trial relations regime were part of the package that ultimately generated increased productivity. In addition, when all is said and done, in a highproductivity regime unit costs need not be lower and profits need not be higher than in a low-productivity regime since the higher labor productivity is gen¬ erated by increased expenditures by the firm. Under these conditions, there are no definitive incentives for members of the firm hierarchy to develop a higher productivity work environment. Unless their utility function con¬ tains arguments related to improving the material and psychological well¬ being of workers, it would be quite rational for members of the firm

PROFITABLE INEFFICIENCY AND MARKET FAILURE

157

hierarchy to maximize their utility in a low-wage environment, increasing their earnings by redistributing income away from labor into their own hands. In this case, the firm would be producing less. It would realize a lower level of labor productivity than it would under a more conducive system of indus¬ trial relations. In effect, it would be producing below potential or x-inefficiently (chapter 9).^ In the long run, under competitive product market conditions, firms can produce x-inefficiently if doing so does not threaten the survival of the firm by raising unit costs above or reducing profits below competitive levels (chap¬ ters 1 and 2). For this reason, the x-inefficient firm must keep input costs relatively low in order to compensate for its relatively low level of productiv¬ ity. Keep in mind that average costs (AC) are a product of the weighted input costs deflated by the weighted average productivity of factor inputs. In a simple world with labor (L) as the only factor input and the wage rate (w) as the only factor price, this translates into:

(8.1)

where Q is real output. In this model, the wage rate also serves as a proxy for a particular system of industrial relations, where a low wage is a proxy for a work environment that is antagonistic and nonparticipatory with little invest¬ ment in organizational capital, and a high wage is a proxy for the opposite. One significant way of maintaining low unit costs is to keep the rates of labor compensation relatively low, while low rates of labor compensation keep pro¬ ductivity at a relatively low level. For this reason, low rates of labor compensa¬ tion protect the x-inefficient firm from the relatively more productive firms just as imperfect product markets, tariffs, or subsidies would. So long as the xinefficient firm can be afforded such protection, it can be viable over the long haul even in the face of severe competitive pressures. And x-inefficient levels of production would constitute one option to utility-maximizing members of the firm hierarchy. On the other hand, relatively high-wage firms could not survive on a competitive market unless their relatively high input costs were compensated by higher levels of labor productivity. The higher productivity would allow them to remain competitive even in the face of severe competitive pressures. In fact, higher wages and a higher wage environment serve to in¬ duce higher levels of labor productivity as a means of keeping the firm com¬ petitive.^ In this case, members of the firm hierarchy would be facing the constraint of relatively high wages when maximizing their utility. As long as productivity rises and falls sufficiently with movements in the

158

CHAPTER 8

level of labor compensation, there is no reason for increasing labor benefits to generate higher production costs or for lower benefits to yield lower costs. It is, in fact, possible for there to be a range of labor benefits associated with a unique unit cost of production—changes in productivity would just com¬ pensate for changes in labor benefits. This assumes that there is no unique wage rate or labor compensation package that will minimize unit production costs, at least over a significant range of wage rates (chapters 1 and 2).^ In other words, two identical plants producing the same output can produce at the same unit costs, even when rates of labor compensation are relatively higher in one plant, if labor productivity is sufficiently higher in the highwage plant. However, it is also possible for low-wage plants to produce at a lower unit cost than high-wage plants if the labor productivity differential between the two does not quite compensate for the wage differential. In this case, it pays the utility-maximizing members of the firm hierarchy to pro¬ duce in the low-wage plants, and low-wage plants would dominate the mar¬ ket in long-run equilibrium in spite of their being x-inefficient. In a simple behavioral model of the firm one assumes that labor and capi¬ tal inputs as well as technical change are constant and that wages and effort per unit of labor input vary. Effort per unit of labor input is assumed to be positively related to the wage rate:

£l

/(w),

Lt

(8.2)

where e is effort and t is time. This in turn causes variation in labor productivity;

L

(8.3)

\ L

From Equation 8.1, since average cost is given by wl{Q/L), variations in labor productivity can affect average cost. The average cost equation can be denoted by w/q when Q/L is given by q. The inverse of w/q yields the elastic¬ ity of labor productivity relative to changes in the rate of labor compensa¬ tion, where this elasticity r\ is given by: dq

w

(8-4) Only when r\ is greater than one does an increase in wages—that is, an im¬ provement in the system of industrial relations—^yield a decrease in average

PROFITABLE INEFFICIENCY AND MARKET FAILURE

159

cost through its impact on effort and hence upon labor productivity. When r\ is less than one average costs rise with increasing wages. Finally, an riof one yields no change in average cost as wages rise and at this point average cost is at a minimum. Conventional neoclassical theory assumes an tt of zero since changing the wage rate of the work environment is assumed to have no effect on effort inputs and therefore upon labor productivity. On the other hand, the wellknown result from efficiency wage literature is that there is a unique wage rate—^the efficiency wage—that minimizes unit costs by rather mechanically maximizing effort in a world in which effort is variable. In this case, it is assumed that p is one for a unique value of w. In effect, the wage rate be¬ comes inflexible given the assumption that the logistic production function best reflects the reality of the firm. In contrast, in the behavioral model of this chapter, there is no unique wage, at least over a range of wage rates, that will yield a unique cost-minimizing level of effort. Therefore, in the behav¬ ioral model p is one for a range of wage rates and there is no unique point of cost minimization over this range. This argument is easily illustrated, borrowing from Stiglitz’s (1987) treat¬ ment of efficiency wages. In Figure 8.2, the efficiency wage is given uniquely by W*, assuming a U-shaped average cost curve, relating changes in average costs as the wage rate varies. The average cost curve is in turn derived from a labor—wage productivity curve that takes the form of a logistic function (Figure 8.3). In this case, average productivity is maximized (average cost is minimized) at e, where the wage is IF*. At this point p is one. To the left of e, p exceeds one and, to its right, p is less than one. The logistic productivity function yields a unique wage productivity—average cost effort set that is invariably and ultimately chosen by the firm. But the functional form of the productivity curve underlying such a unique set is not the only one possible, the form is embedded in the assumptions one makes about changes in effort relative to changes in wages. In the behavioral model, one assumes that some linearity characterizes the production function—for example, from 0 to e, in Figure 8.3. Along Oe the elasticity of productivity to wages is unity. This in turn generates an L-shaped average cost curve in Figure 8.2, with C*B be¬ coming a component of curve 1. In this case, over a certain range of wage rates, there is no unique wage rate. A range of wage rates or systems of industrial relations is consistent with a unique level of average cost such as OC* in Figure 8.2. There is no efficiency wage per se. Over this range, changes in wages are just compensated by changes in labor productivity. Moreover, over this space there is an array of firms, spanning from the most x-ineffiicient (low-wage) to the most x-efficient (high-wage), all of which are cost competitive and therefore economically viable.

160

CHAPTER 8

Figure 8.2

Figure 8.3

PROFITABLE INEFFICIENCY AND MARKET FAILURE

161

How would the explicit introduction of technological change affect the arguments presented above? At the most basic analytical level, according to the conventional neoclassical wisdom, new technology should dominate old technology if it can produce a particular product at a lower cost, holding the quality of the product constant. This argument should hold where factor prices are identical across all firms and where x-inefficiency does not exist or where the level of x-inefficiency is the same for all firms. In Figure 8.1, if curve 2 represents the new technology, this technology should eventually dominate the old technology represented by curve 1 for reasons of lower costs. In Figure 8.4, where capital and labor inputs are mapped out along the vertical and horizontal axes respectively, technological change can be illustrated by a shift inward of the production isoquant from to where the level of output produced by both isoquants is identical. If the initial equilibrium is given by point A along and the new equilibrium is given by point D along —relative factor prices have not changed—^the new technology should dominate the old in terms of unit production costs as the isocost curve shifts inward from BC to point D. However, this type of scenario need not be the only reasonable representation of economic reality. In the tradition of Leibenstein (1973b) one can assume that the pace of technological change is affected by the extent to which x-inefficiency and therefore effort discretion exist and can be expected to persist in the firm. If one begins with a scenario wherein the new technology is exogenously in¬ troduced into an environment where x-inefficiency characterizes all firms in the economy, the new technology might not be viable if the level of x-inefficiency prevents it from realizing its potential. In other words, if the new technology corresponds to a relatively high level of x-inefficieney such that the production isoquant remains at Q^, unit production costs associated with the new technology will not be less than those associated with the old tech¬ nology, and the new technology will not dominate the old even though it is potentially superior in terms of unit costs. With no x-inefficiency, techno¬ logical change would shift the production isoquant to , and the new tech¬ nology would be expected to dominate the old. If reducing the level of x-inefficiency requires changing and investing in the organizational capital of the firm—illustrated, for example, by a pivot in the isocost curve from BC to BC"—unit production costs for the old and the new technology will be identical. Producing at point A along Qq yields the same unit costs as producing at point A' along where represents the new technology that is now incorporated into a relatively x-efficient firm. The high-tech firm fails to dominate the low-tech firm even though it in¬ creases the economy’s per capita output. Only if the new technology, inclu¬ sive of its embodied level of x-inefficiency, serves to reduce unit costs, as

162

CHAPTER 8

Figure 8.4

p, c a

& u

0

C'"

C

Labor Inputs

can be illustrated by isoquant would the new technology dominate the old. In this case, the lower costs are illustrated by the shift in the isocost curve from BC" to B'C" (chapter 6; Altman 1998). As opposed to exogenously determined technological change, one might assume that technological change, as well as corresponding reductions in the level of x-ineflficiency, is endogenously determined and is a costly process

PROFITABLE INEFFICIENCY AND MARKET FAILURE

163

induced by changes in factor prices (chapters 2, 6; Altman 1998; Habbakkuk 1962; Ruttan 1997). Assume that the change in factor prices is once again illustrated by a pivot in the isocost curve from BC to BC". If the new technol¬ ogy, inclusive of the residual level of embodied x-inefficiency, is given by and the old by Q^, the new technology will once again fail to dominate the old since unit production costs at point A would be the same as at point A'. These are given by isocost curves BC and BC". In this case, for technological change along with the necessary reduction in the level of x-inefficiency to transpire requires an investment in the firm. Technological change is here a costly pro¬ cess. Without technological change cost-minimizing production would take place at point ^1* along Qq, generating higher unit production costs. Only if the change in factor prices induces technical change and reductions in the level of x-inefficiency such that the isoquant moves below to Q2, for example, will the new technology dominate the old. In this sense, the old technology re¬ mains competitive if it is path dependent on a low-wage system of industrial organization while the new technology is path dependent on a high-wage sys¬ tem of industrial organization and, given these constraints, the new technology does not yield lower unit costs than the old technology. Even in a world with no x-inefficiency, a movement from points to points' will not result in the dominance of the new technology if technological change is a costly process since unit costs at point vl would be the same as at points'. How does the introduction of effort discretion and induced technical change impact upon path dependency theory? Without deviating from the conven¬ tional assumptions of (constrained) utility-maximizing individuals and longrun competitive product markets, the introduction of effort discretion into the modeling of the economic agent allows for the existence of path depen¬ dent high-and low-productivity firms producing identical products in longrun equilibrium. At a more general level, it also allows for the existence of path dependent high-and low-productivity economies in long-run equilib¬ rium. This is true even without the assumption of increasing returns and externalities that provide some initial advantage to the suboptimal economic regimes. To generalize further, there can exist in long-run competitive equi¬ librium an array of firms or economies characterized by an array of productivities, producing at identical unit production costs, with only one component of the array of firms or economies being efficient. This is so for the simple reason that when productivity is positively correlated with labor compensation packages that in turn encompass the industrial relations envi¬ ronment in which production takes place, different levels of productivity need not be associated with different levels of unit cost or profit. Under these circumstances, market forces cannot easily eliminate the low-productivity economic entities when they are no less competitive than their high-produc-

164

CHAPTER 8

tivity counterparts. This is true even if one introduces into the argument tech¬ nological change where technological change is a costly process. The crux of the Liebowitz and Margolis critique of path dependency theory’s prediction of the persistence of inefficient economic outcomes is that the existence of known, relatively efficient alternatives should be ex¬ pected to trigger an entrepreneurial response that would take advantage of them. Their critique does not hold when the more efficient, higher-produc¬ tivity regime does not carry with it lower unit costs and higher profits and when it is inconsistent with the utility function of members of the firm hier¬ archy. In this case, it would be possible for both efficient and inefficient regimes to exist simultaneously in long-run equilibrium. The economic opportunities that might trigger the elimination of ineffi¬ cient economic systems or products need not exist when economic agents behave in a fashion consistent with effort discretion or when induced techni¬ cal change is a costly process. In fact, it would be possible for the inefficient (low-productivity/x-inefficient) system to dominate if the relatively low la¬ bor costs yield low unit production costs for an identical product. In addi¬ tion, even if the x-inefficient system yields the same unit costs as the more efficient system, the former might dominate if it is consistent with the prefer¬ ences of members of the firm hierarchy since they have traditionally deter¬ mined which path an economic regime should follow. If the efficient system generates relatively higher unit costs but also produces an output of a higher quality, the lower unit cost system need not dominate. In this case, however, one would no longer be modeling identical outputs. The different economic regimes could also be a product of past random events and would then be path dependent. One could not easily predict the convergence of the economy toward the efficient equilibrium. The modeling of the economic agent pre¬ sented in this chapter is therefore able to contribute toward addressing the critical question asked by Liebowitz and Margolis of path dependency theory: why should we expect inefficient economic regimes to persist over time? Of course, if productivity differentials between two systems generate lower unit costs and higher profits to the high-productivity system the critique of Liebowitz and Margolis kicks in, and market forces might very well, as they argue, displace the inefficient with the efficient economic system.^ The Persistence of Inefficiency: Some Examples The possibility, elaborated upon in this chapter, that inefficient economic regimes can dominate or exist simultaneously with relatively efficient re¬ gimes because the more efficient regimes do not provide the mechanism to displace them—a mechanism that clearly exists in the David and Arthur

PROFITABLE INEFFICIENCY AND MARKET FAILURE

165

modeling of path dependency in terms of known and unexploited economic rents—Phelps to address a variety of important apparent paradoxes of eco¬ nomic life such as the persistence of inefficient or low-productivity eco¬ nomic regimes.’® Douglass North, for example, has made the case that through world history inefficient economic regimes have dominated: I will approximate the conditions in many Third World countries today as well as those that have characterized much of the world’s economic his¬ tory. The opportunities for political and economic entrepreneurs are still a mixed bag, but they overwhelmingly favor activities that promote redis¬ tributive rather than productive activities, that create monopolies rather than competitive conditions, and that restrict opportunities rather than ex¬ pand them. The organizations that develop in this institutional framework will become more efficient—but more efficient at making the society even more unproductive and the basic institutional stmcture even less condu¬ cive to productive activity. Such a path can persist because the transaction costs of the political and economic markets of those economies together with the subjective models of the actors do not lead them to move incre¬ mentally toward more efficient outcomes. (1990, 9)” But can low-productivity regimes survive in the face of competitive pres¬ sures? They can, but only under particular circumstances. Certainly unproduc¬ tive economic regimes can survive if well protected from competitive pressures. Such protection was more easily afforded in the past than in the present. How¬ ever, in face of competitive pressures, unproductive economic regimes will have a difficult time of it if relatively low productivity translates into relatively high production costs. This need not be the case when low productivity is balanced by low rates of labor compensation or low rates of investment in organizational capital. And, as outlined above, the low rates of labor compen¬ sation and low-wage enviromnent bundle in themselves can be a key cause of low productivity. Members of a firm hierarchy or more generally of the economic hierarchy, can fare quite well under these conditions since their economic ven¬ tures remain competitive and their own incomes can remain high through re¬ distributive as opposed to productive activities. The existence of low- and high-productivity regimes might be a product of history, of random or deliber¬ ate actions taken sometime in the past. Once a particular path to development is taken, an economy might even be locked into it due to the high costs of transition—a point emphasized by North. But lock-in can be characterized by a certain degree of stability only when a particular economic regime remains competitive. The behavioral modeling of the economic agent presented above is suggestive of important mechanisms that might account for multiple types of economic regimes in long-mn equilibrium.

166

CHAPTER 8

Another intriguing and important paradox of economic life is related to labor market structures. Of particular interest is the survival into the nineteenth century of slavery in the United States and serfdom throughout most of East¬ ern Europe as forms of labor organization that paralleled free labor. Does their survival suggest that these were efficient forms of labor organization that were close substitutes for free labor and that all three general forms contributed to equally efficient systems of production?’^ Such an argument would follow from the conventional assumptions about the economic agent and the assump¬ tion that relatively inefficient regimes generate the incentives for their even¬ tual displacement by relatively more efficient regimes. An alternative argument would be that unfree labor was relatively inefficient. But if it was inefficient, how could it persist over time, practically speaking for centuries on end? Should it not have been displaced by relatively more efficient systems of free labor? The empirical debate on the subject is far from resolved.’^ In terms of a behavioral modeling of the economic agent, one cannot pre¬ dict the disappearance of slavery or serfdom in competition with free labor even if free labor is relatively more productive. The self-interested landlord or slaveowner will refuse to institute a free labor regime unless the productivity differential in favor of free labor exceeds the cost of the labor differential in favor of coerced labor. Only under these circumstances would the unit costs of using free labor be lower than the unit costs of using unfree labor. If unit costs under both regimes are equivalent, the profit-maximizing landlord would be indifferent between using free and unffee labor, and both regimes could exist simultaneously. In addition, ceteris paribus, unfree labor remains viable in face of free labor becoming increasingly productive if the income of the unfree labor can be further depressed without an attendant fall in labor productivity. Suboptimal or inefficient regimes of labor organization can persist over time or even dominate an economy as long as they remain cost competitive and privately profitable for members of the economic hierarchy. Conclusion Scholars arguing on either side of the path dependency debate agree that producing at a relatively low level, ceteris paribus, is indicative of a market failure, where the opportunity cost to society is the loss of output. What has been subject to debate is the ability of the market to correct for such failures within a reasonable time, even under competitive conditions. Modeling the economic agent in terms of more realistic behavioral assumptions allows for the persistence of the inefficient economic regimes, on either a micro or macro level, even in a highly competitive environment. Unlike in the DavidArthur modeling of path dependency, where inefficient paths can persist as a

PROFITABLE INEFFICIENCY AND MARKET FAILURE

167

consequence of the inability of economic agents to take advantage of avail¬ able and known economic opportunities, in the model presented above there need be no economic benefits to be gained by private economic agents from shifting from inefficient to efficient economic regimes. Moreover, both effi¬ cient and inefficient economic regimes can be cost competitive, and ineffi¬ ciencies in production can be consistent with constrained utility maximization on the part of economic agents. In these circumstances, there is no easy mecha¬ nism to direct economic agents from inefficient to efficient paths. This re¬ sults in a market failure. Ceteris paribus, society would be better off if the economy produced x-efficiently as this would increase the standard of mate¬ rial well-being of most individuals. In addition, x-efficient production could be achieved by modifying the incentive system in the workplace. However, if the members of the firm or economic hierarchy prefer the inefficient eco¬ nomic regime and this utility-maximizing preference does not threaten the com¬ petitiveness of the firm, the economy will traverse along the inefficient path, and such a path will continue to be chosen since it is the members of the economic hierarchy who ultimately decide which path to take.'^ Whether or not an economy moves along an inefficient path depends on the preferences of members of the firm hierarchy given the constraints that they face and the economic viability of choosing the inefficient path. The behav¬ ioral model of path dependency does not predict that inefficient or suboptimal solutions to economic problems are inevitable but, rather, that inefficient eco¬ nomic regimes can be viable and, if so, will be chosen by members of the firm hierarchy when they are most consistent with their preferences. Once locked into the inefficient path, there is no good economic reason to expect the economy to break out of it, especially if there are economic or psychic costs involved in a transition to a new, more efficient path. Institutional factors can therefore play an important role in affecting the capacity of the market to motivate eco¬ nomic agents to choose the relatively more efficient path or, once on the inef¬ ficient path, to shift to the more efficient path. On the other hand, they can also play a determining role in encouraging economic agents to choose the ineffi¬ cient path. Therefore, it is not inevitable that economic agents will choose optimal solutions to particular economic problems. Chance events as well as carefully crafted policy can affect whether society takes the efficient or ineffi¬ cient route and whether it locks in or breaks out of one particular equilibrium. A behavioral modeling of path dependency suggests that one cannot predict that market forces in and of themselves will consistently generate efficient solutions to important economic problems. Moreover, one cannot predict which equilibrium solution from a subset of solutions economic agents will choose since they are not necessarily constrained by the market, even a highly com¬ petitive one, into choosing the most efficient of the available solutions.

168

CHAPTER 8

Notes 1. On a related argument with respect to the process of economic development, see Krugman (1991). 2. An interesting perspective on the persistence of inefficiency that is related to the path dependency literature is presented by George (1989). In his modeling, ineffi¬ cient subsystems persist when they are nested within larger relatively efficient sys¬ tems. See also Boyer (1997) and Boyer and Hollingsworth (1997), who also discuss persistent differences in economic and social regimes and the need to better explain their concurrent viability. 3. If increasing returns existed without bound, then the first mover would have a permanent advantage over any and all newcomers unless unit costs fell at a faster pace in the newcomer’s than in the first mover’s plant. In the former case, however, one cannot make the case that the newcomers are more efficient than the first movers. However, what characterizes path dependency theory is the assumption that the newcomer’s product, firm, or industry is initially the relatively more efficient one. 4. See Witt (1997) for an elaborate and more technical critique of the Arthur— David modeling. 5. See Altman (1996a, 1998) for more detailed discussion of the behavioral model of the economic agent that underlies the model of path dependency presented in this chapter. See Stiglitz (1987) and Akerlof and Yellen (1990) for a related discussion of efficiency wage theory. 6. Leibenstein (1966, 1987) coined the term x-inefficiency to describe a level of output that fell below the neoclassical ideal, where all economic agents are maximiz¬ ing effort per unit labor input (see also Frantz 1988; Dean and Perlman 1998). Leibenstein believed that maximum labor productivity (x-efficiency) could possibly be achieved only under conditions of an ideal system of industrial relations. X-efficiency could not be achieved by simply writing an ideal contract and then enforcing the terms of this contract. For a similar argument focusing on a game theoretic ap¬ proach to production inefficiencies, see Miller (1992). 7. This argument is elaborated upon in, for example, chapters 1, 2, and 9; Card and Krueger (1995); Levine and Tyson (1990); and Freeman and Medoff (1984). 8. A standard argument in the efficiency wage literature is that although effort is a discretionary variable with respect to wages, there exists a unique wage rate that will serve to maximize effort per unit of labor, thereby maximizing labor productivity and thus minimizing unit costs. This is the wage rate that efficiency wage theory predicts will be selected by a profit-maximizing firm. Any deviation from this unique wage rate would serve only to increase unit production costs. For the classic state¬ ment on this, see Solow (1986). Refer also to a detailed discussion in Stiglitz (1987). 9. In chapters 2 and 6, the significance of relatively low wage rates and x-inefficiency to the persistence of suboptimal, low-productivity technologies is discussed. 10. In chapter 6 and Altman (1998), I discuss the conditions for multiple long-run equilibrium wage and productivity paths of economic growth using the type of behav¬ ioral modeling of the economic agent presented above. 11. On this point, see also Lane (1958, 1975). It is important to note that North’s point is empirically validated by the time series evidence on the absence of conver¬ gence in terms of real per capita GDP in the world economy. See Altman (1999c) for a summary discussion of the relevant literature.

PROFITABLE INEFFICIENCY AND MARKET FAILURE

169

12. A very basic definition of a slave is an individual who by law is a chattel or property of his or her owner. The serf, on the other hand, is legally free. The serf was bound to his or her “master” by institutional as opposed to contractual ties, which made the serf into a subset of unfree labor. As with the slave, the relative freedom of the serf varied over time and space. 13. The debate on the relative efficiency of American slavery has been particularly heated. Two critical articles on the subject are by Fogel and Engerman (1971) and G. Wright (1975). 14. Domar (1970) provides considerable theoretical insight into the economics of free versus unfree labor systems. He argues that unfree labor becomes the norm when landlords are no longer able to earn a rent from their land on the free market—^when the income of the free peasant is driven up to the point where the only rents earned are on land of superior quality—^and when government provides the landlord with the legal and political wherewithal to create a class of unfree labor. Domar argues that the use of unfree labor would dominate even if free labor were relatively more productive if the productivity differential was outweighed by the relatively higher costs of free labor. Unlike in the behavioral modeling of the economic agent, Domar assumes that productivity is independent of the wage rate—^there is no effort discretion, and wages and working conditions have no effect on labor productivity. For this reason, he ar¬ gues that reducing wages is central to making free labor relatively more attractive to the landlord since a fall in the wage rate has no predictable impact on labor productiv¬ ity. Labor productivity is assumed to be a positive function of the extent to which labor is legally free, irrespective of the working conditions of such labor. 15. See Ichniowski et al. (1996) for a detailed accounting of the literature discuss¬ ing relatively innovative, productivity-enhancing, alternative modes of workplace or¬ ganization that consistently fail to be adopted for material and utility cost consider¬ ations. See also chapter 9.

9

Economic Theory, Public Policy, and the Challenge of Innovative Work Practices

Introduction A fundamental finding of the current empirical industrial relations (IR) and human resource management (HRM) research is that similar types of firms producing similar types of products adopt different sets of work practices or cultures. Among the different available and known sets of work practices or IR systems are those that yield superior levels of output and quality as com¬ pared to alternative IR systems. This is true even when the alternative sys¬ tems are part and parcel of the same external economic and institutional environment. These different types of work cultures exist simultaneously and persist over time. The literature suggests that the higher quantity—quality of work cultures are in some sense superior because of their relatively higher yields. This in turn implies that they yield, in some meaningful sense of the term, a higher level of welfare. These work cultures are associated with a bundle of charac¬ teristics that include more active employee participation in the functioning of the firm, more labor-mianagement cooperation associated with a less hier¬ archical system of management, labor compensation related to labor produc¬ tivity, and a relatively long-term employment relationship with the firm (Alcaly 1997; Appelbaum and Batt 1994; Barney 1995; Becker and Huselid 1998; Buchele and Christiansen 1999; Gordon 1996; Ichniowski et al. 1996; Kochan, Katz, and McKersie 1986; Levine 1995; Levine and Tyson 1990; Logue and Yates 1999; Neal and Tromley 1995; Pfeffer 1995; Tomer 1999; Winther and Marens 1997). These broadly drawn stylized facts of the inner workings 170

WORK PRACTICES

171

of the firm suggest that superior systems of work practices are rejected by the majority of firms and that the majority of firms are able to successfully compete in the marketplace using inferior sets of work practices. In effect, the majority appear to be both performing inefficiently and surviving the test of the market, and there appears to be little convergence of IR systems.* The notion that economic agents systematically reject superior inputs into the production process, of which the work culture is 1, stands in sharp con¬ trast to the prediction of neoclassical theory that flows from its basic premise that all static and dynamic gains from trade are exhausted by utility-maxi¬ mizing and rational economic agents. In more colloquial terms, this predic¬ tion simply follows from the American Question and the Axiom of Modest Greed (McCloskey 1990, 112; see ch. 1 above). If superior work cultures exist and they are known to generate economically significant static and dynamic increases in productivity and quality—$500 bills as opposed to quarters—^why would even mildly rational individuals slide over such an easy mark? Do the findings of the IR and HRM literature pose a dilemma and paradox to neoclassical theory and indeed for public policy? Or upon closer inspection, would the data on diverse work cultures reveal that eco¬ nomic agents are foregoing quarters and not $500 bills when adopting or maintaining the more traditional work practices, thereby confirming the sus¬ picions of the conventional wisdom that individuals do not, as a matter of course, forego economic opportunities of any consequence? Indeed, the con¬ ventional wisdom would strongly suggest that what appear to be superior inputs are not and what appears to be inefficient firms are not once we take into consideration the specific arguments that comprise the objective func¬ tions of the economic agents of the firm, as well as their objective con¬ straints. This follows directly from a basic premise of neoclassical theory that utility maximization entails that all firms in a competitive environment converge toward maximizing output given inputs (chapter 3; Friedman 1953a; Stigler 1976). The central argument of this chapter is that the neoclassical worldview is too narrow to accommodate the possible existence and persistence of infe¬ rior work cultures. As we have discussed, the theoretical framework pre¬ sented in this book addresses the paradox of the persistence of inefficient economic regimes even when both efficient and inefficient firms face the same external economic and institutional constraints.^ Moreover, this revised theoretical framework allows for and helps identify the conditions for the simultaneous existence and persistence of both efficient and inefficient work cultures without violating the neoclassical assumption of utility-maximiz¬ ing, rational economic agents or the Axiom of Modest Greed. An unequivo¬ cal public policy implication is that firms employing the relatively superior

172

CHAPTER 9

work cultures need not pose a threat to firms with the relatively inferior work cultures; therefore market forces alone will not displace the latter. This dis¬ cussion is a subset of the ongoing and controversial question in economics, first articulated by Harvey Leibenstein (1966), as to whether or not x-inefficiency exists and whether or not it can be expected to persist over time.^ Superior Work Cultures and the World of Work This chapter is motivated by the rapidly growing evidence that certain work practices yield relatively large permanent increases in labor productivity, yet they are simply not adopted, and more often than not they are resisted by management. Prior to a detailed discussion of a theoretical frame to explain why this is, it is critically important to gain a handle on the literature that purports to demonstrate that superior work cultures exist that yield higher levels of productivity. It is also important to discuss evidence that hints at why firms typically fail to adopt these work cultures without threatening their competitive position on the market. The point here is not to present a detailed review of the literature, but rather to introduce its critical dimen¬ sions. This literature has fueled a public policy debate, which is constructed implicitly or explicitly by the theoretical frame underlying the discussion. A fundamental finding in the empirical literature, which is assessed in some detail in (for example) Levine and Tyson (1990) and Ichniowski et al. (1996), is that there is a persistent and economically important effect on productivity when firms adopt a bundle of innovative work practices. The whole work culture must be transformed such that it incorporates profit- or gainsharing, long-term employment or job security, narrower wage and sta¬ tus differentials, effective worker input into the firm’s decision-making pro¬ cess, and guarantees on individual workers’ rights so that their inputs into the decision-making process cannot be penalized. Ichniowski et al. (1996) conclude, from a survey of the literature, that the necessary is not typically adopted by American firms. Many firms adopt one or two strands of the bundle. Indeed, only a small percentage of firms have adopted superior cul¬ tures, in spite of their demonstrated productivity-enhancing effect. On the other hand, in both Japan and continental Europe, superior work cultures are much more widespread. On a macroeconomic level Gordon (1996) and Buehele and Christiansen (1995, 1999), for example, show empirieally that economies with more co¬ operative work cultures are characterized by relatively much higher rates of labor productivity growth. This finding holds even after one accounts for the contribution of capital stock to the level and rates of growth in labor produc¬ tivity. It is important to note that economies with relatively cooperative work

WORK PRACTICES

173

cultures tend to have more generous unemployment insurance and income security programs, more effective laws on job dismissal notification and stron¬ ger and more pervasive unions. These institutional parameters enhance the relative bargaining power of employees, thereby encouraging (it is argued) the adoption on a firm level of more cooperative work cultures. Cooperative work culture is defined similarly to innovative work practices. These macro studies focus on developed economies, with the United States, the United Kingdom, and Canada representing the conflictual approach and France, Germany, Italy, and Japan representing the more cooperative approach.*^ A classic U.S. example of the positive effects of a superior work culture on productivity is the New United Motor Manufacturing, Inc. (NUMMI) experiment. A brief narrative of the NUMMI success serves as a controlled experiment, to the extent that this can exist in the real world, demonstrating how the introduction of a bundle of innovative work practices can enhance productivity on a firm level. In this particular situation, Toyota entered into a joint venture with General Motors (GM), taking over the management of a plant in Freemont, California, in 1983 that GM had closed in 1982, laying off 5,700 workers in the process. The plant was closeo after a $150 million investment, making it a state-of-the-art manufacturing facility in terms of plant and equipment. In spite of this investment, it was GM’s least produc¬ tive plant, with labor productivity at half of Toyota’s and one of the highest defect rates in the automobile industry. What characterized the Freemont plant, apart from its disastrous productivity record, was its equally disastrous and conflict-ridden relationship between labor and management. It is into this mess that Toyota entered, hoping to transform this relatively technologi¬ cally advanced plant into a productive enterprise. A key point made by critics of the multifirm statistical studies of superior cultures, which conclude that productivity gains are a product of the intro¬ duction of innovative work practices, is that the studies do not properly con¬ trol for other variables (omitted variable effects) that might also account for the enhanced productivity. Such omitted variables might include changes in capital stock per worker, technical change, and the changing composition of labor inputs, whereby firms that are relatively more productive are so be¬ cause they hire the most productive workers and managers, not because they apply a particular set of work practices. Whatever the validity of these cri¬ tiques, the NUMMI experiment is not vulnerable to them. In addition, the NUMMI success took place in the United States, which is characterized by a macro institutional environment that is severely biased against the adoption of superior work cultures, according to many who have argued in their favor (Levine and Tyson 1990). The major change in the Freemont plant was the introduction of the Toyota

174

CHAPTER 9

system of production, which includes the bundle of innovative work prac¬ tices that most scholars argue is necessary to induce higher levels of produc¬ tivity. Little else changed. Toyota hired about 85 percent its workers from the pool formerly employed by GM in the Freemont plant. Moreover, the same militant trade union representatives and union local became an integral part of Toyota’s plans to transform the facility. But Toyota invested little in plant and equipment. Nor did America’s macro institutional environment change in favor of superior work cultures during the NUMMI experiment. Never¬ theless, in quick order, productivity doubled and the defect rate declined in the Freemont plant. NUMMI became GM’s most productive American plant, manufacturing one car in 19.6 hours, with only 69 defects per vehicle. A defected car required its return to the assembly line for repair. In GM’s most automated plant—40 percent more automated than NUMMI—^it took 33 hours to produce one car, and there were 137 defects per vehicle (Levine 1995; Levine and Tyson 1990; Ichniowski et al. 1996). In spite of this success, the NUMMI example was not followed by GM. The typical GM plant remains largely within the domain of the traditional, rela¬ tively noncooperative, and often conflict-ridden system of IR. Why is the most productive work culture not being adopted by GM? Clearly, the American institutional setting did not prevent Toyota from transforming GM’s old Freemont plant. Moreover, the NUMMI experiment and others like it seem to confirm the findings of the multifirm studies.^ So, to return to a central theme of this chapter, why have superior work practices not spread like wildfire? Can inefficient and relatively unproductive plants survive in a competitive market? Work Cultures and X-Inefficiency As discussed, x-efficiency is defined by the outermost production possibility frontier. X-inefFicient firms operate somewhere in the interior. The existence of x-inefflciency is possible in a world where effort discretion exists. This, in turn, is facilitated by the existence of incomplete contracts and different be¬ havioral functions, especially between employees and members of the firm hierarchy. In terms of a behavioral model of the firm, the existence of supe¬ rior work cultures is suggestive of the existence of x-inefflciency in firms that fail to adopt or develop more productive systems of work organization. Members of the firm, both agents and principals, must work harder and better to generate the relatively higher levels of productivity and quality re¬ vealed by the IR and HRM literature. Therefore, effort discretion in quantity and quality is implied in the firms embraeing the superior work cultures. The literature also suggests that to achieve what may be referred to as relative xefficiency, the firm must invest, at least in the short run, in its organizational

WORK PRACTICES

175

capital (Tomer 1987). This includes improved screening and training of em¬ ployees and managers and redesigning and reconfiguring the shop floor for a more participatory work environment. Members of the firm hierarchy would also have to work harder and better. They would at least have to reinvent their approach to HRM. Moreover, not all layers of the firm hierarchy ben¬ efit from the superior work culture. For example, the ratio of management to employees tends to be much lower in the superior work culture, thereby threat¬ ening the position of many middle managers. In addition, principals may lose power and prestige, to which they attach utility and a heavy weight in their objective function. In this sense, achieving x-efficiency is not a free ride. It clearly involves opportunity costs (both material and nonmaterial) that differ across the spectrum of the firm’s economic agents (Alcaly 1997; Appelbaum and Batt 1994; Gordon 1996; Ichniowski et al. 1996; Klein 1984; Levine and Tyson 1990). Multiple Work Cultures in a Behavioral Model of the Firm The significance of the empirics of the IR and HRM literature for economic theory and vice versa is revealed in a simple behavioral model of the firm developed in this book. Assume that all economic agents attempt to maxi¬ mize their utility in a forward-looking (rational) manner.^ However, the ob¬ jective function of agents and principals can differ, and they may face different objective constraints. It is also assumed that in a competitive product market, principals are subject to the binding constraint that unit costs must be com¬ petitive. Output is a product of inputs such as capital, labor, land, and tech¬ nology, as well as work culture. Ceteris paribus, a more effective work culture generates a higher level of output and quality. Assume also that the product market is subject to severe competitive pressures and that firms are not pro¬ tected in any way (such as through subsidies and tariffs), so that inefficient firms must be subject to the full force of competitive pressures.^ Based on the stylized facts of the firm, also assume that for a superior set of work practices to be effectively adopted, employee compensation must increase and the firm must invest in its organizational infrastructure. What would be the unit costs of x-efficient compared to x-inefficient firms where, for the sake of argument, one assumes that the only difference between two types of firms is the application of different work cultures? Average costs are given by the following basic equation, assuming, for simplicity, that labor and organizational capital are the only inputs into the production function;

AC =

OC+wL

Q

(9.1)

176

CHAPTER 9

where AC is average costs, OC is organizational capital, w is rate of labor com¬ pensation, L is labor input, and Q is output. This equation can be rewntten as;

ac.2£,}L.

Q

Q

(9.2)

L Average costs are determined by the unit costs of organizational capital, OC/ L, and by the per unit labor costs, which can be expressed as the wage rate deflated by labor productivity. Assuming that OC and w increase with the adoption of superior work cultures, it is clear from Equation 9.2 that the higher productivity, x-efficient firm might be characterized by the same av¬ erage costs as the lower productivity, x-inefficient firm if higher productiv¬ ity and corresponding costs rise in an offsetting fashion.^ It is therefore quite possible that the x-inefficient firms can compete on the basis of low rates of labor compensation and a smaller investment in organizational capital. On the other hand, high-wage firms can compete on the basis of more coopera¬ tive and participatory work cultures that generate the higher levels of pro¬ ductivity required for these firms to survive and even prosper in a competitive environment. Under these circumstances there would be no economic rea¬ son for there not to exist simultaneously, in a competitive environment, a multiplicity of firms with a multiplicity of work cultures, each characterized by different levels of output and quality. Both x-inefificient and x-efficient firms can survive even in a competitive environment (chapters 2 and 3). In other words, there is no reason to expect, as a rule, a convergence of work cultures or IR systems. Only when the productivity differential, controlled for quality of output, in favor of firms that use the superior work cultures exceeds the cost differentials between these firms can one expect the com¬ petitive process to result in the dominance of the high-productivity, rela¬ tively x-efficient firms. These arguments are illustrated in Figures 9.1 and 9.2. In Figure 9.1, an array of IR systems (OB) yields a unique average cost OA' and there exist multiple equilibria in terms of different and diverse systems. These include both x-efficient and x-inefficient work cultures. Differences in productivity attributed to the different systems are balanced by differences in associated costs. IR systems where cost differentials exceed productivity differentials generate higher unit costs than OA' (to the right of B) and should be eventu¬ ally eliminated in a competitive market. In Figure 9.2, an array of industrial relation systems (approximately OB) is characterized by different unit costs. There is only one unit cost, OA', that minimizes average costs. In this sce¬ nario, there is a unique IR system, given by B, that generates a unique mini-

WORK PRACTICES

177

Figure 9.1

mum unit cost. In the other systems, the ratio of cost to productivity exceeds what is achieved at B. In such a world, one would expect that competitive forces would drive all system toward B. In this case, B represents the unique cost-effective IR regime. This is what the conventional wisdom would ex¬ pect to characterize superior work cultures. Under the assumption of effort variability, this regime represents the most x-efficient regime if it generates the most effort per unit of labor input.

Conclusion: Superior Work Cultures, X-Efficiency, and Material Welfare If traditional work cultures generate output at the same unit costs as superior work cultures, why is it suboptimal for the traditional work cultures to be maintained? Under the regime with superior work cultures the economy is producing relatively x-efficiently and is therefore closer to the productivity possibility frontier. In other words, superior work cultures allow firms (and

178

CHAPTER 9

Figure 9.2

therefore the larger economy) to produce more per capita output. Thus soci¬ ety is better off with firms exploiting the superior work cultures. In addition, employees are better off in terms of their real income and overall working conditions. Nevertheless, employees do not choose the work culture to be employed by the firm. Rather, members of the firm hierarchy make this choice. If it is utility maximizing for the firm hierarchy to maintain the traditional work culture, given the binding constraint that unit costs must remain com¬ petitive, a superior work culture will not be adopted or developed, even if the traditional culture choice yields x-inefficient results.^ The IR and HRM literature has identified a variety of factors that have impeded the adoption and development of superior work cultures. A major factor is the considerable amount of mistrust between workers and members of the firm hierarchy. Without trust, a more cooperative and participatory work culture cannot be effective. Many members of the firm hierarchy are

WORK PRACTICES

179

also not receptive to power sharing with workers. There is considerable re¬ sistance from managers who believe that they will be disadvantaged by a shift away from the traditional work culture. Moreover, there are consider¬ able short-run costs involved in adopting and developing a new work cul¬ ture, while the returns generally appear in the more distant future. Such longer-run returns are especially problematic because the institutional in¬ vestment environment is most supportive of high, short-run returns that may be more easily realized in the traditional work culture. Under these circum¬ stances, members of the firm hierarchy may find it utility maximizing to maintain the traditional work culture, even if this involves reducing the level of their employees’ pecuniary and nonpecuniary benefits or keeping them below what they might otherwise be. This is especially true when they do not incorporate into their objective function the utility generated to their em¬ ployees by higher levels of income and improved working conditions. These basic impediments to introducing more cooperative and participatory work structures are consistent with the IR and HRM literature findings that supe¬ rior work cultures tend to be adopted when firms are in crisis and reorganiz¬ ing the work system is one approach to restoring the competitive position of the firm. This is especially the case when the superior work culture generates lower unit production costs than the traditional work culture. To neoclassical theory a set of work practices cannot be deemed superior or relatively efficient if they cannot dominate the more traditional work practices in the market place. However, this presumption is largely due to the assumption of neoclassical theory that effort discretion does not exist and that x-inefficiency is impossible, especially in a competitive environment. In sharp contrast, the be¬ havioral model of the firm discussed here reveals when, even under conditions of competitive product markets, rational utihty maximizing individuals will pro¬ duce both x-inefficiently and competitively using traditional work cultures. More¬ over, efficient work cultures need not dominate the less efficient traditional ones if their associated costs are too high. In contrast, superior work cultures allow firms to remain competitive in a relatively high wage environment while contrib¬ uting to an economy’s realization of higher levels of per capita output. The fact that market forces do not displace the traditional work cultures in no way dem¬ onstrates their economic efficiency. Rather, it reveals the extent to which ineffi¬ cient institutions can persist, even under competitive conditions, by virtue of the ability of firms to revert to low wages and poor working conditions that, in turn, protect them from their more efficient competitors. Market forces, under these conditions, cannot force the adoption of the most efficient methods of organizing work. The market only cares that firms remain competitive, even if this on the basis of relatively inefficient, relatively low wage and non-cooperative systems of industrial relations.

180

CHAPTER 9

Notes 1. For a detailed discussion on the more general question of lack of convergence among nations in terms of real per capita GNP given the analytical predictions of the conventional wisdom, see chapter 3, Baumol and Wolff (1988), and De Long (1988). 2. Levine and Tyson (1990) and Levine (1995) elaborate on the significance of different external economic and institutional environments in determining the adop¬ tion of different work cultures by firm management. These environments present ex¬ ogenous constraints to firms, some of which are more conducive to a more participa¬ tory HRM system than others insofar as they affect the material or nonmaterial oppor¬ tunity costs faced by agents and principals in adopting different IR systems. There¬ fore, different external environments can help explain different HRM systems across different external constraints. Of particular significance in facilitating the successful introduction of a more participatory HRM system are long-term staple product de¬ mand, long-term low rates of unemployment, and a legal environment that makes it more difficult and costly to dismiss workers without just cause. Nevertheless, the existence of the different HRM systems within the same external environment re¬ mains of fundamental importance and requires explanation. For a more general dis¬ cussion on how institutional parameters affect an economy’s relative x-efficiency, see Thorbecke (1990). 3. Much of the empirical and theoretical literature on x-efficiency theory is sur¬ veyed in Frantz (1998). 4. Gordon’s analysis also incorporates data from the Netherlands, Norway, and Sweden as representatives of the more cooperative work cultures. 5. See Gordon (1996), Levine and Tyson (1990), Ichniowski et al. (1996), and H. Smith (1995) for a discussion of other case studies. 6. Leibenstein (1966, 1978a, 1979) assumes that economic agents who do not perform x-efficiently in production are only quasi-rational. Rational behavior is as¬ sumed to be behavior that is consistent with x-efficiency. However, rational economic agents need not perform x-efficiently if such behavior is not of a utility-maximizing type (chapters, 1—3). 7. The x-efficiency literature largely assumes that x-inefficiency is possible only when x-inefficient firms are afforded some protection, through monopolistic powers or subsidies, for example, since it is assumed that the x-inefficient firm must be a relatively high unit cost producer (Leibenstein 1966; Frantz 1998). I argue (chapters 1 and 2) that this need not be the case since x-efficient production entails higher costs than x-inefficient production so that, on balance and controlling for quality, the unit cost of the x-efficient and x-inefficient firms might be identical. 8. In the standard efficiency wage literature a unique wage exists that generates a unique level of effort per unit of labor time consistent with maximizing labor produc¬ tivity and thereby unit labor costs. In this case, however, it is not possible for two sets of firms with different wage rates and associated work culture, both producing at the same unit costs, to exist (Akerlof and Yellen 1986; Akerlof and Yellen, eds. 1986). This assumes a traditional production function, generating an inverted-U relationship between wages and effort and effort and productivity. Assuming some linearity in the production function allows for a set of different levels of labor compensation yielding a unique level of unit costs (Chapters 1,7, 8; Stiglitz 1987). 9. These results fit into the unfair Prisoner’s Dilemma game designed by Merrill

WORK PRACTICES

181

Flood and Melvin Dresher (Poundstone 1992,106-116). In this game all players com¬ bined are better off if the cooperative (in this book, the superior HRM system) strat¬ egy is adopted. However, one player, the firm hierarchy, is no better off than if it defects to the traditional work strategy. Labor finds the cooperative strategy superior but only if the firm hierarchy adopts the same strategy. Otherwise, labor would be worse off. If labor expects the firm hierarchy to defect, it is in labor’s self-interest to adopt the same strategy. We end up with a Prisoner’s Dilemma solution (both parties defecting) when the firm hierarchy does not gain from cooperation—that is, the adop¬ tion of the superior HMR system. Assuming that all strategies are consistent with being cost competitive, we end up with a social dilemma where the socially superior results are rejected by the firm given an institutional structure that is not conducive to cooperation (Miller 1992).

10

The Efficiency- and WelfarePromoting Role of Labor Rights and Labor Power in a Market Economy

Introduction

Conventional economic wisdom argues that any interference in workings of the free market with respect to the firm will have negative consequences for the firm itself as well as for the larger economy. In other words, the firm’s competitive position will be threatened, and the larger economy will suffer a loss of output and employment. More specifically, any attempt to force higher wages, labor benefits, or otherwise improve standards above what the mar¬ ket dictates, by legislation or by enhancing the power of labor, is frowned upon. For this reason efforts to promote minimum wage legislation or unions or impose labor standards such as health and safety regulations, normal hours of work, and age restrictions are deemed to be the victory of soft hearts over hard minds without foundation in the economic science. Many scholars more sympathetic to the plight of workers accept the logic of the conventional wisdom but argue that improvements in labor standards are worth the costs, and to prevent any one firm from losing its competitive edge with the introduction of improved labor standards, minimum standards should be legislated across the board and trade policy should be designed to protect high-standard domestic firms from low-standard foreign competi¬ tion. The argument presented in this chapter is that conventional economic wisdom is correct only under extreme and unrealistic behavioral assump¬ tions about the economic agent and the firm. More realistic, generally appli182

LABOR RIGHTS AND POWER

183

cable assumptions generate different predictions and policy prescriptions about the implications of providing labor with more rights and more power inside of the firm. In a behavioral model of the firm, presented below, in¬ creasing labor rights and power need not reduce the competitive position of a firm in either a closed or open economy. This result should hold true no matter the degree of product market competition—even in the extreme case of perfect competition. This chapter is concerned with the implications of labor rights and power for the economy: whether or not such leverage by labor can have persistent negative or positive effects on the firm specifically and more generally on the economy as a whole. What is at issue, of course, is whether or not a meaningful set of labor rights and power yields higher unit costs and lower profits to the firm, with all the deleterious effects upon the firm and the economy that these are expected have. ^ The Conventional View of Labor Rights and Powers The conventional economic wisdom on labor rights and power was plainly and clearly stated by George Stigler (1946) with regard to government legis¬ lation of a minimum wage that is above what the market would generate in the absence of such legislation. The specific case Stigler makes against mini¬ mum wage legislation can be easily generalized, without any loss of rigor or practical analytical context, to the potential impact that labor rights and power might have on the firm and the economy through such intermediaries as unions, health and safety legislation, age restrictions, and any other types of nonmarket intervention. Stigler argues that minimum wage legislation can be expected to reduce aggregate output and result in an increase in unem¬ ployment. It can also be expected to benefit workers who previously earned below the minimum and retain their jobs in face of minimum wages. The net effect of minimum wage legislation is clearly negative from the perspective of the theoretical prism used by Stigler and the conventional wisdom then and now. Stigler admits that these negative effects can be partially or even completely avoided if the productivity of labor is increased as a result of the minimum wage legislation. This can occur only if workers and management were not previously as efficient as they could possibly be and if the mini¬ mum wage legislation shocked them into becoming more efficient. How¬ ever, this possibility is considered to be little more than a fantasy, a theoretical special case buried in the general and more realistic case of economic agents working as hard and as well as possible along the production possibility frontier (Stigler 1946, 1976). In the same vein, the conventional wisdom predicts that the realization of more rights and power by workers will indubi¬ tably have negative effects on the economy, with few exceptions.^

184

CHAPTER 10

An important twist to the conventional wisdom on introducing or improv¬ ing labor standards is articulated by Richard Freeman (1994). He assumes, along with the conventional wisdom, that new labor standards or improve¬ ments therein increase unit production costs. Clearly, in this case, ceteris paribus, firms with more stringent labor standards will incur higher unit costs. For this reason, economists who accept the logical imperative of the conven¬ tional wisdom but who nevertheless believe that labor standards are worth the cost from a social welfare perspective argue for legislating minimum standards so as to create an equal playing field for all affected firms. Free¬ man argues for labor standards but against legislating labor standards. He maintains that improved labor standards associated with the production of a particular product should be viewed as a distinct commodity and are viewed as such by the typical consumer. For example, two otherwise identical auto¬ mobiles produced under different conditions of work would be identified as two distinct products and. Freeman argues, the typical consumer would be willing to pay a premium for the product produced under better working conditions. The premium covers the additional production costs generated by improvements in labor standards, and it can be expected that the demand for the product will diminish as the premium increases—one has the typical negatively sloped demand curve, in this case for labor standards piggy-backed on a particular product. In effect, one can therefore model a market for labor standards, where labor standards impose increased production costs, as pre¬ dicted by the conventional modeling of the firm. Moreover, the costs of im¬ proved labor standards would be borne by consumers and producers in aceordance with the elasticities of demand and supply. This is illustrated in Figure 10.1 for a multifirm, one-produet, perfectly competitive economy, where the introduction of labor standards shifts the supply curve of a par¬ ticular product upward to S',. Assuming that at least some firms in an indus¬ try adopt a bundle of labor standards, the increased cost per unit of output can be expected to generate a higher equilibrium price, P,, thereby reducing demand and equilibrium output to In this case, the direct increased eosts generated by improved working conditions are borne by both consumers and producers. Freeman argues, however, that if consumers know that the higher price is the result of better labor standards, they will reallocate their demand toward the higher-priced product, thereby shifting the demand for better la¬ bor standards upward from Z)q to Z), and reducing the demand for the lower priced product of the low labor standards firms. The shift in demand for the higher-prieed product serves as a countervail against the negative impact on demand that the higher price would otherwise have. According to Freeman, in spite of the cost-augmenting nature of labor standards, they can be intro¬ duced and even improved upon in a free market economy since effective

LABOR RIGHTS AND POWER

185

Figure 10.1

labor rights and power are sustainable and even welfare maximizing when they generate economic results that are consistent with consumer preferences. The key to the successful implementation of labor standards is the correc¬ tion of one serious product market imperfection: the costliness of obtaining information on labor standards, which make it impossible for consumers to choose the product—labor standards bundle consistent with their preferences. Freeman suggests that government intervention is required to minimize the market failure that flows from inadequate information. Government should administer a system of compulsory product labeling that would identify the working conditions surrounding the production of a particular product. Such product labeling would ideally allow consumer preferences to reign supreme, thereby providing the consumer a clear choice of buying a product produced under superior or inferior working conditions. As noted. Freeman assumes that consumers will tend to prefer a more highly priced commodity pro¬ duced by well-treated adult labor to a lower-priced commodity produced with child labor or low-paid, poorly treated adult labor. In this case. Freeman argues, the low labor standards producers will lose market share and suffer from a decline in profits, ultimately resulting in the dominance of the high labor standards producers. Nevertheless, there might be two markets in equi-

186

CHAPTER 10

librium for a particular product, one for the higher-priced, high labor stan¬ dards product and another for the lower-priced, low labor standards product. The latter would consist of individuals for whom lower prices dominate higher labor standards. Often such individuals simply cannot afford the more ex¬ pensive product. There is no way to predict, a priori, whether consumer preferences will result in the high labor standard firms’ products dominating the market. However, as discussed below, in an economy where economic agents are not typically and automatically working as hard and as well as they can and where working conditions affect labor productivity, effective product label¬ ing can contribute toward the elimination from the marketplace of low labor standards firms. In this economy, one might end up with identical unit costs for both low and high labor standards producers. And the choice afforded to consumers will be the less difficult one between the same product offered at the same price but produced under different bundles of labor standards. Here there would be no trade-off between price and labor standards, and there would be no premium paid for commodities produced under superior labor standards. But this takes us beyond the limits of conventional economic wis¬ dom into the realm of behavioral economics and, more specifieally, to a dis¬ cussion of x-efficiency and efficiency wage theory. An Alternative View of Labor Rights and Powers It bears repeating that critical to the prediction that improved labor stan¬ dards, inclusive of higher real wages, inevitably generate higher unit costs is the assumption that economic agents are typically producing efficiently along the production possibility frontier. Empirical work that suggests otherwise is dismissed as inconsistent with the theory and therefore probably character¬ ized by methodological flaws or simply sampling error. However, if the eco¬ nomic theory itself is derived from questionable behavioral premises, it might very well be the empirical evidence that is correct and the theory that is wrong (chapter 3). An underlying theme throughout this book is that the quantity and quality of effort is a discretionary variable, and therefore labor productivity is not simply a function of capital per worker or technological change. This premise underlies a behavioral model of the firm that suggests that the conventional modeling of human agency is a special case and that, more often than not, economic agents are operating below their potential— that is, below the production possibility frontier.^ Moreover, such modeling is consistent with the literature on unions and minimum wage legislation that suggests that these have little if any negative economic effect and might very well positively affect the economy in terms of productivity growth.'^

LABOR RIGHTS AND POWER

187

For effort to be a discretionary variable, the marginal transaction costs of metering, monitoring, enforcing, and writing effort-specific contracts must exceed the marginal benefits. This must be true with regard to both manag¬ ers managing the effort inputs of workers and owners managing the effort inputs of managers. Once effort is discretionary, indirect incentives must be developed to ensure that the quantity and quality of effort are maximized. There is a growing literature suggesting that only in more trusting, coopera¬ tive forms of industrial organization is effort per unit of labor input maxi¬ mized (chapter 9; Gordon 1996; Ichniowski et al. 1996; King 1995; Levine and Tyson 1990; Miller 1992; Pfeffer 1995). Under such a regime, labor has a direct and meaningful input into the decision-making process; labor com¬ pensation is related in part to productivity; there is relative job security and a relatively less hierarchical administrative structure. In this case, labor pro¬ ductivity is affected by the quantity and the quality of effort inputted into the production process, and it can be expected to be below the maximum unless an ideal system of industrial relations is adopted by the firm hierarchy. In this behavioral model of the firm unit costs are in part determined by effort per unit of labor since this variable, unlike in the conventional model, im¬ pacts upon labor productivity. Ceteris paribus, unit costs rise as the quantity and quality of effort diminish. This is clear from Equation 10.1, where v4C is average cost, Pj is the price of inputs, Q is output, / is inputs (incorporating labor, capital, land, and labor standards), and e is a multiplier that is at a maximum of 100 when effort per unit of labor is maximized. As e falls be¬ low the maximum, input productivity diminishes, which in turn increases unit or average cost.

AC^

Pi ^ Qxe^

(10.1)

Leibenstein refers to firms with these effort-related higher costs as being xinefficient. These firms can survive, according to Leibenstein, under a variety of conditions whereby they are protected from relatively x-efficient firms. The traditional x-efficiency literature emphasizes the protection afforded to the relatively inefficient firms by monopolistic market structures, govern¬ ment subsidies, and tariffs. In effect, in a world with competitive markets, sustained x-inefficiency in production becomes impossible. Economic agents must become x-efficient if the firm is to survive. All surviving firms, there¬ fore, must be x-efficient in competitive equilibrium. However, I argue that more important to the survival of the x-inefficient firms is the price of inputs, in particular the price of labor and related costs (chapters 1 and 2). For ex-

188

CHAPTER 10

ample, an x-inefficient firm can remain competitive if the price of inputs is kept to a sufficiently low level. From Equation 10.1, if the only input is labor and the firm is relatively x-inefficient, average cost can remain competitive if the level of labor compensation is sufficiently low. In fact, low wages and substandard conditions of work can serve as substitutes for more efficient firm performance. They protect the x-inefficient firm from competitive threats posed by more x-efficient firms. In this case, relatively x-inefficient firms can survive in competitive equilibrium, and all surviving firms need not be x-efficient.' On the other hand, high-wage—high labor standard firms need not be uncompetitive when the wage rate and work standards affect labor produc¬ tivity via effort inputs. Simply put, higher wages and overall improvements in working conditions, in the context of a more cooperative and trust-based system of industrial relations, induce higher quantity and quality of effort yielding higher levels of productivity that serve to compensate for the higher production costs. To become more productive or x-efficient, as discussed in chapter 9, the firm must invest, at least in the short run, in the organizational capital of the firm (Tomer 1987). Members of the firm hierarchy would also have to work harder and smarter. They would, at least, have to reinvent their approach to human resource management (chapter 9; Alcaly 1997; Appelbaum and Batt 1994; Gordon 1996; Ichniowski et al. 1996; Levine and Tyson 1990). Achieving more x-efificiency in production is clearly not a free ride. But the higher wage rates and improved labor standards need not yield higher aver¬ age costs if they induce sufficiently higher levels of labor productivity. When the overall conditions of work affect labor productivity, high-priced labor working in a high labor standards environment can compete with low-priced labor working in a low labor standards environment if labor productivity is related to the level of labor compensation and labor standards. This would be true even in a perfectly competitive product market environment. Imper¬ fect product markets and other forms of protection are not required to protect high-priced labor firms that are relatively efficient, and, on the other hand, competitive pressures on the product market cannot undermine the economic viability of relatively inefficient firms that are protected by low-priced labor and low labor standards. From Equation 10.1, unit costs can be identical for the relatively efficient and inefficient firms if the input prices are adjusted sufficiently to accommo¬ date different levels of productivity or, alternatively, if productivity levels are adjusted sufficiently to accommodate different levels of input prices. In this case, one can have the simultaneous existence, in competitive equilibrium, of both high and low labor standards firms, where the former firms are no more costly to operate than the latter. The higher labor standards firm can survive if

LABOR RIGHTS AND POWER

189

efficiency improvements, triggered and associated with improved labor stan¬ dards, suffice to compensate for increased costs. It is assumed here that there exists an array of wage rates or, more generally, rates of labor compensation inclusive of an array of bundles of labor standards or working conditions that is associated with an array of labor productivity levels that yield identical unit costs. Unlike in the standard efficiency wage literature, some linearity in the relationship among labor standards, effort, and productivity is assumed (Akerlof and Yellen 1990; Stiglitz 1987).^ Of course, a point may be reached where further increases in rates of labor compensation and labor standards result in higher unit costs. What this point is is an empirical question. To repeat, for improvements in labor standards not to have a positive impact on production costs requires that effort be a discretionary variable and that x-efficient firm performance is not a given but rather is a function of a complex set of factors inclusive of the wage rate and labor standards. This argument can be taken one step further by introducing technical change. We have argued elsewhere that rates of labor compensation can af¬ fect the rate of technical change (as opposed to simply the choice of tech¬ nique). The introduction of new technology can help keep unit costs competitive in the face of higher rates of labor compensation and improve¬ ments in labor standards. In this sense, higher labor standards serve to in¬ duce technical change. Alternatively, technological change need not be introduced for low-wage—low labor standards firms to remain competitive if such firms can produce at the same unit costs as the higher labor standards firms that have opted for the new technology. In addition, the level of xefficiency can affect the new technology, which might be viable only at a relatively high level of x-efficiency (chapters 2 and 6). In this more dynamic modeling of the firm, one can imagine a situation where higher-priced labor need not generate higher unit costs as a conse¬ quence of the possible impact of increased wages and improvement in labor standards upon labor productivity. This argument may be further illustrated in Figure 10.2. Along JK, unit costs are identical. As working conditions improve from to 0 along the horizontal axis, inclusive of wages and labor standards, the level of x-efficiency must increase sufficiently, moving from 0 to /along the vertical axis. If not, and one ended up at a point such as A, unit costs would rise since for the given level of working conditions {KC), the level of x-efficiency would not be sufficiently high. The level of x-efficiency would have to be at D. Alternatively, at a point such as B, unit costs fall since the level of x-efficiency is more than enough to compensate for KC of work¬ ing conditions. As long as changes in technology or the level of x-efficiency compensate for changes in the price of labor and associated costs, there is no need for unit production costs to change with changes in the price of labor

190

CHAPTER 10

and associated costs. This is in stark comparison to the conventional model¬ ing of the firm, where it is assumed that firms are operating x-efficiently irrespective of wage rates and labor standards and any increase in wages or improvement in labor standards necessarily produces higher unit costs. To the extent that the behavioral modeling of the economic agent and the firm is more appropriate than the conventional modeling, the negative pre¬ dictions on the repercussions of higher wages and improved labor standards that flow from the conventional thinking would not hold. In Figure 10.1, improvements in labor standards need not shift the industry supply curve upward to Sy It is even possible that the efficiency effect of such improve¬ ments might shift the supply curve outward, generating opposite effects to what are predicted by the conventional wisdom. Figure 10.3 illustrates the key differences in the implications of the conventional and behavioral mod¬ eling. According to the conventional wisdom, firms with increasing rates of labor compensation and improvements in labor standards experience higher unit costs, making them less competitive and forcing them into bankruptcy, into soliciting protection from the state, or into engaging with like firms so as to increase market power. Such actions, if successful, would afford these firms some degree of freedom to raise prices to compensate for higher unit costs. In the behavioral model, unit costs need not rise as direct labor costs increase since firms possess the capacity to improve upon productivity, and increases in wages and improvements in labor standards themselves can con¬ tribute to the structuring of the organizational capital and the technology of the firm so as to increase productivity. At some point, the behavioral model joins with the conventional one in the sense that when productivity is at its maximum, higher-priced labor will generate higher unit costs. Much of the literature that favors improvements in labor standards, unions, and minimum wage legislation implicitly assumes that a behavioral model of the economic agent and the firm and the assumptions that underlie it best explain the workings of a market economy (Campbell and Sengenberger 1994; Feis 1994; R. Marshall 1994; Sengenberger 1994a, 1994b; Wilkinson 1994). For example, Ray Marshall argues that from 1945 to 1973, “Labor standards (whether enforced by government regulations or collective bar¬ gaining) improved economic efficiency by removing worker (or public) sub¬ sidies [low wages and poor labor standards] to firms that could not provide acceptable working conditions. In forcing companies to compete by increas¬ ing efficiency rather than by reducing labor standards, this, in turn, shifted resources to more efficient uses and allowed countries to protect and develop human resources—their most valuable assets” (1994, 69). More often than not, these scholars argue in favor of government intervention to protect the high-wage-4iigh labor standards firms, including trade protection from low-

LABOR RIGHTS AND POWER

Figure 10.2

Figure 10.3

Working Conditions

191

192

CHAPTER 10

wage—low labor standards countries. They also argue for the imposition of like labor standards across different firms and economies so as to create a level playing field for all firms in a relatively competitive market. As Ray Marshall puts it: “We cede the right to set our own standards if we fail to regulate the goods coming into a country because a basic principle of highly competitive markets is that bad standards tend to drive out the good. Com¬ petitive markets make it difficult for employers who want to have good stan¬ dards to do so, even though in the long run good labour practices enhance economic efficiency” (1994, 72). In fact one of the issues raised in the litera¬ ture sympathetic to improved labor standards is discussed by Freeman (1995), “Are Your Wages Set in Beijing?” If indeed wages and labor standards are not set by the lowest common denominator in terms of either local or global competition, then the question arises as to why rules or regulations must be set for wages or labor standards. More specifically, why are not the highest labor standards chosen by economic agents irrespective of political bound¬ aries? The latter question directly relates to the issue of market failure and social welfare. Who Chooses Conditions of Work? Market Failure in Wage Determination and Labor Standards Where are wage rates and, in particular, labor standards set when labor mar¬ kets are relatively imperfect and there does not exist an easy and free flow of labor from one geographical place to another for economic, cultural, or po¬ litical (among other) reasons? Labor market imperfection is a key character¬ istic of market economies, especially with respect to the flow of labor across international boundaries. A critical component to the argument that wages in the more developed countries are being set by those in the less developed countries is the view that market forces set in motion by trade will force the equalization of wages across these two groups of countries since the highwage countries will find their products outpriced by the like commodities produced in the low-wage countries using identical technologies. (One as¬ sumes that factor price equalization theory actually speaks to the workings of a modem market economy.) As indicated in note 2, the OECD (1996) study finds little evidence to support the view that the existence of or improvements in labor standards have a negative effect on a country’s competitive position in the interna¬ tional market, and there is no evidence to support the hypothesis that low labor standards are correlated with low labor costs or export prices: “The view which argues that low-standards countries will enjoy export market shares to the detriment of high-standards countries appears to lack solid

LABOR RIGHTS AND POWER

193

empirical support” (OECD 1996, 105). This study also finds that labor standards do not play a determining role in the investment decisions of multinational enterprises (OECD 1996, 123—124). Thus improvements in labor standards need not drive out investment toward low labor standard countries. Speaking largely to the wages of unskilled workers and after surveying a wide array of the empirical literature on the subject. Freeman (1995) concludes that there is little evidence that the wage rates in the more developed countries are now being set by the wages in the less devel¬ oped countries as trade between these two economic spheres has increased. This is especially true of the growing proportion of unskilled workers in the nontraded services sector. David Gordon (1996, ch. 7) reaches a simi¬ lar conclusion for American workers in general. Paul Krugman addresses the more general question, from both an em¬ pirical and theoretical bent, of how increasing trade with less developed countries and growing international trade in general might impact upon per capita income in the more developed countries. Krugman finds the sug¬ gestion that wages in the more developed economies have been reduced by trade with the low-wage economies is “flatly rejected by the data” (1994, 121). Moreover, he concludes that in theory, growth in trade with the poorer, low-wage economies is as likely to increase as to decrease per capita in¬ come and real wages in the wealthier, high-wage economies. Critical to Krugman’s theoretical case is the argument that wages are typically and roughly proportional to productivity, with productivity changes driving wage rates, especially in sectors subject to international trade. Hence, low wages will not generate a price advantage to low-wage economies. This line of reasoning is provided a strong theoretical foundation by the behavioral model of the firm presented here, wherein changes in productivity are driven by changes in labor costs. In this case, relatively higher labor costs are sufficiently matched by higher labor productivity to keep unit costs from rising and high-priced labor competitive. Alternatively, lower labor costs need not generate lower unit costs because of the productivity effect of lower wages or lower labor standards—low-priced labor need not generate a competitive advantage to low-wage economies.^ These arguments, par¬ ticularly in the context of the behavioral model of the firm presented here, strongly suggest that increasing trade with low-wage economies need not pose a competitive threat to economic jurisdictions characterized by rela¬ tively high wages and labor standards. If international, national, regional, or local trade does not necessarily pose a threat to relatively high wages and high labor standards and regulatory intervention is not necessarily required to protect high-wage firms or econo¬ mies, can one still argue that regulatory intervention in the economy is re-

194

CHAPTER 10

quired with respect to wages and labor standards? According to the behav¬ ioral model of the firm, both high- and low-wage economies can be competi¬ tive in terms of both unit costs and profits, whereas only in the former is the workers’ level of material welfare relatively high.^ Why then should firms choose the low-wage road to economic growth? First, it is important to re¬ peat that such a decision is typically made in the market economy by the firm’s managers/owners. If it is utility maximizing for the firm hierarchy to maintain the low labor standards, given the binding constraint that unit costs must remain competitive, improvements will not be implemented. It is clear that the low-wage road is preferred by the typical corporate leader in North America (Kochan, Katz, and McKersie 1986; Gordon 1996) and, increas¬ ingly, the corporate leadership in Europe (Albert 1993). Given their prefer¬ ences, one would not expect utility-maximizing managers/owners to choose the high-wage-diigh standards path if both the low- and high-wage-diigh stan¬ dards path yield the same unit costs and rates of profits. Only if one of the arguments in the principals’ utility function is the utility of the workers, whose utility is positively affected by higher levels of x-efficiency, would the prin¬ cipals be expected to proactively work toward more x-efficiency in produc¬ tion. Otherwise, the utility of the majority of firm members remains an externality to the principals. Market forces, even highly competitive market forces, need not cause this important externality to be internalized by the principals, when the x-efficient firms remain cost competitive with the xinefficient firms. Why might managers/owners choose the low standards firm if the high standards firm is equally cost competitive and profitable? Moving from a low to a high labor standards firm involves both net pecuniary and/or nonpecuniary costs. The high-wage road involves investing in organizational capi¬ tal, the returns to which are not instantaneous nor without at least some risk. For example, the high-wage-Tiigh standards road involves restmcturing labormanagement relations and reengineering the shop floor design. Moreover, there is no evidence to suggest, either on average or on the margin, that there are any net benefits to managers/owners from investing in organizational capital in terms of higher profits, higher managerial income, or lower unit costs. In this sense the private rate of pecuniary return to managers/owners would be zero. Given the private costs and benefits of adopting higher labor standards, the higher the rate of time preference, the higher the risk premium (which is positively affected by the degree of labor-mianagement distrust), the more likely it is that managers/owners will avoid choosing the highwage road. This is so since, ceteris paribus, the greater the rate of time preference or myopia and the greater the risk premium, the lower is the net present value of income to the firm, where the former negatively affects in-

LABOR RIGHTS AND POWER

195

come and the latter positively affects costs. In addition, the high labor stan¬ dards firm might even require the downsizing of management, as well as a change in the power relationship among various (although not all) levels of management and workers. These changes might also negatively impact upon the utility of management, reducing the chances of a high-wage road being freely chosen by management, even in a world where future benefits are not discounted and are riskless (chapter 9; Gordon 1996; Ichniowski et al. 1996; Miller 1992). To the extent that managers/owners, when free to choose, choose the low-wage path in spite of the high-wage road’s being competitive and preferred by workers, one enters into the realm of possible market failure. The calculus of determining net social benefits as opposed to net private benefits is quite different. The behavioral model of the firm suggests that the benefits of the high-wage road are higher levels of real per capita output as well as higher real wages and higher labor standards. The economic costs of the high-wage road are related to the organizational capital and other costs required for a firm to minimize the level of x-inefficiency and adopt the appropriate technology in the face of higher labor-related costs. In terms of the private calculus, the key decision makers of the typical firm do not nec¬ essarily accrue any direct or clear benefits from workers being materially better off and/or from society being characterized by higher real output per capita, although they bear the in-firm short-run costs of shifting to the highwage road. Since the managers/owners need not incorporate into their objec¬ tive function the benefits to their employees of the high-wage-high standards firm (a component of the social benefits), they would be underinvesting in the high-wage-Tiigh standards firm, and herein lies the potential for market failure. In this case, managers/owners fail to internalize a crucial externality. A clear exception is the case where the managers/owners are altruistic in the sense that their utility is positively affected by the anticipated net material and nonpecuniary benefits accruing to their workers from the high-wagehigh standards firm. In this context, rules, regulations, and laws that facilitate unionization and labor-mianagement cooperation can serve to indirectly promote the highwage-Tiigh labor standards road to growth without introducing “excessive” government intervention. Unionized labor, in the context of the marketplace, can serve to prevent the managers/owners from realizing their preferences and to encourage them to choose the high-wage road. Similarly minimum wage laws, health and safety rules, and limits to the legal working day estab¬ lish a floor that forces firms to do better—to approach their production pos¬ sibility frontier—^preventing managers/owners from moving to an even lower-wage road. Such intervention can serve to reduce, if not prevent, the type of market failure involved in adopting the low-wage road. The eco-

196

CHAPTER 10

nomic theory discussed here suggests that this should be as true on an inter¬ national as on a national level.* Conclusion Arguing that relatively high wages and labor standards need not pose a com¬ petitive threat to a firm or to the economy at large in no way diminishes the difficulty involved in choosing a high-wage road to growth. One obstacle to choosing this road is that making a firm and an economy more efficient takes time. This exposes them to a potential competitive threat from low-wage and low labor standards firms and economies. This threat must be kept at bay during the transition period. In addition, the high-wage road might require investing in what Moses Abramovitz (1986) refers to as social capabilities— investments by government in the social and economic infrastructure that facilitate or permit economies to realize higher levels of per capita income and growth. The level of social capabilities sets an exogenous upper-bound limit on how much labor productivity can be increased to compensate for higher wages and labor standards. On the other hand, roadblocks against movements toward the low-wage road pressure governments into ensuring the development of the social capabilities necessary for the high-wage road to become sustainable and for an economy to remain competitive. Increasing labor rights and power can have welfare-enhancing effects on the economy, especially if the appropriate social capabilities are developed. The behavioral model of the economic agent and the firm suggests that in¬ creasing labor rights and power, translated into higher wages and improved working conditions, need not increase unit costs and therefore need not threaten the economic viability of the firm since such an economic milieu can serve to pressure and motivate economic agents to work harder and better. In this case society as a whole is better off. Increasing labor rights and power threaten the economic viability of the firm only if they generate higher unit costs and lower profits. There is no clear evidence that this is the case. To the extent that higher unit costs result, as suggested by Freeman, the careful classification of prod¬ ucts by the working conditions under which they were produced would allow consumers to impose high labor standards by voting with their purchases. In this case, the higher prices generated by improved working conditions would be consistent with the maximization of social welfare in that they would be consistent with consumer preferences, unless, of course, negative externalities are associated with the higher product prices. Under more reasonable assumptions than those of conventional economic wisdom, it is more likely that enhancing labor rights and power will not have a negative effect on the economy, and, indeed, it is more probable that the

LABOR RIGHTS AND POWER

197

opposite would be true. Only empirical research can determine the manner in which higher wages and improved labor standards will impact upon a firm or economy. The alternative approach to modeling the economic agent and the firm presented here also suggests that there is nothing in theory to con¬ demn as economically catastrophic efforts to improve conditions of work. For this presumption of catastrophe alone, economic policy, which is largely derived from the conventional economic wisdom, will be unnecessarily bi¬ ased against the high-wage-high labor standards road to economic growth. Notes 1. This chapter’s focus is on the possible cost implications of improving labor’s rights and related standards. For this reason, I abstract from the important question of the possible employment and welfare implications that bans or further restrictions on child labor might have, especially on low-income families, even if such tightened standards are shown not to have any positive effect on unit production costs or any negative effect on profits. Some of these issues are addressed in chapter 11 and Basu (1999a, 1999b). 2. A recent OECD (1996) study using conventional neoclassical analysis exam¬ ines what impact can be expected on the efficiency of market economy from the elimi¬ nation of child labor, forced labor, and discriminatory hiring practices and the free¬ dom of association and the right to organize and bargain collectively—referred to as core labor standards. The study argues that child labor, forced labor, and discrimina¬ tory hiring practices distort the allocation of resources away from the free market ideal, thereby shifting the production possibility frontier inward. Ending these prac¬ tices would have the opposite effect. Neoclassical theory, however, would suggest that successful profit-maximizing firms would not adopt such practices if these were efficiency reducing since this would only increase unit costs and the competitive po¬ sition of these firms. The report further argues that the absence of freedom of associa¬ tion and the inability to bargain collectively might result in workers being exploited— being paid below the equilibrium market rate—^and that labor power might force em¬ ployers to pay workers their marginal product and might even have the dynamic effect of increasing labor productivity. However, the latter possibilities are marginalized in the standard neoclassical analysis of the labor market. This entire gamut of theoretical possibilities can be encompassed within the framework of the behavioral model of the firm, which broadens the horizons of the neoclassical analytical framework. It is note¬ worthy that this OECD report finds no empirical evidence to support the view that improvements in core labor standards have negative economic effects. 3. See Leibenstein (1966, 1987); Button, ed. (1989); Frantz (1997). See also Leibenstein and Maital (1994) and Tomer (1987) for important extensions of the xefficiency model. Akerlof (1984), Akerlof and Yellen (1986,1990), and Stiglitz (1987) pioneered work on efficiency wage theory. See also Akerlof and Yellen, eds. (1986). Miller (1992) places this literature in a game-theoretic framework. On this, also see Leibenstein (1982). 4. This literature is summarized in chapter 5. See Freeman and Medoff (1984) for the impact of unions on economic performance and Card and Krueger (1995) on the economic impact of minimum wage legislation.

198

CHAPTER 10

5. This assumption is unlike the one that underlies the traditional efficiency wage literature, which presumes without any evident empirical bearing that there exists one unique wage that yields a unique minimum unit production cost through its inpact on effort per unit of labor input (Akerlof and Yellen, eds. 1986; Stiglitz 1987). Any devia¬ tion from this efficiency wage yields higher unit costs. No rational firm hierarchy should therefore attempt to reduce or increase wages, ceteris paribus. 6. See Altman (1996b) for a detailed discussion of the relationship among wages, levels of x-inefficiency, and labor productivity and how a behavioral model of the economic agent and the firm better explains the lack of convergence between lowand high-wage economies than do the conventional economic models. 7. For m6re details on this point, see chapters 5 and 6. 8. One argument against the imposition of international labor standards is that this will reduce the competitive advantage of less developed low-wage-low labor stan¬ dards economies (Amsden 1994; Krugman 1994). However, to the extent that such international standards pressure these economies into becoming more efficient, there is no reason to expect that they should become less competitive. In this sense, what’s good for the goose is good for the gander.

11

A Revisionist View of the Economic impiications of Chiid Labor Regulations

Introduction The existence and persistence of child labor, variously defined, remains an issue of great concern and contention among social and religious activists, entrepreneurs, politicians, and scholars alike, given the large number of chil¬ dren still engaged worldwide in economic activities, especially in the lowerincome countries. According to the most recent and comprehensive estimates by the International Labour Organisation (2000a), in 1995 there were about 250 million children between five and fourteen years of age engaged in eco¬ nomic activities throughout the world, with about 120 million of these work¬ ing full time. Of the working children, 140 million are boys.' These estimates do not include children engaged in household work, largely girls, since their labor does not contribute directly to marketed output. Africa’s share of the world child labor force amounts to 32 percent, Asia’s to 61 percent, and Latin America’s to 7 percent (Table 11.1). The estimated 250 million child laborers is a large number given the fact that the world’s labor force stood at about 2.9 billion people in 1995 and that these children were working almost entirely in the low-income and middle-income countries, with a labor force of about 2.5 billion people. Children therefore contributed about 10 percent to these countries’ labor force in 1995. Most children, about 94 percent, are employed in economic activities ori¬ ented for domestic consumption. At least 70 percent are employed in the agricultural sector. The large majority are unpaid family workers engaged in small production units found mainly in the agricultural sector but also in the 199

200

CHAPTER 11

Table 11.1

Child Labor in the World Economy Boys and girls

Boys_Girls

250

140

110

80 15 18 1

45 82 12 0.29

35 70 6 0.22

World

100

6

44

Africa Asia Latin America Oceania

32 61 7 0.2

56 54 67 57

44 46 33 43

World (millions) Africa Asia Latin America Oceania Percentage share

Source: Derived from Box 1, International Labour Organisation (2000a).

urban informal sector (Lansky 1997). Since about half of the children are employed full time, they are deprived of the education (the investment in human capital) that might provide them with the opportunity to earn higher incomes in the future. Moreover, it has been argued that even if children work on a part-time basis but more than twenty hours per week, it can nega¬ tively affect their capacity to learn and thus their ability to accumulate hu¬ man capital. In addition, many children are employed in jobs that are hazardous and prone to injury, in the sex industry, or as bonded or slave laborers, albeit these represent only a minority of child workers (Fallon and Tzannatos 1998; Mendelievich, ed. 1979). Children are also poorly paid rela¬ tive to adults, no matter their skill level. Finally, less than 10 percent of the time do children make their own decision to work (Siddiqi and Patrinos 2000). The International Labour Organisation has been long in the forefront of efforts to abolish child labor. As early as 1919, by adopting Convention Num¬ ber 5, it prohibited work in industrial establishments by children less than fourteen years of age. This convention was then extended to other sectors of the economy. In 1973 Convention Number 138 was adopted banning child labor for children fifteen years of age or younger. This convention contains exceptions and also makes provision for the improvement of working condi¬ tions when children are allowed to work in a transition to the abolition of child labor. It demands that that the minimum age of work not be less than

CHILD LABOR REGULATIONS

201

the age required to complete compulsory schooling. The focus of the Inter¬ national Labour Organisation is on eliminating the worst forms of child la¬ bor, including slavery, debt bondage, prostitution, pornographic and related activities; illegal work such as drug trafficking; work that takes place in un¬ safe, unhealthy, and immoral environments; and labor performed by chil¬ dren less than twelve years of age (Fyfe 1993; International Labour Organisation 1993,2000b). As an intermediate step, the International Labour Organisation is pressing for the implementation of informal education and apprenticeship and skills development programs for child laborers (Lansky 1997). Many countries, although far from the majority, are signatories to the International Labour Organisation conventions on child labor, but much heated debate remains as to the pace and extent to which child labor should be abolished, where, and, indeed, what child labor consists of A fundamental point of contention focuses upon the predicted economic im¬ pact of either reducing the child labor force participation rate or, as an interim solution, improving the working conditions of child laborers. Another point of concern is the potential economic impact of improving the adult wage, which many scholars now agree is a key component to any practical strategy designed to eliminate child labor (Basu 1999a; Lansky 1997). The focus of this chapter is to address the critically important question as to whether or not changes de¬ signed to rid society of child labor can be expected to increase unit production costs, thereby impairing the competitive position of affected economies. Amother important issue to be addressed, although it is not a critical bone of conten¬ tion among those opposed to the abolition or radical reduction of child labor, is the potential impact of eliminating, reducing, or regulating child labor upon fam¬ ily income. With regard to the cost side of the debate, the consensus view, based on the neoclassical model, is that child labor produces output at a relatively lower cost than adult labor. Therefore, firms using child labor or relatively more child labor will be more cost competitive than firms that are more reliant on adult labor. It follows that jurisdictions that limit or ban child labor do so at a cost to their competitive position. Jurisdictions that impose few if any limitations on child labor will be relatively more competitive, especially in sectors where child labor tends to be concentrated. For these reasons, in the relatively more devel¬ oped economies there are calls for trade restrictions against nations where there are few (if any) restrictions on child labor. Alternatively, there are calls for re¬ strictions or bans on child labor across all nations so as to create a level playing field between the more and less developed economies. Otherwise, it is argued, there will invariably be a race to the bottom as all economies strive to maintain their competitive position via child employment rates matching those set by the child-employment-intensive economies or by slashing the rates of labor com¬ pensation and working conditions of adult workers.^

202

CHAPTER 11

Among the less developed economies, the ability to use child labor is viewed as contributing to a competitive advantage in trade with the more developed economies that in turn contributes to economic growth and devel¬ opment. Efforts to restrict trade with jurisdictions employing child labor and efforts to impose significant limits or bans on child labor, or even to improve the working conditions of child workers, are viewed as impediments to the process of economic development in the poorer economies on the part of the more developed economies. Today’s wealthier countries, it is noted, were once characterized by high child labor force participation rates that were typically not any lower than those in contemporary less developed economies. These rates diminished signifi¬ cantly only when the economies became relatively developed (Basu 1999a, 1087—1089). Indeed, there has been a downward trend in the percentage of economically active children aged ten to fourteen, from 28 to 11 percent, from 1950 to 2000. A similar downward trend is found for all regions in the world except Eastern Europe (Basu 1999a, 1087). Moreover, there is a negative rela¬ tionship between the labor force participation rate of children in the 10—14 age cohort and a country’s per capita real income; it is especially evident between the $500 or less (1987 U.S. dollars) per capita income range and the $500— $1,000 range, where the participation rate falls from 30-60 percent to 10-30 percent. There is no trend movement in the participation rate as per capita income increases from $1,000 to $4,000. After the $4,000 per capita income threshold is passed, the child labor force participation rate tends to fall, albeit with significant variation about the mean, as in the above cases, suggesting that variables apart from per capita income play an important role in its deter¬ mination (Fallon and Tzannatos 1998). Given the wide variation in the rate about the mean for each range of per capita income, increases in per capita income do not appear to be a sufficient condition for a reduction in the partici¬ pation rate. Therefore, reductions in the rate cannot be expected to follow natu¬ rally with increases in real per capita or per family income. Nevertheless, although few would argue that child labor is an economic or social good in and of itself, it is considered by many to be better than the opportunity cost of restricting its use in the less developed economies until the process of eco¬ nomic development takes hold and rising per capita incomes dramatically drive down the participation rate. This opportunity cost includes a less competitive economy and less per capita family income. In other words, it is argued by many scholars and public policy experts that controls on child labor in general harm the less developed economies and also, most particularly, the poorest families of these countries. These very strong conclusions, however, critically hinge upon very specific assumptions about the behavior of economic agents. In this chapter, 1 map out the implications of introducing alternative, more

CHILD LABOR REGULATIONS

203

realistic behavioral assumptions for our understanding of interventions against the use of child labor. A behavioral model of the firm, as opposed to the tradi¬ tional neoclassical model, predicts that shifting from low-priced child labor to relatively high-priced adult labor need not increase unit production costs. In other words, a reduction in the percentage of child laborers or even the elimi¬ nation of child labor need not reduce the competitive position of the firm. Therefore, restrictions on child labor need not generate the economic costs predicted by the conventional economic wisdom. In addition, interventions in the economy that serve to improve the working conditions of child laborers need not increase production costs. The behavioral model of the firm predicts that under reasonable assumptions restricting or eliminating child labor yields improvements in productivity as the higher-cost adult labor creates incentives for firms to reduce inefficiencies and adopt and develop more productive, costefficient technologies. The same holds true for efforts to improve the working conditions and wages of child laborers and adult workers alike. A world with less or without child labor or one with higher-priced child and adult workers might be more productive. Such a world may also be characterized by a higher level of per capita and family income, depending on the interaction among child labor force participation, the adult wage rate, and family income.^ More¬ over, I argue that the extent to which child labor is reduced by market forces alone critically depends on the behavioral assumptions underlying the deci¬ sion-making agents within the household. I find that under reasonable assump¬ tions, effective state interventions against child labor or to protect child laborers can increase per capita and even family income without any negative conse¬ quences for the competitive state of the affected economy.

Modeling Child Labor and Production Costs The conventional wisdom predicts that shifting from child to adult labor in¬ creases unit production costs. This follows from the assumption that child la¬ bor is characterized by a lower rate of compensation than adult labor and that any productivity differential in favor of adult labor does not suffice to over¬ come the compensation differential favoring child labor. If, however, the adult labor-child labor productivity differential just compensates for the lower rate of pay to children, shifting between child and adult labor will not affect unit production costs. Assuming, for simplicity, that labor is the only input in the process of production, this argument can be expressed as follows:

AC =

Qa

E

Ok

E

(11.1)

204

CHAPTER 11

Average cost (AC) is a product of the average cost of adult labor (WJ(QJ L^) weighted by its share in employment (L^/E) plus the average cost of child labor {Wf./(Qf./Lf.)) weighted by its share in employment (Lj^/E). Since average cost is a product of the wage rate (fV) which can be taken to repre¬ sent all employment-related costs inclusive of all aspects of the firm’s indus¬ trial relations system) and labor productivity (Q/L), average cost is determined by the relationship between the wage rate and labor productivity, not simply by the level of wages. From Equation 11.1 it is clear that even if child labor can be had at a much lower wage than adult labor, adult labor need not be more costly if it is sufficiently more productive than child labor. In this case, the elimination of child labor, where (Lf^/E) becomes zero and (L^/E) be¬ comes 100 percent, would not increase average production costs. For ex¬ ample, if the adult wage is $10 per hour and the child wage is $5 per hour, average production costs when using adult and child labor would be identi¬ cal if labor productivity is 10 units and 5 units respectively. If adult labor productivity is identical to child labor productivity, the average costs of us¬ ing adult labor would be double those using child labor. On the other hand, if adult labor productivity exceeds 10 units, the average costs of using adult labor will be lower than when using child labor. Whether or not using relatively more child labor yields a more competitive firm ultimately depends on how productive adult labor is compared to child labor relative to their respective rates of compensation. The wage rate alone is no indicator of the relative competitive position of the firm. Only in a world where the productivity of adult and child labor is equal and invariant to differ¬ entials in pay, working eonditions, and physiological characteristics will lowwage child labor yield relatively lower average costs. There is no evidence that such an assumption is empirically valid. Moreover, the efficiency wage literature strongly suggests that labor productivity is affected by conditions of work inclusive of wages. This is apart from the physiological advantages that adults have over children, on average, with regard to productivity. A key com¬ ponent of this line of argument is that labor productivity is affected by the quantity and quality of effort inputted into the process of production and that there is no empirical basis for the conventional assumption that effort per unit of time is fixed at some maximum for all economic agents. The empirically based efficiency wage theory suggests that effort supply is positively affected by the level of labor compensation through its impact upon the nutritional intake and overall health of workers. This is especially true for workers at lower absolute levels of calorie consumption and health. The higher is the wage, the more productive the worker (Bliss and Stem 1978; Dasgupta and Ray 1986; Leibenstein 1957, ch. 8; Stiglitz 1976). Whether or not child labor yields lower average costs than adult labor is therefore an empirical question and cannot

CHILD LABOR REGULATIONS

205

simply be assumed. As already mentioned, it is quite possible, given the exist¬ ence of effort variability, for adult labor to yield the same average costs as child labor. It is important to emphasize that increasing productivity is not a free ride. For workers to become more productive they must be paid more. For this reason, neither the low-wage-low-productivity, child-labor-intensive economy nor the high-wage-diigh-productivity, adult-labor-intensive economy need hold a competitive advantage over the other. Moreover, under these cir¬ cumstances, there is no clear, unequivocal incentive for the employers of child labor to eliminate it volimtarily since they will not gain economically in the process; they must invest in their adult workers so as to remain cost competi¬ tive in a higher-wage-adult-labor-intensive environment. Employers would have to be altruistic in their utility-maximizing objectives for child labor to be vol¬ untarily reduced here. Even if the elimination of child labor need not make affected firms uncompetitive, the case can be made that the elimination or even reduction of child labor would indirectly cause average production costs to increase by way of its impact on aggregate labor supply. Reducing child labor effec¬ tively diminishes the overall supply of labor to the economy. Ceteris pari¬ bus, this has the effect of increasing the wage of adult workers. This argument is illustrated in Figure 11.1, where labor supply is assumed for simplicity to be perfectly inelastic. The adult labor supply curve is given by FB, while the adult and child labor supply curve combined is given by GC. This yields an equilibrium adult wage of Children are assumed to earn only a portion of the adult wage. Any reduction in the supply of child labor shifts the aggre¬ gate labor supply curve to the left, yielding a higher wage rate for both adults and the remaining child laborers. With the elimination of child labor, the aggregate labor supply curve becomes FB and the adult wage becomes W^. According to the conventional wisdom, such wage increases, an indirect re¬ sult of reducing child labor, cause an increase in average costs, thereby nega¬ tively impacting upon the competitiveness of the economy. Although adult workers who remain employed are clearly themselves better off, the increas¬ ing average production costs will harm the economy as a whole. More spe¬ cifically, an economy where child labor is relatively important is said to have a competitive advantage over an economy where child labor is restricted due to the impact that such restrictions have upon the adult and child wage rates. Therefore, it is argued that economies that act to restrict child labor should be afforded protection from economies that are said to take advantage of relatively low-priced child labor. The prediction of higher average costs emanating from higher wage rates critically depends upon the behavioral assumptions made with respect to effort variability. As already discussed, the conventional wisdom holds that effort per

206

CHAPTER 11

Figure 11.1

unit of time is invariant to changes in wages or working conditions. However, there is strong evidence suggesting that effort is a discretionary variable that is affected by the level of labor compensation and the overall work environment, independent of the level of workers’ nutrition and health (chapter 9; AJcaly 1997; Appelbaum and Batt 1994; Barney 1995; Becker and Huselid 1998; Gordon 1996; Ichniowski et al. 1996; Kochan, Katz, and McKersie 1986; Levine 1995; Levine and Tyson 1990; Neal and Tromley 1995; Pfeffer 1995). For effort to be a discretionary variable, the marginal transaction costs of me¬ tering, monitoring, enforcing, and writing effort-specific contracts must ex¬ ceed the marginal benefits. In the behavioral model presented in this book, economic agents typically are not working as hard or as well as they might— they are operating below potential, inside the production possibility frontier. In effect, effort per unit of labor input is not being maximized. Only in a more ideal work environment, where there is a more cooperative and trust-based system of industrial relations, do economic agents tend to perform x-efficiently (chapter 9; Gordon 1996; Ichniowski et al. 1996; King 1995; Levine and Tyson 1990; Miller 1992; Pfeffer 1995; Tomer 1999). In the traditional x-efFiciency literature pioneered by Harvey Leibenstein,

CHILD LABOR REGULATIONS

207

x-inefficiency in production implies, ceteris paribus, higher unit costs, whereas increases in the level of x-efficiency imply lower unit costs. This is evident from Equation 11.1, where changing the level of labor productivity by changing the level of effort changes the average cost of production. In this case, x-inefficient firms survive only if protected through a monopolistic market structure or subsidies and tariffs. Such high-cost firms cannot survive in a competitive environment. In contrast, in the behavioral model of the firm, where effort inputs are modeled as a function of labor compensation and the overall work environment, x-inefficient firms can survive even in a competitive environment. Low-wage firms tend to be relatively x-inefficient, whereas high-wage firms tend to be relatively x-efficient. Higher wages and improvements in working conditions not only induce firms to become more x-efficient to survive in the marketplace, but they also represent a necessary investment in the organizational capital of the firm if productivity is to be increased (Alcaly 1997; Appelbaum and Batt 1994; Barney 1995; B. Becker and Huselid 1998; Gordon 1996; Ichniowski et al. 1996; Tomer 1987,1997). Reducing the level of x-inefficiency is therefore not a free ride. To the extent that the wage differentials between the low- and high-wage firms just offset the productivity differences between them, the low-wage firm will produce at the same average cost as the high-wage firm. In this sense, low wages serve to protect the x-inefficient firms from competitive pressures. On the other hand, higher levels of productivity serve to maintain the competitive¬ ness of high-wage firms. Under these circumstances, as discussed in previous chapters, there might be a unique average cost associated with an array of wage rates and bundles of working conditions (see also Figure 10.3). Overall, both x-efficient and x-inefficient firms can be cost competitive, and one cannot readily predict that low-wage firms will drive high-wage firms out of the mar¬ ket or vice versa. In this scenario, as in the efficiency wage scenario discussed above, there is no reason to expect rational, profit-maximizing firm decision makers to choose the high-wage, x-efficient route since maintaining a lowwage, x-inefficient system need not yield relatively higher costs whereas mov¬ ing to the more x-efficient alternative need not yield any cost advantage. Applied to child labor, the behavioral model of the firm suggests that in a world of effort discretion and x-inefficiency, low-wage child labor need not provide any competitive advantage to firms, nor need high-wage adult labor be a competitive disadvantage. Thus there is no reason to expect rational, profit-maximizing firm decision makers to shift from a child-labor-intensive firm to an adult-intensive firm. However, to the extent that higher wages and better working conditions to adult labor trigger higher levels of x-efficiency, firms dominated by adult labor can be cost competitive with firms domi¬ nated by child labor. This would be true even in economies where the in-

208

CHAPTER 11

comes of both child and adult labor exceed subsistence. Here, changes in wages and working conditions induce productivity changes through their impact on incentives. More generally, high-wage, adult-labor-intensive econo¬ mies can be competitive with low-wage, child-labor-intensive economies. If restrictions are placed on child labor such that the macroeconomic effect is to increase the price of adult labor, through the impact on the aggregate sup¬ ply of labor (as is illustrated in Figure 11.1), one cannot simply predict from theory that this will increase average production costs and make this economy relatively less competitive. To the extent that the higher wages induce man¬ agers and owners to make their firms more x-efFicient, the high-wage firms can remain cost competitive with the lower-wage and relatively x-inefficient firms. Interventions on the labor market that serve to increase wages need not have a negative effect on the firm or the larger economy. In fact, the effects might be positive to the extent that they induce increases in labor productivity and thereby, ceteris paribus, increase per capita output without increasing average costs. The gist of this argument is illustrated in Figure 11.2. In the conventional framework, as the percentage of adult workers increases in a firm, average unit costs increase since the weighted average wage of adult and child labor increases. (It is assumed that the adult wage exceeds the child wage.) This relationship is given by line C*B. However, if labor productivity differen¬ tials between adults and children offset wage differentials between the two groups, increasing the percentage of workers who are adults need not cause average costs to rise, and average costs remain at OC* irrespective of the percentage of workers who are adult. In this case, while the conventional wisdom predicts an average cost gap of AB between a firm employing only child labor and one where QD percent of the workers are adult, in the behav¬ ioral model no such cost gap need exist. On a more general level, increasing the adult wage rate while holding the percentage of workers who are adult constant also need not result in a cost gap between these two types of firms. This is true as long as the increasing adult wage is compensated for by im¬ provements in productivity. This argument can be easily extended to an analysis of the impact upon production costs of introducing improvements to the working conditions of child laborers. These improvements include increased real wages, improve¬ ments to health and safety conditions, and the introduction of educational programs, be they job-specific or of a more general type, where the costs of these programs must be borne by the firm. To the extent that child labor is xinefficient, improvements in working conditions can be expected to induce improvements in productivity. To the extent that improvements in productiv¬ ity compensate for increasing labor costs, unit production costs remain in-

CHILD LABOR REGULATIONS

209

Figure 11.2

variant to improvements in the working conditions. This would be tme up to some point, given technology, when further improvements in working con¬ ditions could no longer be matched by improvements in labor productiv¬ ity—one enters into the domain of diminishing returns—^and unit costs would begin to increase. But until one reaches this threshold, there is no reason to expect that children working under relatively better working conditions and being paid relatively higher wages than other children will produce at rela¬ tively higher unit costs. The case made here is only strengthened if technical change is introduced into the argument. We have argued elsewhere that rates of labor compensa¬ tion can affect the rate of technical change (as opposed to simply the choice of technique). Introducing new technology can help keep unit costs competi¬ tive in the face of increasing costs. Generally speaking, higher rates of labor compensation can induce technical change. However, technological change need not be introduced for low-wage firms to remain competitive if they can produce at the same unit costs as the higher labor standards firms that have opted for the new technology. In addition, the level of x-efficiency can affect the viability of the new technology (chapters 2 and 6; Altman 1998). From

210

CHAPTER 11

this dynamic perspective, higher wages and labor costs that are a product of restrictions upon child labor or improvements in the working conditions of child laborers can serve to increase the level of per capita output through their impact on the rate of technical change, without negatively impacting upon the competitive position of the firm. Child Labor, Per Capita Output, and Family Income Although a fundamentally important result of introducing a behavioral ap¬ proach to an analysis of child labor is that eliminating child labor need not increase average production costs, at the level of the household a key ques¬ tion is whether or not restrictions to child labor result in a reduction in house¬ hold income. More generally, whether improvements in labor productivity yield improvements in per capita output, thereby increasing the potential level of material welfare afforded on average to all members of society, criti¬ cally depends upon the joint and offsetting impact that decreases in the child labor force participation rate and increases in labor productivity have upon the level of per capita output. The relationship among these variables can be isolated in the following equations. Per capita output can be written as:

L

Q

J

(11.2)

p where Q is real output, P is population, and L is labor input. This equation can be rewritten to highlight the potential importance of labor force partici¬ pation, or more specifically the employment to population rate, to per capita real output:

P

L

P

(11.3)

The level of per capita output is directly and positively linked to labor produc¬ tivity {Q/L) and the labor force participation rate (L/P).'^ Ceteris paribus, in¬ creasing the labor force participation rate increases per capita income. The labor force participation rate can be segmented into two components: the adult and child participation rates. Ceteris paribus, increasing the child participation rate has the effect of increasing per capita output through its positive impact on the average labor force participation rate. Reducing the child labor force par¬ ticipation rate has the opposite effect. By definition, there is a trade-off be-

CHILD LABOR REGULATIONS

211

tween the level of per capita output or income and the child labor force partici¬ pation rate, ceteris paribus. However, this static perspective abstracts from the potential impact that a reduction in the child labor force participation rate might have on adult wages and thereby upon labor productivity. Reducing the child labor force participation rate serves to increase per capita output only if this contributes toward increasing labor productivity in a more than offsetting fash¬ ion. In effect, from Equation 11.3, the elasticity of the labor productivity to the labor force participation rate must be greater than one. Clearly, reducing the child labor force participation rate need not reduce per capita output when this reduction positively affects adult labor productivity. The same holds true at the more fundamental level for the relationship between the child labor force participation rate and average or per capita family income. Average family income is a product of the income earned by the adults plus the income brought in by the children. Ceteris paribus, a reduction in the family’s child labor force participation rate yields a fall in average family income. This can be expressed as:

= A

(, ) 1 4

+ K.

where FI is family income, L^/Pf. is the family adult labor force participa¬ tion rate, and is the family child labor force participation rate. Any reduction in the latter yields a reduction in average family income unless accompanied by a sufficient increase in the adult wage rate, ceteris paribus. Increases in the adult labor force participation rate, such as through increases in the adult female participation rate, would also serve to compensate for any loss of income generated by a reduction in the number of children employed. In addition, the adult male labor force participation rate might increase to the extent that men substitute for child laborers whose wage has risen beyond the point where sufficient reductions in x-inefficiency can be generated to compensate for the wage increases. Adult wage increases are to be expected when the fall in the child labor force participation rate is an economywide phenomenon that has the effect of shifting the labor supply curve inward as per Figure 11.1.^ At this point, it is important to note that increases in adult wages need not take place as predicted by a simple supply and demand mod¬ eling of the labor market when the capacity of adult workers to secure higher real wages is inhibited by institutional constraints, such as antiunion legisla¬ tion and policy, the absence of effective minimum wage legislation, and weak or nonexistent unemployment insurance packages. These constraints weaken the bargaining power of adult labor. Therefore, reducing child labor without

212

CHAPTER 11

providing an institutional framework that facilitates the realization of the predicted increases in real adult wages might very well generate some reduc¬ tion in family income. Nevertheless, in this scenario, to the extent that the predicted increases in adult real wages can be secured, there is no unambigu¬ ous trade-off between family income and child labor force participation. Fi¬ nally, to the extent that the reduction in child labor takes place over time and is accompanied by improvements in the wages of children remaining in the labor force, this can further serve to compensate the family for the loss of income caused by either a reduction in the number of children working or a reduction in the hours they work. Conclusion A fundamentally important public policy issue that flows from the analysis presented here is that effective restrictions on child labor and improvements in the working conditions of the remaining employed children need not cause harm to an economy’s competitive position. For this reason, there is no basis in economic theory, nor indeed evidence, to support the argument defending child labor as a necessary means toward economic development. In fact, reducing the child labor force participation rate might even result in improv¬ ing an economy’s average level of real per capita output in a world where xinefficiency exists and where technical change is induced by increasing the price of labor. More generally, there is no theoretical reason to expect that an adult-labor-intensive economy will be less competitive than a child-laborintensive economy or, alternatively put, that a child-labor-intensive economy will be more competitive than a adult-labor-intensive economy. This is in stark contrast to what the conventional economic wisdom predicts. Needless to say, government may still oppose change in the status quo for child labor since eliminating child labor need not provide any direct benefit to decision makers, and, moreover, child labor may be supported by political allies of government who themselves make use of child labor in the production pro¬ cess. Finally, the conventional economic wisdom provides good reason for government to tread carefully on efforts to eliminate child labor when child labor is viewed as a vehicle for economic development. This worldview is tied to the more general analytical perspective of mainstream reasoning that higher wages and improvements in working conditions serve to impede eco¬ nomic development through their positive impact on labor costs. A problem arises when the conventional wisdom is incorrect in its modeling of the eco¬ nomic costs of child labor—^and adult labor, for that matter—and it is this modeling that informs government decision makers. Another question that flows from the analysis presented here is whether one

CHILD LABOR REGULATIONS

213

can expect household decision makers to choose to reduce the family’s child labor participation rate. This issue speaks directly to whether government should intervene to reduce child labor force participation. The first point that needs to be made is that any individual family’s decision to reduce its children’s employ¬ ment can have no impact on the labor market adult wage. Therefore, ceteris paribus, reducing the number of children working within a family unit will un¬ ambiguously reduce average family income. In this context, rational household decision makers cannot be expected to choose to reduce their family’s child labor force participation rate independently and on their own. This is tme espe¬ cially when household decision makers (typically adult males) use child labor as a means to ensure against the risk of income loss in economies when no such insurance is provided by the state and as a means to secure particular, very often subsistence, target levels of income (Grootaert and Kanbur 1995). This behavior is typical where child labor is of numerical consequence, and it locks the economy into a particular equilibrium child labor force participation rate since no indi¬ vidual family unit has the incentive to break out of the equilibrium. Only if the supply of child labor is reduced in the economy as a whole can one expect adult wages to increase as the aggregate supply of labor diminishes. Of course, this would be tme only if institutions exist, such as unions, that facilitate wage in¬ creases. This suggests that some form of government intervention that will result in a reduction in the child labor force is called for since only then can one expect an economywide increase in the adult wage in the immediate future. However, efforts to ban or significantly reduce child labor, independent of other measures, have not had much success. More often than not, children end up working ille¬ gally (Grootaert and Kanbur 1995; Lansky 1997; Mendelievich, ed. 1979). More¬ over, efforts to ban or restrict child labor will have the immediate effect of reducing family income. This suggests that an institutional setting should be developed in economies that are home to child labor that will facilitate in¬ creases in the adult wage rate independent of any change in the child labor force participation rate. Not only would this cushion the negative effect that the reduction in the child labor force participation rate would otherwise have on family income, it would also create the incentives for the household deci¬ sion makers to reduce the child labor force participation rate within their own households (Basu 1999a; Cummingham and Viazzo 1996; Grootaert and Kanbur 1995). This reduction should, in turn, cause a fall in the aggregate supply of labor that should enforce or provoke further increases in the adult wage. In addition, legislated increases to the wages or working conditions of children can provoke employers to reduce the employment of children to the extent that such increases in labor costs are viewed as causing increases in unit costs, as compared to what unit costs would be when relatively more produc¬ tive adults are employed (Grootaert and Kanbur 1995).

214

CHAPTER 11

AJthough it is quite possible that increasing adult real wages will result in a reduction in the child labor force participation rate even in the absence of gov¬ ernment intervention, this assumes that the decision makers behave altmistically toward their children and that they are willing to sacrifice income that their chil¬ dren would otherwise earn. Basu (1999a) argues that assuming altruism of par¬ ents toward their children, as real wages increase, household decision makers will reduce the labor force participation of their children, and once real wages passes some threshold, parents will eliminate their children’s labor market par¬ ticipation in its entirety. In this scenario, there is no reason to expect any volun¬ tary diminution in the child labor force participation rate imless the adult real wage rate increases sulficiently. However, if the desired threshold of real family income increases with increasing real wages—the altmism assumption becomes invalid—there is no reason to expect adult decision makers to voluntarily reduce their children’s labor force participation even if the real wage rate increases (Cummingham and Viazzo 1996). In this case, increasing adult wages and legis¬ lation restricting child labor go hand in hand since unless parents are altruistic and their threshold income does not increase with increasing real income, one cannot expect parents to voluntarily reduce their children’s labor force partieipation rate. Government efforts to restrict child labor in the context of increasing adult wage rates serve to reinforce the private incentives of altruistic households to reduce their child labor force participation rate and to foree nonaltmistic house¬ hold decision makers to reduee their households’ child labor force participation rate since increasing real adult wages alone will not do the trick. The behavioral model suggests that government poliey that provokes or facili¬ tates increases to the adult real wage need not negatively alFect the competitive state of an economy when increased wages provoke increases in the level of xefficiency and the rate of technical change. This is important to any discussion of approaches to eliminate child labor since increasing the adult real wage is regarded by many scholars as a key eomponent of any such effort (Basu 1999a, 1999b; Grootaert and Kanbur 1995; Lansky 1997). In the behavioral model, child labor can be reduced and eventually eliminated without damaging an economy’s long¬ term prospects or reducing the already very low real incomes of those families where most child laborers reside. The opportunity cost of eliminating child labor is not, according to the behavioral model, higher unit produetion costs or lower fam¬ ily income. Eliminating child labor, however, requires that firms and governments invest in higher-wage and more productive economic regimes.^ Notes 1. For a discussion of the difficulties in constructing estimates for child labor, see International Labour Organisation (2000a) and Lansky (1997). Often much lower estimates are used, but these exclude child workers who are less than ten years of age.

CHILD LABOR REGULATIONS

215

2. For a detailed review of the literature on child labor, see Basu (1999a, 1999b); Cummingham and Viazzo (1996); Human Rights Watch (1996); Mendelievich, ed. (1979); and Rodrick (1996). See Sengenberger (1994a, 1994b) and OECD (1996) for a review of the more general issue of labor standards. 3. The argument presented here applies more generally to the relative competitive¬ ness of low- and high-wage regimes, where x-inefficiency exists and where the level of x-efficiency is sensitive to relative labor costs and the rate of technical change is also sensitive to this same variable (chapters 2, 6, and 9). 4. The employment to population rate equals the labor force participation rate only when full employment exists in the economy. 5. There is a possibility, of course, that higher wages, by inducing increased levels of x-efficiency and technological change, will generate long-term increases in the unemployment rate, thereby serving to reduce family income by reducing the house¬ hold employment rate. However, there is no evidence that such a long-term effect is present. In fact, unemployment rates tend to be lowest where rates of technical change, measured by improvements to labor productivity, tend to be greatest. Unemployment rates tend to vary more as a result of differences in macroeconomic policy and savings rates (Maddison 1995). 6. Other key components of strategies to eliminate child labor include access to low-cost quality education, more equitable economic growth, and stabilizing the in¬ come flow of households. It is also argued that increases in the adult female real wage is more strongly correlated with a reduction in the child labor force participation rate than are increases in the adult male real wage (Grootaert and Kanbur 1995).

12

How Discriminatory Pay Inequality Can Persist Even in Competitive Markets

Introduction Pay inequality between women and men for market work continues to be a well-documented central characteristic of the market economy.* Its causes, in particular the extent to which it is a product of labor market discrimina¬ tion, continue to be debated.^ One apparent problem with many of the theories that relate gender pay inequality to labor market discrimination is their difficulty in explaining its persistence over time in the face of competitive pressures. Gary Becker’s (1971) classic work on labor market discrimination has served as the basis for many of these theories, that ultimately rely on the persistence of market imperfections.^ On the other hand, theories that are better able to explain the endurance of pay inequality over time—^as a relatively stable equilibrium solution for the price of labor inputs—have relied largely on supply-side arguments related to human capital formation or the specialization of women in housework.'* Such theories view pay inequality as the payment of differ¬ ent wage rates for labor of differing relative marginal productivities. From this perspective, labor market discrimination is not the ultimate cause of pay inequality. Rather, if discrimination is of any consequence, it is in the social¬ ization of women outside the labor market to engage in particular labor mar¬ ket or household tasks that ultimately result in the development of sex-based labor productivity differences and concomitant gender pay inequality. All of these models of pay inequality have difficulty explaining the per216

PERSISTENCE OF PAY INEQUALITY

217

sistence of pay inequality in the face of competitive pressures and the im¬ provement in information available to economic agents. This is not to say that market imperfections, which of course exist, do not play an active role in maintaining gender pay inequality. However, in theory, it is possible to show that pay inequality caused by discrimination in the marketplace can explain long-run pay inequality even in the face of strong competitive pres¬ sures and improved information. The model developed in this chapter builds on different and more realistic behavioral assumptions than those that underlie the standard economic theo¬ ries of discrimination.^ Its basic proposition is simple: once discrimination leads women to be paid less than men, women become less productive than men. In this case, hiring lower-paid women does not give nondiscriminating employers a competitive advantage. Pay inequality becomes an equilibrium solution that camiot be eliminated by competitive pressures. These results can be generalized to include any pay inequality in which one group is sub¬ ject to discrimination. Discrimination and Pay Inequality Discrimination in the labor market has been interpreted as the payment of different wage rates to equally productive individuals. The individual en¬ gaged in discriminating, with a “taste for discrimination,” as it were, must act as if he or she is willing to pay for preferring one group over another where the two groups have the same productivity characteristics (Becker 1971, 14; 1985, S42). In Gary Becker’s discussion of discrimination, em¬ ployers do not necessarily maximize profits, for discriminating employers must forfeit profits (Becker 1971, 40). In Arrow’s (1973, 1980) articulation and elaboration of the Becker model, employers are utility maximizers, where utility is a positive function of profit and a negative function of psychic costs, expressed in terms of monetary equivalents of employing the individual or group whose employment yields a negative utility. This negative utility rep¬ resents an additional cost—for example, hiring women-—^to a discriminating employer. This negative utility can be expressed by the “coefficient of dis¬ crimination.” For this reason, the exercise in short-run utility maximization, in which the capital stock is held constant, reduces to an attempt to maxi¬ mize the difference between total revenue and total costs. This can be ex¬ pressed as: n^{PxQ)-FW{\ + d)-MW{\ + d),

(12.1)

where n is profits, P is product price, Q is quantity of output, F is female

218

CHAPTER 12

employees, Wis the wage rate, M is male employees, and d is the coefficient of discrimination. When a positive psychic cost is associated with the em¬ ployment of women but not with the employment of men, the time costs to the discriminating employer of employing women whose monetary wage is the same as that of men become greater than the time costs of employing men. In the basic Becker—Arrow model of discrimination one assumes perfect competition in the product market. Men and women are also equally produc¬ tive; the wage rates are given to the employers (by the market, by legislation, or by the negotiation process); and all plants being compared are identical (characterized by the same production function). Given these assumptions, the utility-maximizing employer must equate the marginal cost and marginal revenue product of the last worker employed, where the marginal cost of labor incorporates the marginal disutility of employing a member of a dis¬ criminated group. Moreover, the marginal cost of each employee must be equal. This type of behavior yields the following pay ratio:

W, . W^-W,{d)

(j22)

w,+w,(dy Here, W is the wage rate, M stands for male employees and F for female employees, and d is the coefficient of discrimination. Throughout this chap¬ ter, I use women as my proxy for the discriminated group. When a positive psychic cost is associated with the employment of women and none with the employment of men, women are paid less than men to an extent specified by the coefficient of discrimination, which represents the psychic costs to dis¬ criminating employers of employing women. Graphically, this argument can be expressed by examining the employ¬ ment of women and men separately. Each group of employees is assumed to be characterized by the same marginal revenue product curve (since men and women are assumed to be perfect substitutes), and all discriminators are characterized by the same coefficient of discrimination against women. Nondiscriminating employers associate no positive psychic benefits per se from employing men; their d is zero. In Figure 12.1, MRP^ is the marginal revenue product curve for both women and men. The wage rate and the marginal revenue product are measured along the vertical axis and hours of labor employed along the horizontal axis. If the wage rate of males is given by OA, OLq hours of male labor are employed by the discriminating employer. The same number of hours of male labor is employed by the nondiscriminating employer if no females can be hired. In this case, ceteris paribus, both sets of firms earn the same profits and incur the same production costs. With respect to the employment of women, however, the situation changes. The monetary

PERSISTENCE OF PAY INEQUALITY

219

Figure 12.1

wage of women is given by OB. It is less than OA by AB{d), the psychic costs of employing women, or lVF(d) in Equation 12.2. The discriminating em¬ ployer employs OLq hours of female labor, but the nondiscriminating em¬ ployer employs OLg hours of female labor. The former employer employs fewer women than the latter because the costs of employing women, mon¬ etary and psychic combined, are greater. Since the monetary costs of employing women are the same for both groups of employers, the same unit costs of employing female labor are incurred by both. However, the nondiscriminating employer benefits fi'om the extra profits (BCD) earned by employing the additional hours of female labor {L^^. This in itself is enough to give the nondiscriminating employer a competitive ad¬ vantage over the discriminator. The extra profits can be used to cut the product price or to invest in growth or in new cost-effective technology. But the nondiscriminating employer will employ women as opposed to men as long as they are available at a lower monetary cost. The nondiscriminating employer’s female employees, who serve as substitutes for male employees, provide this employer with both higher profits and lower unit costs. Unit costs will be lower by AB. Herein lies the major competitive advantage of the nondiscriminating

220

CHAPTER 12

employer. In the long run, it eliminates pay inequality between women and men but only if nondiscriminating employers exist or can come to exist in the marketplace. This point has been long recognized in the literature, inclu¬ sive of Becker (1971, 20, 441) and Arrow (1973, 10; 1980, 124, 126). Pay inequality is eliminated when the nondiscriminating employer ex¬ pands plant size or as more nondiscriminating employers enter into the rel¬ evant industries to take advantage of relatively inexpensive female labor. This bids up the price of female labor and bids down the price of male labor. Discriminating firms will be eliminated, or they will employ male labor only once pay equality is achieved—labor markets will become segmented. These results are complicated but not changed in kind once one allows for employ¬ ers to be characterized by different coefficients of discrimination. In this case, employers with the lowest coefficient of discrimination determine the extent of pay inequality in the long run. Such employers will hire the lowerpriced female labor until the pay ratio of female to male wages is equal to that of the lowest coefficient of discrimination. When this is zero, the degree of pay inequality will be zero as well. The extent to which pay inequality is established in this model is then ultimately determined by the coefficient of discrimination of the marginal employers, by the elasticity in the supply of such employers, and by the degree to which unit costs are independent of output. The last is related to the production function of those plants relevant to the employment of the discriminated group. If the production function is linear homogenous, the marginal employer—^the employer with the lowest discrimination coeffi¬ cient—will produce all the output since unit costs would be lowest and inde¬ pendent of output. If production functions are not linear homogenous, the existing marginal employer will not produce all the output, and pay inequal¬ ity need not fall to the level of the marginal employer’s coefficient of dis¬ crimination. However, if a sufficient number of new employers characterized by the marginal employer’s coefficient of discrimination enter the relevant industries in the long run, it is possible for pay inequality to be reduced to its lowest possible rate. Of course, if the coefficient of discrimination is zero, there would be no pay inequality in the long run.^ Given the theory, this does not appear to be an unreasonable long-run scenario.^ Effort Discretion, Discrimination, and Long-Run Pay Inequality To eliminate pay inequalities due to labor market discrimination, the rela¬ tively non-discriminating employers must possess some economic advan¬ tage over their discriminating counterparts. But this need not always be the case. Indeed, in the neoclassical model of pay inequality, such an advantage

'PERSISTENCE OF PAY INEQUALITY

221

exists largely because it is assumed that there is no relationship among labor compensation, the organization of the firm, and the effort of economic agents per unit of time in the production process, and therefore among wages, orga¬ nization, and labor productivity. If such a relationship does exist, changes in wage rates and/or in the organization of the firm would, through labor pro¬ ductivity, affect unit and marginal costs of production. It has long been recognized in the x-efficiency literature (chapters 1 and 2; Button, ed. 1989; Frantz 1988, 1997; Leibenstein 1966, 1974, 1987; Rozen 1991; Tomer 1987) and in the efficiency wage literature (Akerlof 1982, 1984; AJcerlof and Yellen 1986, 1990; Akerlof and Yellen, eds. 1986; Bowles 1985; Bulowand Summers 1986; Leibenstein 1974; Stiglitz 1976,1987; Yellen 1984), both of which are constructed on solid empirical foundations, that labor pro¬ ductivity is not simply a function of technology, capital intensity, and plant size. It is also affected by the quantity and quality of effort per unit of time (effort intensity) contributed to the process of production. Ceteris paribus, the more effort applied, the greater the labor productivity. The extent to which productivity increases as a consequence of effort intensity depends on the elas¬ ticity of labor productivity with respect to changes in effort. Therefore, once effort per unit of time is variable and a function of wages and/or firm organiza¬ tion, labor productivity becomes a function of wages and/or firm organization through the intermediary of effort intensity, and this indirectly affects produc¬ tion costs (chapter 9; Arrow 1980, 126). The introduction of effort variability allows for an explanation of pay inequality that originates with labor market discrimination and is also stable over the long run, even in the face of stringent competitive pressures.^ The Fair Wage Hypothesis, Discrimination, and Pay Inequality An equilibrium theory of long-run discriminatory pay inequality can be built upon the fair wage hypothesis of Akerlof and Yellen (1990).^ This hypothesis is itself derived from a rich sociological and psychological literature on human behavior. Akerlof and Yellen argue that effort varies positively with the wage rate and that the worker will contribute frill effort to the production process only if the wage rate is perceived to be fair. If the wage rate falls below the fair wage, effort intensity falls proportionately; “When people do not get what they deserve, they try to get even” (Akerlof and Yellen 1990,256). The fair wage is defined to be;

W* = —,

(12.3)

where W* is the fair wage, W is the received wage, and e^ is the energy or

222

CHAPTER 12

effort supplied per unit of market time. The fair wage is then simply a particu¬ lar rate of compensation per unit of effort that is deemed fair by the employee. In this model one can assume that the product market is perfectly com¬ petitive and that the employer attempts to choose a wage rate that minimizes the marginal cost of effective labor, Wle^. Such a wage rate is not unique (Akerlof and Yellen 1990,268,274, n. 9; see also chapters 1 and 2, above).*® Any wage rate below the fair wage yields the same minimum cost. As the wage rate falls, effort intensity declines sufficiently so that the marginal cost of effective' labor does not change. In other words, as has been argued through¬ out this book, there is a unique marginal and average cost associated with a range of wage rates, in this case wage rates falling below the fair wage. Moreover, it is assumed that if the employer’s profit is not affected by the wage rate, the employer chooses to pay the worker the fair wage (Akerlof and Yellen 1990, 274). If, however, as I am assuming, there are employers who have a taste for discrimination, they will prefer not to pay a fair wage but will be utility maximizing as in the Becker—Arrow model. Women will be paid less than men. The effort intensity of female employees will fall as a result. This con¬ trasts sharply with the original neoclassical model developed by Becker, in which a lower wage for female employees does not affect effort intensity. By reducing their effort intensities, the lower wage for women reduces their labor productivity, so that the costs and profits associated with female em¬ ployment remain at par with those associated with the employment of the higher-priced men. In this scenario it is feasible for nondiscriminating em¬ ployers, who have a preference for paying a fair wage, to pay women the same wage as men. In equilibrium, therefore, only discriminating employers will continue to pay women a lower wage. However, given the stipulated relationship among wages, effort, and productivity, either a fair or an unfair (discriminatory) wage is consistent with the long-run competitive well-be¬ ing of the firm. X-Efficiency Theory and Pay Inequality If employers are modeled as utility maximizers as opposed to profit maxi¬ mizers, even nondiscriminating employers may pay women less than men in equilibrium. A basic premise of the behavioral model presented in this book and of x-efficiency theory per se is that wage rates can affect productivity through the intermediary of firm organization by affecting effort intensity.'* The behavioral model further assumes that employers maximize utility, not necessarily profits. This would modify Equation 12.1 above by including leisure and stress in the employer’s objective function, with leisure posi-

' PERSISTENCE OF PAY INEQUALITY

223

lively and stress negatively related to utility. To the extent that increasing employees’ effort per unit of time requires less leisure and more stress on the part of employers, they will prefer lower-wage-lower-effort-intensity em¬ ployees so long as this does not make them uncompetitive in terms of costs or profits (chapters 1, 2, and 9). There is no reason to believe that such an employer would want to pay a fairer wage. This stands in marked contrast to the suggestion in the fair wage hypothesis that the employer would prefer to pay a higher wage if this did not affect profits. To the extent employers, including nondiscriminating employers, prefer to pay workers (female or male) lower wages, pay inequalities associated with labor market discrimination would not only be consistent with long-run equilibrium in x-efficiency theory, but they would also be pervasive because lower rates of pay for women would be consistent with utility maximization on the part of both discriminating and nondiscriminating employers. Thus the pay inequalities caused by discrimination can be expected to persist as long as discriminating employers can affect the rate of pay of female em¬ ployees characterized by the same basic productivity as men. The Basic Model Whether one uses the fair wage hypothesis or x-efficiency theory to moti¬ vate this model of long-run pay inequality, the core results remain unchanged. These are driven by the assumption that effort is a variable that is affected by the wage rate and firm organization. Changes in effort intensity affect pro¬ duction costs, which in turn determine the ability of the competitive process to eliminate pay inequality due to labor market discrimination. Average costs can be expressed as: TC Q~

FC+N^W _FC

a

W

~a^(a.’

(12.4)

N where TC is total cost, Q is total output, FC is fixed costs, N is total labor input, W is the labor compensation per unit of labor input (our proxy for which is wage rates), and e is effort intensity (0 < e < 1). Marginal costs can be expressed as;

224

CHAPTER 12

Reducing the wage rate while holding labor productivity constant reduces unit and marginal production costs. This provides a low-wage firm with a com¬ petitive advantage. If, however, as wage rates fall, labor productivity durunishes sufficiently to offset the decline in labor cost, as a result, for example, of a reduc¬ tion in effort intensity, unit and marginal costs will not decline. In this case the employment of relatively low-priced labor need not provide an employer with an economic advantage over another employer who employs high-priced labor. If women are relatively poorly compensated because of discrimination and consequently they reduce their effort intensity (unlike their behavior in the Becker—Arrow world), nondiscriminating firms need not realize a com¬ petitive advantage over discriminating firms by employing women. Once one assumes that productivity is influenced by wage rates and the organiza¬ tion of work, the advantage of the nondiscriminating employer dissipates quickly. From Equations 12.4 and 12.5, to the extent that the fall in female productivity offsets the fall in the wage rate, the nondiscriminating employer has no cost advantage over the discriminating employer. Therefore, the em¬ ployment of women instead of men does not yield a competitive cost advan¬ tage to the nondiscriminating employer. This is illustrated in Figure 12.1, where the discriminator employs OLq of men at wage rate 0^1, earning profits of ZAE. If the nondiscriminating employer hires women at the lower wage rate OB, only OLq of women are employed since the lower wage results in an inward shift of the marginal revenue product curve for women from MRP^ to MRPq, yielding a profit of only WBC, where WBC is not necessarily greater than ZAE. In contrast, in the Becker—Arrow world, where the lower wages paid to women are not expected to negatively affect their pro¬ ductivity, the nondiscriminating employer hires OL^ of women, earning a profit of ZBD, since the relevant labor demand curve would be MRP^ When wage rates or firm organization affect labor productivity through their impact on effort intensity, the employment of women instead of men does not enable nondiscriminating employers to drive the discriminators out of business. Nor do they have the incentive to bid up women’s wages by employing women as opposed to men. In other words, low-wage labor is no longer necessarily cheap labor, and the nondiscriminating employer loses the incentive to employ members of the discriminated group. Likewise, em¬ ployers with the lowest discrimination coefficients do not have the means to reduce pay inequality through the competitive process. Pay inequality caused by discrimination is stable even in the long run and consistent with pay in¬ equality for the same job or for different jobs of the same potential produc¬ tivity. In this sense, whether or not segmented labor markets develop does not affect the extent or the stability of pay inequality. As long as the supply of female labor or that of any discriminated group is large enough to require

PERSISTENCE OF PAY INEQUALITY

225

their employment by discriminating employers at prevailing low-wage rates, pay inequality will persist. Conclusion A basic finding of this chapter is that pay inequality caused by discrimina¬ tion in the labor market can persist even under the assumption of perfect competition in the product market. Once one admits the possibility that ef¬ fort intensity can be a variable input in the production process that is at least partly a positive function of movements in labor compensation, one removes the primary incentive to rid the market economy of pay inequality due to labor market discrimination from the marketplace. This model stands in sharp contrast to standard neoclassical theory, which assumes that effort inputs are fixed at some optimum level irrespective of the wage rate. In this case, if discriminating employers pay the discriminated group less than the nondiscriminated group and regard the marginal cost of the discriminated group to be greater than the actual labor compensation (by the coefficient of discrimination), these employers provide the non¬ discriminating employers with the means to either bankrupt the discrimina¬ tors or bid up the wages of the discriminated group to that of the nondiscriminated group. In my model, the extent to which pay inequality is stable over time depends, of course, on the degree to which effort intensity varies with changes in labor compensation and the extent to which labor productivity varies with effort intensity. This relationship is ultimately an empirical question. The model presented in this chapter allows us to explain the existence of long-run pay inequality that cannot be explained by nonmarket supply-side factors. Moreover, it also suggests that to the extent that labor market dis¬ crimination causes differences in productivity to develop, one should not treat pay inequality that is consistent with productivity differences as neces¬ sarily indicative of an absence of labor market discrimination. Indeed, as demonstrated above, labor market discrimination can in itself be the cause of differences in productivity. This direction of causality runs contrary to what standard theory accepts as normal, given its standard assumption about ef¬ fort intensity. To determine whether pay inequality is a product of labor mar¬ ket discrimination, therefore, one must determine the extent to which differences in productivity between groups are a product of differences in labor compensation, which in turn are a product of labor market discrimina¬ tion.*^ The model presented here, calls into question empirical tests of labor market discrimination based on models assuming that pay differences be¬ tween economic agents that are correlated with productivity differences can-

226

CHAPTER 12

not be a product of labor market discrimination. To know anything it is nec¬ essary to determine more carefully the causal link among discrimination, pay inequality, and productivity differentials. In a world where labor market discrimination exists and effort intensity is affected by labor compensation, one cannot expect market forces, no matter how competitive, to eliminate pay inequality due to discrimination. Further¬ more, reducing market imperfections cannot be expected to eliminate such pay inequality. This is not to say that market pressures cannot contribute to a reduction in pay inequality in the long run. However, to eliminate pay in¬ equality that results from labor market discrimination would require much more than that. Critical to an elimination of this type of pay inequality is the development of mechanisms that would prevent the wages of those who are discriminated against from falling below the wages of others who are char¬ acterized by the same basic productivity or, alternatively, would prevent dis¬ criminating employers from actualizing their preferences. Once such discriminatory preferences are actualized, the market will have a difficult time eliminating the pay inequality they produce. Notes 1. See Gunderson (1989, 47) for more contemporary estimates of gender pay inequality for relatively developed market economies as well as the former USSR. See Altman and Lamontagne (1996) and Goldin (1986, 1990) for historical gender inequality estimates for Canada and the United States respectively. See also Groshen (1991) for a detailed empirical analysis of labor market segmentation within and across firms and occupations. 2. Theories of labor market pay inequality are critically discussed in Bergmann (1989), Cain (1986), Darity (1989), Lamond (1977), and Marshall (1974). 3. In elaborating upon Becker’s influential work, Kenneth Arrow (1973, 1980) re¬ lies on the persistence of imperfect information over time, and thus on the persistence of statistical discrimination, or on the costs of hiring and firing workers (nonconvexities in employment) to explain the fact that the market has not forced the elimination of pay inequality. Lester Thurow (1975) also relies on the persistence of imperfect information and statistical discrimination to explain pay inequality. However, Thurow also argues that wage competition is typically unimportant, even over the long run. This prevents women, for example, from bidding low for what will become a high-wage job. By bidding low, women would be compensating their employers for their negative percep¬ tions of the potential lower productivity of women. Barbara Bergmann’s (1971, 1986) classic research on labor market segmentation argues that discrimination causes the segmentation of labor markets; increasing the labor supply (shifting the labor supply curve outward) in one market and reducing it in another (shifting the labor supply curve inward) from what it would be in a world with an integrated labor market results in pay inequality. However, labor market segmentation must persist over time; the lower paid women must be kept from entering the higher paid labor market dominated by men. See also note 2 above and note 8 below.

PERSISTENCE OF PAY INEQUALITY

227

4. Gary Becker (1985) argues that married women can be expected to be paid less than men due to their specialization in household work in general and child-rearing in particular. This has the effect of leaving women with less effort per unit of time avail¬ able for market work and lower productivity than men per hour worked. Thus women are paid less than men. See also Fuchs (1988), who argues that the portion of pay inequality between men and women that cannot be accounted for in terms of such differences as schooling and work experience can largely be explained in terms of women’s specialization in child care. 5. The model developed in this chapter draws upon the behavioral model devel¬ oped in this book, the fair wage-effort hypothesis (Akerlof and Yellen 1990), and an important paper exploring the significance of effort variability in the workplace by Becker (1985). See also Drago and Heywood (1992). 6. The above argument is drawn from but not identical to Arrow’s (1973, 6-8). 7. Gary Becker very skillfully elucidates some of the dynamics involved in estab¬ lishing a long-run equilibrium rate of pay inequality: “If all firms had the same linear and homogenous production function, firms that discriminated would always have larger unit net costs than firms that did not. . . . The smaller (in absolute value) the DC [discrimination coefficient] of any firm, the less would be its unit net costs. The firm with the smallest DC would produce total output, since it would undersell all others; therefore the equilibrium MDC [market discrimination coefficient] would equal the firm’s DC.... If firms did not have homogenous production functions, unit costs would rise v/ith output, and the firm with the smallest DC would not produce every¬ thing. ... In general, firms with DC’s [^/c] less than the MDC are profitable and tend to expand relative to other firms. The ease with which a firm expands is determined by the relation of unit costs to output; if unit costs are independent of output, expansion is easy; if costs rise sharply with output, expansion is difficult. Firms with small DC’s expand more in comparison with other firms, the less this expansion increases their costs relative to others; hence production conditions facing firms must be important determinants of the MDC” (1971, 44) See also Arrow (1980, 124, 126). 8. In an effort to explain the persistence of gender pay inequality between women and men with the same productivity characteristics, Gary Becker (1985, S43, S49, S52— 53, S55) has argued that labor productivity is affected by effort intensity. His focus is not on intrafim dynamics but rather on the allocation of time and effort to nonmarket activities. A critical assumption of Becker’s model (1985, S45) is that each identical economic agent completely allocates his or her fixed amount of energy among an array of market and nonmarket activities, inclusive of leisure. Married women can be ex¬ pected to be less productive than married men because women choose to devote a dis¬ proportionate amount of their time and effort to household work. Thus less effort is available to married women to devote to market work, resulting in their market-related effort intensity being less than married men’s (Becker 1985, S52—S53). In Becker’s model, effort devoted to market activities is determined by the amount of effort allo¬ cated to nonmarket activities. In equilibrium, married women will be paid less then married men even in the absence of labor market discrimination because they are less productive than their male counterparts. This model allows the economic agent to choose not to exhaust the effort constraint either within each period of time or over his/her life cycle in market activities. But once effort is allocated to nonmarket activities, it is as¬ sumed that the economic agent will be working as hard as possible, or at what one might refer to as the effort possibility frontier. However, an economic agent who works rela¬ tively hard in the household does not necessarily devote relatively less energy to market

228

CHAPTER 12

activities. The married woman can be as productive as the married man if she applies more total effort across all activities than the married man, working as hard as the mar¬ ried man in market activities and harder in household activities. He is left with more residual effort (leisure) than the married woman, and she is operating closer to the effort possibility frontier. Indeed, some empirical evidence derived from 1973 and 1977 U.S. survey data, now exists to support the hypothesis that women who engage in household work do not do so at the expense of the energy devoted to labor market activities (Bielby and Bielby 1988, 1043, 1050, 1055; Bielby 1991, 1001). The allocation of effort be¬ tween market and nonmarket activities does not appear to be the key to an explanation of the persistence of gender pay inequality. 9. M. Reich (1981, 204-215) also develops a model of pay inequality in which effort is variable, but it is quite distinct from the model of pay inequality presented in this chapter. In Reich’s formulation, effort intensity is negatively related to the bar¬ gaining power of workers. Reich assumes that as bargaining power improves, effort intensity declines and this, ceteris paribus, reduces profits. To reduce the bargaining power of workers, employers set out to pay black and white workers different wages, with the white workers being paid a premium above their marginal product. This tends to divide workers and reduce their bargaining power. For this model to work, firms must be integrated by race (or by sex). Otherwise, workers will fail to develop feelings of resentment and antagonism that cause a reduction in their bargaining power. Presumably, labor market segmentation increases the effort intensity among all work¬ ers and increases the overall profits of the discriminatory employers. In this model “unfair” wages seem to have the effect of increasing the productivity of labor. The model also predicts that racist firms should drive out nonracist firms from the market¬ place and that the winning firms should be highly integrated, with a record of poor if not abysmal labor relations. These predictions do not seem to be consistent with the reality of labor market segmentation: fair wages inducing higher productivity and unions having a positive effect on productivity. 10. The traditional efficiency wage theory assumes that profit-maximizing em¬ ployers choose the wage rate, the efficiency wage, that will minimize labor cost per efficiency unit (the real wage divided by effort per unit of time). There is a unique wage rate consistent with profit maximization. Under such assumptions one would not expect profit-maximizing employers, such as nondiscriminating employers, to pay women less than men if the wage paid to men is the efficiency wage. Moreover, discriminating employers, by paying women less than men, will be at a competitive disadvantage. If they pay women less than the efficiency wage, women will supply less effort per unit of time to market work, shifting their marginal revenue product curve inward and causing higher labor and unit costs than would otherwise exist. Pay inequality could not therefore persist in the long run. On this point see Bullow and Summers (1986, 398). 11. See Leibenstein (1966, 1974). See Button, ed. (1989) for a collection of some of Leibenstein’s most important contributions to the subject. See also chapters 1 and 2 for a detailed discussion of x-efficiency theory. See also Leibenstein (1987) and Frantz (1988, 1997). 12. On a related point Gary Becker suggests that if women and men are character¬ ized by the same basic productivity and women are paid less than men due to dis¬ crimination, the lower rate of pay might induce women to specialize in household work, reducing their effort intensity in market work and thereby reducing their pro-

PERSISTENCE OF PAY INEQUALITY

229

ductivity in market work relative to that of men. This would further increase the ex¬ tent of pay inequality. In this case, Becker points out, a decomposition of the pay differential would attribute most of the pay differential to differences in human capi¬ tal. This would be so in spite of the fact that: “the average earnings of men and women would be equal without discrimination.... More generally, discrimination and other causes of sexual differences in basic comparative advantage can be said to explain the entire difference in earnings between men and women, even though differences in human capital may appear to explain most of it” (Becker 1985, S42).

13

When Green Isn’t Mean , The Economics of Environmental Regulations

Introduction To what extent need environmental regulations increase production costs, lower returns to investment, and reduce the level of per capita Gross Domes¬ tic Product (GDP)? Related to this question, can “green” be competitive? The conventional economic wisdom maintains that the opportunity costs of environmental regulations are positive and reasonably high, increasing costs and reducing profits, and thereby negatively affecting growth and the level of per capita GDP. The conventional wisdom does not deny that social ben¬ efits might be generated by such regulations, where pollution represents an externality in production, and that the marginal social benefits might even outweigh the marginal private costs. It denies only that these benefits can be obtained at economically insignificant private costs (Cropper and Oates 1992; Jaffe, Peterson, and Portney 1995; Palmer, Oates, and Portney 1995; Stewart 1993). This is in contrast to a minority view, found largely outside of the economics domain, that environmental controls induce cost offsets such that the opportunity costs predicted by the conventional wisdom are minimal. The minority view, clearly articulated by Michael Porter (1991) and Porter and Van der Linde (1995a, 1995b) and referred to as the Porter Hypothesis, is marginalized by the conventional wisdom, where it is assumed that eco¬ nomic agents are x-efificient in production at all points in time. In other words, economic agents are assumed to be operating along the production possibil¬ ity frontier. From the neoclassical perspective, the opportunities for economic 230

ENVIRONMENTAL REGULATIONS

231

efficiency are exhausted independently of environmental regulation and would be exhausted prior to the introduction of such regulation (Palmer, Oates, and Portney 1995; Jaffe, Peterson, and Portney 1995; Stewart 1993). Following the Axiom of Modest Greed (McCloskey 1990, 112), rational, profit-seeking economic agents would not allow any significant or economi¬ cally meaningful economic opportunities to go unexploited. To wit, ineffi¬ ciencies in the economy of any substantive order of magnitude should not be assumed to be lingering, waiting to be exploited through the actions of regu¬ lators and prescient economic agents. Therefore, forcing firms to become greener, although generating social benefits, necessarily elicits private eco¬ nomic costs. Moreover, any evidence that greener economies appear to be economically vibrant, and therefore viable and competitive, should not be taken as evidence that environmental regulations have induced appropriate cost countervails, as such a deduction runs contrary to the analytical predic¬ tions of the conventional economic wisdom (Jaffe, Peterson, and Portney 1995; Palmer, Oates, and Portney 1995; Stewart 1993).' The objective of this chapter is to develop a behavioral model of the firm where pollution is a joint output of production where both “goods” and “bads” are produced. In this model, x-inefficiency prevails even in a world of per¬ fect product market competition dominated by rational, utility-maximizing economic agents. In this scenario, environmental regulations affect both the level of x-efficiency and the extent of technological change. Therefore, con¬ ditions are established whereby regulations induce cost offsets. Greener firms can be more productive than polluting firms, and greener economies might generate higher levels of per capita output than polluting economies. In other words, both green and polluting firms can be cost competitive and profitable. However, private economic agents cannot be expected to adopt green eco¬ nomic policy independently of regulations since, in this model, there need not be any economic advantage accruing to firms in becoming greener. There is no free ride to becoming green. The induced changes inside of the firm serve largely as a countervail to the higher costs incurred in reducing the level of pollution. Therefore, the induced productivity increases that account for the cost offsets need not generate further benefits to firm owners. For this reason, one might have the simultaneous existence, even in a competitive economy, of green and polluting firms, which is one of the stylized facts of economic life. Contrary to what is argued by the conventional wisdom, eco¬ nomic theory per se does not preclude environmental regulations or environ¬ mentally friendly firm owners motivating increases in firm productivity as cost offsets to the development of environmental friendliness. It is therefore possible for green economies to remain cost competitive and profitable rela¬ tive to pollution-intensive economies. Whether or not green and cost com-

232

CHAPTER 13

petitiveness and profitability are mutually consistent hinges upon whether or not an economy is typically x-inefficient and whether or not technological change is independent of environmental regulation. The Porter Hypothesis and Its Critics The multipart Porter Hypothesis was first put forth in Michael Porter’s (1991) oft-cited Scientific American essay. The least emphasized component of the Porter Hypothesis is that environmental regulation will induce companies to produce new, environmentally friendly products that prove internationally com¬ petitive, adding to an economy’s overall competitiveness. Few critics of the hypothesis deny this possibility. What they deny is that these new products do not add to an economy’s overall costs of production using the preregulation period as a baseline of comparison. Also not in dispute is another component of the Porter Hypothesis that stipulates that the manner in which regulations are implemented will affect their costliness. Porter argues: “Turning environ¬ mental concern into competitive advantage demands that we establish the right kind of regulations. They must stress pollution prevention rather than merely abatement or cleanup. They must not constrain the technology used to achieve them, or else innovation will be stifled. And standards must be sensitive to costs involved and use market incentives to contain them” (1991, 168). What is in dispute is whether or not net private costs would be lower in the absence of pollution regulations. It is the following and key component of the Porter Hypothesis that has engendered the most controversy and criticism. Porter stipulates that con¬ trary to conventional economic theory, the introduction or strengthening of environmental legislation “[does] not inevitably hinder competitive advan¬ tage against foreign rivals; indeed [standards] often enhance it. Tough stand¬ ards trigger iimovation and upgrading.” Moreover, “Exacting standards seem at first blush to raise costs and make firms less competitive, particularly if competitors are from nations with fewer regulations. This may be true if everything stays the same. . . . But everything will not stay the same. Prop¬ erly constructed regulatory standards, which aim at outcomes and not meth¬ ods, will encourage companies to re-engineer their technology. The result in many cases is a process that not only pollutes less but lowers costs and im¬ proves quality” (1991, 168). Thus environmental regulation is expected to induce productivity increases in the pollution-producing firms that will, it appears, offset or even more than offset the unit costs to the firms of such legislation. Nevertheless, “This is not to say that all companies will be happy about tough regulations: increased short-term costs and the need to redesign products and processes are unsettling at the leasf’ (Porter 1991, 168). There-

ENVIRONMENTAL REGULATIONS

233

fore, Porter expects companies to successfully resist, even over the long haul, for short-run reasons, what is to their long-mn economic advantage. This argument is elaborated in some detail elsewhere (Porter and Van der Linde 1995a, 1995b; Van der Linde 1993).^ Porter and Van der Linde con¬ clude that contrary to the mainstream economic view, the economy must be regarded as a dynamic system wherein companies are not always making optimal choices in terms of input combinations and technical change and where, therefore, there are profitable opportunities waiting to be picked up if only there is a regulatory push: “The belief that companies will pick up on profitable opportunities without a regulatory push makes a false assumption about competitive reality—namely, that all profitable opportunities for inno¬ vation have already been discovered, that all managers have perfect informa¬ tion about them, and that organizational incentives have aligned with innovating. In fact, in the real world, managers often have highly incomplete information and limited time and attention” (1995b, 127; see also Porter and Van der linde 1995a, 99). The conventional wisdom therefore exaggerates the true private costs of pollution control by assuming optimal behavior on the part of firm managers and owners in the absence of strict regulatory stand¬ ards and so overlooking the possibility that there are profitable opportunities as a direct consequence of pollution regulations. This line of argument, often making direct reference to the Porter Hypothesis, is adopted by leading ad¬ vocates of pollution regulation (Ehrlich and Ehrlich 1996; Elkington 1998; Von Weizsacker, Lovins, and Lovins 1997). From the perspective of the con¬ ventional wisdom the fundamental dilemma of the Porter Hypothesis is the assumption that corporate leaders will systematically choose not to reduce pollutants in spite of the fact that it is to their economic advantage to do so. The focus of criticism of the Porter Hypothesis is not empirical. Rather, it is based on the belief, rooted in the conventional economic wisdom, that rational economic agents will not persistently forsake profit opportunity in the absence of environmental regulations. There is a consensus, on all sides of the debate, that environmental regulations impose substantial costs to firms and to economies at large (Arrow et al. 1995; Cropper and Oates 1992; Jaffe, Peterson, and Portney 1995; Jorgenson and Wilcoxen 1990; Marsh 1993; OECD 1985; Palmer, Oates, and Portney 1995; Porter and Van der Linde 1995a; Weizsacker, Lovins, and Lovins 1997). These include the cost of plant and equipment, operating expenditures, investment in research and develop¬ ment, and the administrative and legal fees associated with environmental regulations, as well as the indirect costs incurred through the payment of higher input prices from regulated suppliers (assuming that input prices must increase as a result of environmental regulations). Moreover, there are oppor¬ tunity costs of investing in environmentally friendly technologies as opposed

234

CHAPTER 13

to the (assumed) more productive alternative technologies. There are also the costs to govermuent of drawing up, monitoring, and enforcing these regu¬ lations; these costs must ultimately be paid by consumers and firms (Jaffe, Peterson, and Portney 1995, 38—139; Stewart 1993, 2061—2065). Neverthe¬ less, it is agreed that in spite of these induced costs, “Overall, there is little evidence to support the hypothesis that environmental regulations have had a large adverse effect on competitiveness, however that elusive term is de¬ fined. Although the long-run social costs of enviromnental regulations may be significant, including adverse effects on productivity, studies attempting to measure the effects of environmental regulation on net exports, overall trade flows, and plant-location decisions have produced estimates that are either small, statistically insignificant, or not robust to tests of model specifi¬ cation” (Jaffe, Peterson, and Portney 1995, 157; see also Cropper and Oates 1992; Stewart 1993; Tobey 1990). In the literature critical of the Porter Hypothesis, the key reason for environ¬ mental regulation costs not having a negative impact on the economy (includ¬ ing the relatively heavily regulated American economy) is that such costs are small relative to either GNP or total production costs. They amount to no more than 2—3 percent of GNP or total costs, although there is considerable variation about the mean (Cropper and Oates 1992; Jaffe, Peterson, and Portney 1995). For this reason Palmer, Oates, and Portney argue these costs are “sufficiently small (in most instances) to be swamped by international differentials in labor and material costs, capital costs, swings in exchange rates and so on” (1995, 130). In addition, it is argued that competing economies and competing firms in different economies have in place like sets of environmental regulations— this is true even among many developing economies—thereby limiting the competitive advantage that might otherwise accrue to a particular economy or firm. Moreover, multinational companies have been inclined to build environ¬ mentally friendly plants even in locations with poor or nonexistent environ¬ mental regulations. Therefore, argue the Porter Hypothesis detractors, the fact that more stringent environmental regulations have not, for example, damaged America’s competitive position is not evidence for such regulations inducing cost offsets as the Porter Hypothesis would have it. Of course, it is quite pos¬ sible that cost offsets were produced and that these have protected the firms that have invested in pollution abatement. But this possibility is mled out largely for theoretical reasons. The traditional models used to critique the Porter Hypothesis assume that it is not possible for cost offsets to occur that would render the net costs of pollution abatement negligible. In one lucid critique. Palmer, Oates, and Portney (1995, 121—125) argue that what distinguishes the conventional neo¬ classical wisdom from the mindset articulated in the Porter Hypothesis is the

ENVIRONMENTAL REGULATIONS

235

assumption in the latter that economic agents systematically overlook profit¬ able opportunities that would result in both less pollution and lower costs and, moreover, that government can correct for this “market failure” by de¬ signing, implementing, monitoring, and enforcing the appropriate environ¬ mental regulations at a low cost that is well below the private benefits induced by these regulations. This implies that regulations provide the possibility of reducing pollution at no cost as a result of the cost offsets they induce. Pollu¬ tion abatement would then be a free lunch (see also Jaffe, Peterson, and Portney 1995, 154-157; Stewart 1993,2079-2082). Palmer, Oates, and Portney develop a simple model that assumes perfect product market competition, profit maximizing on the part of all polluting firms, and given levels for competitors’ outputs and research and develop¬ ment expenditures. More important, this model is designed such that envi¬ ronmental regulations—increases in the set of constraints on the firm’s choices—necessarily increase the net marginal costs to the firm. The authors conclude: “In this model of innovation in abatement technology, an increase in the stringency of environmental regulations unambiguously makes the polluting firm worse off. Even if the firm can invest in and adopt a new, more efficient abatement technology, if that technology wasn’t worth investing in before, its benefits won’t be enough to raise the company’s profits after the environmental standards are raised, either” (1995, 125). In this model, since no incentives exist for firms to adopt environmentally friendly technology prior to the introduction or toughening up of environmental standards, it is assumed that the private production costs must increase as a consequence of regulations. This argument sits well with the logic put forth in another cri¬ tique of the Porter Hypothesis by Jaffe, Peterson, and Portney: One must be careful when claiming that firms are not operating on their production possibility frontiers [this claim is implicit in the Porter Hypoth¬ esis]: if there are managerial costs to investigating in new production tech¬ nologies, then firms may be efficient even if they do not realize that new, more efficient processes exist until regulations necessitate their adoption. In other words, there may be many efficiency-enhancing ideas that firms could implement if they invested the resources required to search for them. If firms do successfully search in a particular area for beneficial ideas, it will appear ex post that they were acting suboptimally by not having inves¬ tigated this area sooner. But with limited resources, the real question is not whether searching produces new ideas, but whether particular searches that are generated by regulation systematically lead to more or better ideas than searches in which firms would otherwise engage. (1995, 156) In this worldview there are invariably costs of one type or another that

236

CHAPTER 13

make the current production setup achieved by private economic agents op¬ timal, at least from the point of view of the private economic agents. In this case, any forced deviation from the optimal production setup, such as envi¬ ronmental regulation, will incur additional costs to the firm.^ X-Inefficiency and Environmental Regulation In the behavioral model of the firm detailed in this book, it is quite possible for firms to'systematically avoid developing and implementing environmen¬ tally friendly production processes even if these are not expected to generate higher unit production costs or lower rates of profit. This would be true even if product markets were perfectly competitive and even if one assumed ra¬ tional, utility-maximizing behavior on the part of all economic agents. In this scenario there need not be any benefits to firms, as is assumed in the Porter Hypothesis, from engaging in environmentally friendly behavior. Rather, it is assumed only that environmental regulation creates the incen¬ tives for cost offsets to be developed. In this case, the pollution-intensive production setup and the relatively environmentally friendly method of pro¬ duction yield the same average costs and the same rates of return. The essence of the behavioral model of the firm adopted here is that firms are typically x-inefficient. From Leibenstein (1978, 206) x-efficiency is de¬ fined by the outermost production possibility frontier. This is determined by the state of technology and the ideal, relatively cooperative, industrial rela¬ tions system."* X-inefficient firms operate somewhere in the interior. Funda¬ mentally, x-inefficiency implies that firms are producing a lower quantity and quality of output than is technically and economically feasible. There is now considerable evidence in favor of the existence of x-inefficiency (Frantz 1997). As we have discussed above, a necessary condition for the existence of xinefficiency is the existence of effort discretion, which in turn is facilitated by the existence of incomplete contracts (based on their transaction costs) and different behavioral functions, especially between employees and members of the firm hierarchy. In other words, economic agents are assumed to be maxi¬ mizing different functions, in contrast to what the conventional theoretical wis¬ dom implies. As noted, for x-efficiency to be achieved, the firm must invest, at least in the short mn, in its organizational capital (Tomer 1987). Members of the firm hierarchy would also have to work harder and better. Moreover, not all layers of the firm hierarchy will benefit from the superior, more cooperative work culture required to construct a relatively x-efficient firm. Thus achieving x-efficiency clearly involves opportunity costs (both material and nonmate¬ rial) that differ across the spectrum of the firm’s economic agents (chapter 9; Appelbaum and Batt 1994; Gordon 1996; Ichniowski et al. 1996; Klein 1984).

ENVIRONMENTAL REGULATIONS

237

Be this as it may, if effort discretion exists, the quantity and quality of effort is not fixed at some maximum and is affected by the work culture of the firm. Xefficiency then becomes a benchmark for maximum output per unit of input that occurs under ideal circumstances. The Porter Hypothesis can be investigated in light of a behavioral model of the firm where x-inefficiency in production is a possibility. Assume that all economic agents attempt to maximize their utility in a calculating, for¬ ward-looking (rational) manner. Moreover and more specifically, polluting firms are assumed to be profit maximizing. As per the conventional model, it is assumed that given the constraints, firms attempt to equate marginal costs and benefits. However, agents and principals may face different objective constraints. In a world of effort discretion, utility maximization does not imply effort maximization unless the marginal costs of effort maximization equal the marginal benefits. Also assume that in a competitive product mar¬ ket, principals are subject to the binding constraint that unit costs must be competitive. Output is a product of inputs such as capital, labor, land, and technology, as well as work culture. Ceteris paribus, a more effective work culture generates a higher level and quality of output. Assume also that the product market is subject to severe competitive pressures and that firms are not protected from the competitive process in any way. Assume further that firms take regulations and competitors’ output and investment in research and development as exogenously given. Based on the stylized facts of the firm, also assume that for the quantity and/or quality of effort to be increased, the firm must also invest in its organizational infrastructure. These assump¬ tions differ from the standard model adopted by Palmer, Oates and Portney (1995), in their critique of the Porter Hypothesis, only with respect to the assumptions allowing for effort discretion and x-inefficiency. A fundamental question related to the Porter Hypothesis is whether the environmentally friendly firm must be a higher-cost firm than a polluting firm where both firms produce the same output and use the same technology. To address this question we assume that the polluting firm is x-inefficient. This firm produces two goods: the marketed good and pollution, which has no market value. However, pollution abatement incurs a positive production cost. The average cost of producing marketed output to the firm (the private economic cost) is given by: EC OC AC =-+ ■

Q

Q ’

(13.1)

where AC is average costs, EC are the abatement costs of reducing pollution, OC are the other production costs, and Q is output. EC and OC are inclusive

238

CHAPTER 13

of labor and capital costs. Assume that for a polluting firm EC is zero. Ceteris paribus, average costs increase as pollution abatement increases. Equation 13.1 can be transformed into:

Prr X El Pqc AC = -^-+

0

X

e

01

(13.2)

where and Pq^ are the prices of inputs related to the EC and to the OC respectively. £7 and OI are the inputs related to£C and OC respectively. For simplicity, assume that El and OI are composed of labor and capital. Aver¬ age costs are positively affected by increases to the price of inputs. But even if input prices are held constant, ceteris paribus, average costs are positively related to increases in the quantity of inputs used in the production process. Therefore, increasing environmental inputs serves to increase average costs, unless they are offset by some positive impact on output. What is typically assumed, however, is that increasing only 01 will increase output, whereas increases in El serve only to reduce pollution. The latter is a public good with no corresponding private material benefits to the firm. The relationship among average cost and input prices, inputs, and output is illustrated by Equa¬ tion 13.3, which is in turn derived from Equation 13.2:

AC =

^EC

,

Pqc

Q

Q

El

OI

’

(13.3)

Average eost is positively affected by the price and quantity of inputs and is negatively affected by the productivity of these inputs. Assume that for pol¬ lution abatement to take place either El or P^^ or both increase. This need not yield a higher average cost if input productivity increases sufficiently. In this case, not only would inereasing environmental inputs reduce the level of pollution associated with a particular level of output, but it would also serve to increase the firm’s average productivity. If the polluting firm is x-inefificient, the increased costs required to reorga¬ nize the process of production so as to reduce pollutants can create the incen¬ tives—^the need to keep average costs from rising—^necessary to reduce x-inefficiency. The firm that is becoming environmentally friendly either vol¬ untarily or through the force of regulations in effect becomes relatively more x-efficient so as to keep average costs competitive. In other words, the require¬ ment or desire on the part of firm decision makers to become greener forces them to reconfigure their firms into more x-efflcient units of production so as to survive in a competitive market.^ Indeed, the higher productivity and envi-

ENVIRONMENTAL REGULATIONS

239

ronmentally friendly x-efficient firm might be characterized by the same aver¬ age costs as the lower productivity x-inefficient and polluting firm if productiv¬ ity and the costs associated with pollution abatement rise in an offsetting fashion. It would therefore be possible for the x-efficient firm to compete on a basis of more effective and cooperative work cultures that generate the higher levels of productivity required for the environmentally friendly firms to survive and even prosper in a competitive environment. Under these conditions, it is possible for there to exist simultaneously, in a competitive environment, an array of firms producing the same product, each firm characterized by an array of pollution abatement rates and related costs and offsetting input productivities. In this sce¬ nario, the relatively heavily polluting firms will not have a cost advantage over the relatively environmentally friendly firms. Of course, assuming no techno¬ logical change, it is possible that in a world of little x-inefficiency; the costs of reducing pollution will not be offset by reductions in the level of x-inefficiency; this is the world of the conventional neoclassical wisdom. Under these circum¬ stances, efforts to reduce pollution will increase the average costs of the greener firms, and the analytical predictions of the conventional wisdom will be opera¬ tional. This point is illustrated in Figure 13.1, where along the horizontal axis there is an array of firms characterized by different levels of pollution abatement, beginning with zero. In this figure, firms in the 05 range are producing output at the same average cost, OA', as a. result of the costs of pollution abatement being compensated by higher levels of x-efficiency. At a certain level of abatement, given by point B, average costs rise since input productivity can no long increase sufficiently to offset the rising cost of abatement, given technology. There is no competitive advantage accruing to the polluting firm from engaging in a pollution-abatement or green strategy if productivity increases just offset the firm’s pollution-abatement costs, inclusive of costs related to increasing the relative level of a firm’s x-efficiency. Adopting a green strat¬ egy then becomes a matter of preference or taste among the managers and owners of the relatively x-inefficient polluting firms in the absence of envi¬ ronmental regulation. Clearly, becoming greener is not a free ride even when the costs of pollution reduction do not increase average production costs. Unlike in the Porter Hypothesis, it is not stipulated here that adopting an environmentally friendly system of production is profitable in the sense of yielding increased profits or lower unit production costs. Failing to become greener, therefore, does not imply that owners and managers are somehow systematically ignoring economic opportunities for gain in a market economy. In a world of x-inefficiency this dilemma of the Porter Hypothesis would not exist, and environmental regulations would be afforded the opportunity to push the decision makers of firms into adopting greener production pro¬ cesses that in the long run need not generate higher costs or lower profits.

240

CHAPTER 13

Figure 13.1

Induced Technical Change and Environmental Regulation The Porter Hypothesis maintains that firms can go well beyond adjusting the production process within the constraints of the prevailing technology so as to become greener without generating higher costs or lower profits. If we apply a simple behavioral model of induced technical change, it can be shown that the Porter Hypothesis may hold tme without posing the dilemma to the conventional economic wisdom of producing profit opportunities that are that systematically bypassed by the decision makers of the firm. Following in the tradition estab¬ lished by John Hicks (1932; see also Altman 1998; Bins wanger 1978;Habbakkuk 1962; Hayami and Ruttan 1971; Ruttan 1997), I assume that changes in factor prices and relative factor prices affect the pace and path of technological change.

ENVIRONMENTAL REGULATIONS

241

Moreover, in the tradition of this literature increasing unit costs of production, irrespective of changes in relative factor prices, also serve to induce technical change. In the tradition of Leibenstein (1973) it is assumed that the pace of technological change is affected to the extent to which x-inefficiency exists and can be expected to exist in the firm. As before, it is assumed that firms are profit maximizers and that perfect competition in the product market exists.^ In Figure 13.2, capital and labor are the two factor inputs that comprise both Other Inputs (01) and Environmental Inputs (El). The latter are required to reduce pollutants. Capital and labor are mapped out along the vertical and horizontal axes respectively. The initial equilibrium for marketed output is given by point A along isoquant Q^. For pollution abatement to take place more El are required, and therefore more capital and labor are required to produce a given level of marketed output. This is illustrated by a shift out¬ ward of the budget or isocost line from B'C to BC and of the isoquant from to Qq, where the level of output at and are equal. The new equilib¬ rium is given by point E along Q^. In other words, pollution abatement in¬ creases the average cost of marketed output, given by Qq, as more expenditure is required to produce a given level. For average cost to remain unchanged with pollution abatement requires technical change or a reduction in the level of x-inefficiency such that the isoquant shifts from to Q^, where the level of output produced at and is equal. In this case, unit costs need not increase if pollution abatement induces a sufficient amount of technical change or increases in the level of x-efficiency to neutralize the cost of the abate¬ ment. However, such improvements in productivity also typically require and go hand and hand with improving the conditions of labor, resulting in increasing the relative price of labor, which is illustrated by increasing the slope of the isocost line from B'C' to B'C." Without further technological change and or further reductions in the level of x-inefficiency, the profitmaximizing firm will adjust its input combination along ray 2, at of isoquant Q^, so as to minimize unit cost. To produce the initial level of out¬ put given by isoquant translates into producing marketed output, now with less pollution, at a higher unit cost—the isocost curve shifts outward along ray 2 to isoquant iQ,. However, this expected or realized increased cost of pollution abatement might induce further technological change, shifting the isoquant from g, to Q^, where the level of output at to is equal. The firm could then produce the original level of output at the original unit cost since isocost curves B'C and B'C" represent the identical level of ex¬ penditure. The technological changes and reductions in the level of x-inefficiency are discussed here sequentially for illustrative purposes only and can be expected to take place simultaneously in response to the cost increases flowing from pollution abatement.^

242

CHAPTER 13

Figure 13.2

In this scenario, the relatively pollution-intensive firm produces its mar¬ keted output at the same unit costs as the relatively green firm once techno¬ logical change is introduced into the equation. From this perspective, the pollution-intensive firm has no incentive to adopt the new technology since the new technology yields no unit cost advantage. Moreover, the old and new technologies are linked to specific and differential costs reflected in isocost curves B' C to B'C." In other words, pollution-abatement technol¬ ogy requires an investment in time, effort, and capital. In addition, if firms are required to become more x-efficient for the new technology to become

ENVIRONMENTAL REGULATIONS

243

cost competitive, it becomes even more costly to adopt the greener technolo¬ gies. For this reason, there would be no economic imperative, in the absence of regulation, for polluting firms to become greener, unless the preferences of the firms’ decision makers drove them in this direction. In effect, environ¬ mental regulations serve to induce environmentally related technological change. Only if the relatively green firms developed even more advanced technology, given by a shift in the isoquant from Q to Q^, where the quan¬ tity of output at Q equals that at and where the isocost curve shifts in¬ ward from B'C" to B"C", would the unit cost of the greener firm drop below that of the more pollution-intensive firm. In this case, market forces induce the pollution-intensive firms to become greener. At a more general level, Nathan Rosenberg argues with respect to the process of technological change that “there is a threshold level at which the costs of the new technology become competitive with those of the old” (1982a, 27). For this reason, there need not be a widespread adoption of the new technology unless the production costs generated when using the new tech¬ nology fall below those associated with the old. If the threshold level is not passed, only environmental regulation will result in the widespread adoption of the greener technologies. Moreover, if the threshold is not passed, the relatively pollution-intensive firms would not be forsaking economic gain by avoiding greener technologies. Induced technological change adds an¬ other degree of freedom to the capacity of firms to become greener without causing unit production costs to rise or profit rates to fall. But it bears repeat¬ ing that even if greener technologies are cost competitive and do not cause average costs to increase, this in no way implies that the free market will force their adoption by polluting firms as long as these firms remain cost competitive and profitable. If the new technology is known to all firms and is not adopted by the pollution-intensive firms, this represents a type of xinefficiency insofar as the new technology yields a higher level of output than is possible with the old. Although the existence of x-inefficiency and induced technological change provides the possibility of pollution abatement at no increased unit cost, this possibility is contingent upon the elasticity of productivity increases to in¬ creases in pollution abatement. The latter is, of course, an empirical ques¬ tion. However, the extent to which productivity must increase to prevent, for example, unit costs from increasing, depends not only on the percentage increase in pollution-abatement costs, but also on the relative importance of these costs to the firm’s total costs. In reality, abatement costs represent only one component of total costs such that introducing or increasing environ¬ mental regulation need not increase unit production cost by much, even in the absence of cost offsets. In the absence of these offsets, the increase in 2

2

244

CHAPTER 13

average cost is given by the share of abatement costs to total costs multiplied by the percentage change in abatement costs. This is given by; A AC EC ARC -=-X-, AC EC+OC EC

(13.4)

where is average cost, EC are the abatement or environmental costs, and OC are the other costs. At one extreme, if abatement costs represents 100 percent of total costs, the doubling of such costs would increase average cost by 100 percent, and output would have to increase by the same percentage to keep average cost from rising. If the environmental costs constituted 50 per¬ cent of total costs, the doubling of such costs would increase average cost by only 50 percent unless output increased by the same percentage. If the share of environmental costs in the total falls to 3 percent—^which is much closer to the stylized facts—the doubling of such costs results in a 3 percent in¬ crease in average cost and requires only a 3 percent increase in productivity as a cost offset. If the environmental costs increased by only 50 percent, average cost would increase by 1.5 percent, necessitating only a 1.5 percent increase in output as a cost offset. In effect, it is easier for firms to manage environmental regulations when these costs represent only a small percent¬ age of total costs. Needless to say, introducing or tightening such regulations does not imply high or higher production costs, as long as labor productivity can be sufficiently increased. The probability of this occurring through re¬ ductions in the level of x-inefficiency and induced technological change in¬ creases as the share of environmental costs in the total diminishes. And the literature has documented that the share of these costs in the total is rela¬ tively small (Cropper and Oates 1992; Jaffe, Peterson, and Portney 1995). Conclusion A fundamental challenge to the Porter Hypothesis posed by the conventional neoclassical worldview is that it fails the Axiom of Minimum Greed. The conventional wisdom also takes issue with the view that environmental regu¬ lations can induce firms to behave more efficiently than they would other¬ wise in the free market. From the logic of this argument it is assumed that environmental regulation must increase the private economic costs to firms and reduce their profits. Moreover, evidence that such regulation has typi¬ cally not affected the competitive advantage of firms is attributed to factors unrelated to the firm’s capacity to offset the costs of pollution abatement. Contrary to the conventional wisdom, in a world where effort discretion and therefore x-inefficiency exists, the decision makers within firms can be

ENVIRONMENTAL REGULATIONS

245

induced into becoming more efficient and greener via environmental regula¬ tions. Firms can be induced to produce more marketed output (“goods”) and fewer pollutants (“bads”). In the model presented in this chapter, rational, profit-maximizing firms can be expected to avoid greener systems of pro¬ duction and technologies if they serve only to offset the costs of pollutionabatement. On the other hand, if being green yields lower unit costs of production than are generated by the relatively pollution-intensive firms, one would expect, as predicted by conventional theory, that the greener firms will drive their environmentally derelict counterparts out of business. In this case, there would be no need for environmental regulations. However, on theoretical grounds there is no reason not to expect that decision makers will avoid becoming greener if such a transformation is a costly process and serves only to offset private costs, leaving the decision makers no better off eco¬ nomically than they were from the get-go. However, the costs to society from such evasion are not only a higher level of pollution, but also a less efficient (more x-inefficient) economy than is necessary. This is apart from the loss of well-recognized social benefits accruing from pollution abate¬ ment. Under these circumstances environmental regulations serve to induce both a more efficient and greener economy. Notes 1. For the methodological worldview of the Chicago School, see Reder (1982). See also chapter 3. 2. The Porter Hypothesis was articulated in a weaker form in an OECD (1985) study on technical change. 3. Stigler makes a similar point in his critique ofLeibenstein’s x-efficiency theory: “The near-universal tradition in modem economic theory is to postulate a maximum possible output from given quantities of productive inputs—that is the production function—^and to assert that each firm operates on this production frontier as a simple corollary of profit or utility maximization. ... In neoclassical economics, the pro¬ ducer is always at a production frontier, but his frontier may be above or below that of other producers. The procedure allocates the foregone product to some factor, so in turn the owner of that factor will be incited to allocate it correctly” (1976,214—215). 4. There is now an extensive empirical literature relating various forms of coopera¬ tive work cultures to higher levels of productivity (chapter 9; Alcaly 1997; Appelbaum and Batt 1994; Barney 1995; B. Becker and Huselid 1998; Gordon 1996; Ichniowski et al. 1996; Levine and Tyson 1990; Neal and Tromley 1995; Logue and Yates 1999; Pfeffer 1995). 5. This argument, specific to the potential impact of environmental regulations or greener preferences on the part of firm decision makers on the level of x-inelficiency in a competitive environment, is a subset of a more general theory developed in this book, whereby increased costs to the firm potentially induce offsetting increases in productivity. This alternative theoretical framework helps explain the stylized fact of the simultaneous existence over time of both efficient and inefficient economic re-

246

CHAPTER 13

gimes (see chapter 8). A pollution-intensive firm, in this light, can be viewed as a subset of an x-inefficient economic regime that can compete with its more efficient and greener counterparts. 6. This discussion follows the argument developed in chapters 2 and 6 above and Altman (1998). 7. The notion that environmental regulations might induce technological change has been explored by Faucheux and Nicolai (1998).

14

Big Is Not Always Better A Critical Appraisal of the Transaction Cost-Economizing Paradigm

Introduction The new transaction cost economics developed by Williamson (1975,1985) has had major implications for applied macroeconomic theory. What under¬ lies this influence is that Williamson specifies conditions other than tradi¬ tional economies of scale under which vertical and horizontal integration or integration resulting in the formation of conglomerates, either through the construction of new plants or through merger, can be economically benefi¬ cial to society. He argues that integration can result in, and even be a prod¬ uct of, transaction cost economizing on the part of the firm that causes a reduction in unit costs. Williamson thereby posits one more reason why integration, inclusive of mergers, can be a source of economic efficiency. For this reason, even if mergers cannot be justified by any savings in pro¬ duction costs through economies of scale or by downward shifts in the up¬ ward sloping marginal cost curve, they might very well be justified, ex ante, for transaction cost reasons. Thus where the conventional economist of old, who ignores transaction cost problems, might argue that mergers are typi¬ cally a source of economic inefficiency in the absence of evidence of pro¬ duction cost savings (due to their generating the classic deadweight losses), Williamson argues that more often than not integrations/mergers are a source of transaction cost savings that outweigh any resulting deadweight losses. By focusing on transaction costs, Williamson has contributed to seriously weakening, from a theoretical perspective, economic arguments favoring vig247

248

CHAPTER 14

orous antitrust activity. In other words, one might predict from Williamson’s modeling that the private market typically generates firms of efficient size and organization.' Williamson emphasizes that the typieal firm ultimately searches for ways and means with which to minimize transaction costs. In this light, transac¬ tion cost analysis can supplement orthodox macroeconomic theory, all the while remaining consistent with its logical and causal premises. Underlying the new transaction cost economics is the hypothesis that “economizing is central to economic organization” (Williamson 1985, xii).^ The fundamental objective function of the firm is ultimately dominated by the argument of cost minimization. To this behavioral assumption, Williamson recognizes only few exceptions. He also assumes that the efficiency-seeking (cost-mini¬ mizing) firm best explains the changing organizational forms that embodied and embody the firm (Williamson 1975, 2; 1980, 197; 1981, 1564; 1985, 103, 232—236, 273, 387—388, 408).^ Thus a modified orthodox economic theory, which includes transaction cost minimizing as part of the cost-mini¬ mizing objective function of the firm, can explain the changes and develop¬ ments in firm size and hierarchy over time. In this chapter I attempt to determine the extent to which Williamson’s modification of standard theory allows us to argue on a theoretical level that integration is economically beneficial to society, that it reduces the economic costs of production, inclusive of transaction costs, from what they would otherwise be without creating other costs (such as deadweight losses) that exceed the savings in production costs. I focus largely on one issue, that of the implications of Williamson’s analysis for introducing the assumption of noncost-minimizing behavior into the objective function of the firm. This assumption is consistent with the behavioral assumptions of x-efficiency theory. I find that the introduction of x-efficiency theory into an analysis of integrations/mergers weakens the theoretical base of transaction cost theory with respect to its evaluation of the development of the large corporation—^it introduces an additional potential cost of integration that has been largely marginalized by Williamson and by conventional economic analysts in gen¬ eral. However, as we shall see, x-efficiency theory can contribute to an as¬ sessment of integrations/mergers only if, and the extent to which, the creation of the larger corporation diminishes competitive pressures. I also examine the manner in which the institutional framework in which the firm operates affects the economic efficiency of integration. The ques¬ tion is raised as to whether a firm, even if it is completely cost minimizing, is necessarily cost efficient. Inappropriate legal constraints imposed on the firm can result in private economic agents minimizing private production costs at a level higher than could be obtained under different legal constraints. There-

TRANSACTION COST-ECONOMIZING PARADIGM

249

fore, transaction cost—minimizing integration can be economically inefficient given inappropriate legal constraints. Thus an evaluation of laws and institu¬ tions is shown to be critical to an assessment of Williamson’s predictions of the efficiency of large corporations. The introduction of x-efficiency theory and institutional parameters into an analysis of Williamson’s work suggests that the environment in which the firm operates is critical to any theoretical determination of whether the be¬ havior of the firm (more particularly, its economic agents) can be expected to be economically efficient. Both severe competitive pressures and efficient institutions are necessary (but by all means not sufficient) conditions for efficient firm behavior and for an analyst to expect and consequently predict market-induced integrations/mergers to be economically efficient. With few exceptions, Williamson assumes that these two conditions pervade the mar¬ ket economy. To the extent that they do not, my analysis provides more scope and justification for government antitrust activity. Transaction Costs, Competition, and Corporate Bigness Williamson argues that the pursuit of efficiency by the firm is ultimately a product of external pressures. For example, he writes: “The argument relies in a general, background way on the efficiency of competition to perform a sort between more and less efficient modes and shift resources in favor of the former. This seems plausible, especially if the relevant outcomes are those which appear over intervals of five and ten years rather than in the very near term” (1985, 22). Inefficient organizational modes are not tolerated in the marketplace. Only the relatively efficient (cost-minimizing) modes persist. Thus the marketplace by and large is able to assure that the firm acts as an efficiency-seeking entity. Transaction cost economizing, as well as cost econo¬ mizing writ large, is guaranteed by competitive market pressures. It also be¬ comes clear from Williamson’s analysis (1975, 135; 1981, 1564; 1985, 120, 129-130, 236) that less than efficient organizational modes can persist if external pressures (market forces) are relaxed from some unspecified opti¬ mal level. Indeed, it appears that firms need not be efficient unless the sur¬ vival of the firm is threatened. This is particularly true if the firm hierarchy or workers prefer leisure or other nonpecuniary benefits to cost minimiza¬ tion. Most important are the preferences of members of the firm hierarchy, who are the firm’s decision makers. Only the force of competitive pressures can convince economic agents to change the arguments in their objective functions to ones that are consistent with the realization of cost-efficient or¬ ganizational modes. It is possible, therefore, for economic agents to postpone or avoid the

250

CHAPTER 14

development of an efficient corporation (Williamson 1975, 38—39, 54—55, 135; 1985, 129-130, 236-237, 319-320). However, Williamson considers competitive pressures to be typically sufficiently severe to prevent ineffi¬ cient organizational modes from becoming the rule. A key proposition im¬ plicitly underlying Williamson’s line of reasoning is that integration does not significantly reduce competitive pressures. It is assumed that integration does not typically generate barriers to entry that in turn prevent competition from enforcing efficient behavior on economic agents."* It follows that to the ex¬ tent that integration generates a reduction in competitive pressures, the orga¬ nizational modes designed by the firm can be expected to be less than efficient and therefore need not be minimizing transaction and other costs. In this case, integration can result in the traditional deadweight losses plus higher than necessary unit costs. These unit costs can potentially be greater than they would be prior to integration. The extent of the integration-related eco¬ nomic costs can be determined only empirically. But these economic losses can be eliminated or reduced to an insignificant level only when competitive pressures are maximized. In a very substantial manner, transaction cost economizing is not inconsis¬ tent with the traditional critique of the big corporation. Central to the conven¬ tional critique of unbridled mergers and integration is the belief that these acts reduce competitive pressure significantly enough to adversely affect the eco¬ nomic welfare of society. The existence of competitive pressures is also an important ingredient of Williamson’s analysis. According to Williamson, inte¬ gration can generate economic costs to society even if one introduces transac¬ tion cost economizing into the objective function of the firm, to the extent that the development of the larger corporations reduces competitive pressures. X-Efficiency Theory and Corporate Bigness At this point, it is important to consider the implications of x-efficiency theory to an evaluation of Williamson’s perspective on corporate size. Leibenstein (1966, 1973a, 1979) relies heavily upon the assumption that competitive pres¬ sures are such that economic agents are not pressured into minimizing produc¬ tion costs. Thus he assumes that product markets are typically imperfect. External pressures upon the economic agents of the firm are therefore slack. This introduces a degree of freedom to economic agents that allows them to supply less than some optimal level of effort per unit of time to the firm. There¬ fore, the firm’s output per unit of input is less than it would be otherwise. Alternatively, unit costs are higher than they would be otherwise. This devia¬ tion from minimum unit costs or from maximum output is what Leibenstein refers to as x-inefificiency. Moreover, it is critical to Leibenstein’s line of rea-

TRANSACTION COST-ECONOMIZING PARADIGM

251

soning that x-efficient production is a possibility that goes unrealized due to the lack of external pressure on the firm. Therefore, what approaches x-effieiency can be realized under a different economic environment eharacterized by more severe eeonomic pressure. This assumes, of eourse, that the firm’s work culture or work environment, coincident with a more competitive market stmcture, is conducive to x-efficient behavior (ehapters 1, 2, and 9). What makes x-inefificiency a possibility when external pressures are relaxed is the assumption that economic agents typically possess objective functions that are not necessarily dominated by cost-minimizing arguments—^what Williamson also allows for. This assumption holds even for firm decision makers. Only external pressures force economic agents to engage in what approaches cost¬ minimizing behavior. The more the pressure increases, the more the economic agents become cost minimizers. In other words, they input a cost-minimizing level of effort into the process of production. Specifically, it is assumed that even in the absence of an ideal work environment, marginal and average costs are lower in a more eompetitive product market environment because mem¬ bers of the firm hierarchy are forced to work harder and better and to take fewer perks so as to keep eosts competitive. The gist of this argument can be gleaned from the following equation:

AC =

Q

(14.1)

L+H where ACis average cost, is the wage rate, is the rate of pecuniary benefits aecruing to members of the firm hierarchy, Q is output, and L and H are the inputs of labor and members of the firm hierarchy. In this scenario, the only inputs are those of labor and managers and owners. Leibenstein argues that product market production allows the firm hierarchy to increase and decrease their effort inputs, thereby reducing input productivity {Q/ (L + H)). This would increase average costs. X-efficiency theory appears to be consistent with the logic of Williamson’s thinking, except that Williamson assumes optimum external pressure on the firm to be the rule, whereas for Leibenstein such a level of pressure is the exception. Given imperfect prod¬ uct markets or other forms of protective belts around the firm, it would not be inconsistent with Williamson to argue that x-inefficiency exists and can be significant. Thus one by-product of integrations/mergers, if they reduce the external pressures on the firm, can be not only to generate allocative inefficiency, but also x-inefficiency. X-inefficiency is consistent with what would be to Williamson an inefficient organizational mode within the firm, wherein all costs inclusive of transaction costs are not minimized.^

252

CHAPTER 14

At this point, one can conclude that unless market pressures remain suffi¬ ciently severe, integration can result in economic inefficiencies of the allocative and x-ineflficiency type. Therefore, the introduction of x-efficiency theory into a critical evaluation of the economic benefits of corporate size introduces an additional potential cost of integration. Given that integration generates reductions in competitive pressures, the true cost of corporate size would, in all probability, exceed the cost of allocative inefficiency. For this reason, traditional critiques of corporate size have by definition underesti¬ mated the true costs of integration.^ Unless integration results in cost savings stemming from scale econo¬ mies, downward shifts in the marginal costs curve, or savings in transaction costs that exceed the costs in allocative and x-inefficiency, integration is not economically beneficial. Williamson’s theoretical framework does not and cannot unambiguously predict that integration yields net reductions in eco¬ nomic costs. Williamson can argue as strongly as he does in favor of the larger corporation because he assumes that external pressures enforce rela¬ tively strict economizing behavior by economic agents. Of course, it is pos¬ sible that integration yields net benefits to an economy. However, whether it does or not cannot be determined theoretically but only by designing tech¬ niques to measure the costs versus the benefits of the larger firms. Williamson (1980, 197; 1981, 1564; 1985, 129—130, 147—150) recognizes that it is pos¬ sible for the costs of integration to exceed any net benefits. However, he relegates such a possibility to the realm of the improbable. Nevertheless, presently the evidence does not exist to support the view that mergers by and large result in the reduction of unit production costs (see Mueller 1986,226; 229-230; Shepherd 1986, 41^2, 52).

Transaction Costs and X-Inefficiency The assumption of the existence of significant transaction costs underlies Williamson’s positive assessment of the benefits of the larger corporation. Transaction costs can be significant, according to Williamson (1981, 1345— 1349; 1985,45-60), only when there exist at one and the same time bounded rationality (BR), opportunism with guile (OG), and asset specificity (AS). BR refers to the fact that the collection and processing of information can be done only at some cost. OG refers to objective functions of economic agents that contain the proclivity to deceive and cheat as significant arguments. AS refers to assets having an opportunity cost that approaches zero. Therefore, once deployed, assets cannot be redeployed except at a significant cost to their owners or users. Given BR, AS would not yield significant transaction costs given no OG. If there is no OG, one can trust economic agents to pro-

TRANSACTION COST-ECONOMIZING PARADIGM

253

vide information honestly and to adhere to contracts. Therefore, BR would not result in significant data-collection and processing costs. AS would not generate significant costs in the enforcement and adjustment of contracts as one can trust economic agents not to renege on contracts and not to find ways to evade their terms. With OG, BR potentially yields significant trans¬ action costs as it requires economic agents to invest more time in searching for accurate information. This potential is realized, however, only if AS ex¬ ists as well. If assets are not specific, then economic agents need not be concerned with obtaining honest information, as assets can be redeployed at no or negligible cost.^ Imperfect information and breach of contract would not be costly. Therefore, in a world without OG, BR and AS would not yield significant transaction costs. In a world where OG increases, given BR and AS, transaction costs increase. Also, given OG and BR, increases in AS yield increasing transaction costs. When events occur that increase transaction costs, firms, Williamson (1985, 61, 94, 103) argues, search for means to economize. One way of economiz¬ ing is to remove from the market to the firm certain transactions by way of integration/merger. In this way, one avoids the costly process of developing, signing, enforcing, and adjusting contracts among independent economic agents. Instead, economic agents are united under the domain of one eco¬ nomic organization. Williamson clearly recognizes, however, that there are costs to the firm that engages in integration. The more significant of these costs are: the losses in economies of scale, when an integrated firm does not generate a demand sufficient to bring its input supplier (now part of the inte¬ grated firm) to the minimum point of its average total cost curve; losses in economies of scope, when the firm requires more than one input, now sup¬ plied internally, that can be supplied at a lower average cost when supplied outside of the firm; and the costs of replacing contracts among independent economic agents (or among independent firms) with governance structures internal to the integrated firm. These internal governance structures yield monitoring and metering costs to the firm (Williamson 1975,117—130; 1981, 1347; 1985, 90-91, 131, 140). Williamson (1975, 124, 125, 128; 1985, 90, 140) also mentions that integration results in the loss of high-powered incen¬ tives to the economic agents of the larger economic organization that en¬ courage economizing behavior. These incentives are directly related to market pressures, which can be reduced if integration confers upon the firm a cer¬ tain degree of protection from market forces. The loss of high-powered in¬ centives results in firms producing in a noncost-minimizing fashion. Hence unit costs would be higher than they would be otherwise. In effect, firms would be producing x-inefficiently. However, if market pressures are not reduced as a consequence of integra-

254

CHAPTER 14

tion, the x-inefficiency-related costs would not arise and only the other inte¬ gration-related costs would have a bearing upon an assessment of the eco¬ nomic efficiency of integration. Moreover, when market pressures remain intense prior to a decision to integrate, the firm hierarchy will attempt to inte¬ grate only when it is expected that the costs of integration will be less than the savings yielded by the larger economic organization. If, however, market pres¬ sures are less than at some optimal level, the firm hierarchy need not attempt to integrate even if this is expected to generate net transaction-cost savings to the firm. Integration would not occur if the firm hierarchy was interested in maxi¬ mizing nonpecuniary benefits such as leisure. If integration were not expected to maximize such a non-cost-minimizing objective function, it would not be pursued unless costs were too high in the smaller corporation to allow the firm to remain competitive. Therefore, the existence of potential net transactioncost savings through integration is no guarantee that the firm hierarchy will pursue the objective of developing a larger corporation. On the other hand, if the firm hierarchy possesses a profit-maximizing ob¬ jective function, integration might take place even if the costs of integration were expected to exceed all transaction-cost savings. If integration was ex¬ pected to increase the firm’s market power, thereby generating an increase in income to the firm above any integration-related costs, integration would take place. The new-found ability of the larger corporation to raise unit prices com¬ pensates the firm for any expected increase in unit costs. Profit maximization takes place although costs are not being minimized. Ultimately, the reduction in market pressures permits such noneconomizing behavior to transpire. To determine if a larger corporation benefits society, one must determine if it yields lower unit costs to the firm than the smaller corporation and if such lower costs outweigh the costs to society of the allocative inefficiency possibly generated by the larger corporation. The latter costs, of course, are not experienced by the firm because they are a function of the artificially high prices established by the larger corporation as it increases its market power. These higher prices yield, in turn, greater revenue to the firm. The allocative and x-inefficiency costs to society are illustrated in Figure 14.1.^ Assume for simplicity that the x-inefficiency costs incorporate all of the integration-related costs discussed above. The conventional critique of bigness assumes no x-inefficiency, the absence of which is characterized by marginal cost curve MC. Allocative inefficiency is given by ABC. This fol¬ lows from the fact that under the assumption of imperfect product markets, price is greater than marginal cost and output is less than it would be under competitive conditions. X-inefficiency is given by marginal cost curve MC^. Allocative inefficiency is now given by [(A'B'C) + (ABC)-(AJC')], which equals yf'5'JSC. Price {Pm') is greater than when one assumes no x-inelfi-

TRANSACTION COST-ECONOMIZING PARADIGM

255

Figure 14.1

ciency and output {Qm') is lower. The cost of x-inefficiency is given by PcPxB' Y znd measures the cost to society of using more than the necessary amount of inputs to produce a given output. A' B'JA is the portion of allocative inefficiency due to the existence of x-inefficiency. In this sense, there is some overlap between allocative and x-inefficiency. Both sources of ineffi¬ ciencies are given by [{PcPxB'Y) + {A'B'JBC). These social costs consider¬ ably exceed the deadweight losses identified by conventional wisdom, which assumed no x-inefficiency (ABC). Moreover, if one assumed away x-ineffi¬ ciency, even though it exists, then one would observe marginal cost MC^ and assume it to be the x-efficient marginal cost curve even though it is not. One would then assume that allocative inefficiency is only A'B'C.' How¬ ever, if one can identify the existence of x-inefficiency, A'B'C' clearly un¬ derestimates the actual deadweight loss, A' B'JBC. Of course, ^4'5'C' alone excludes the economic costs of bigness stemming from x-inefficiency.^

256

CHAPTER 14

In Figure 14.1, the preintegration scenario is illustrated by , where this marginal cost curve is assumed to incorporate preintegration transaction costs. The conventional wisdom typically ignores such costs. Integration would result in the reduction of transaction costs such that the marginal cost curve shifts from T^T^'Xo MC. If integration is to be economically beneficial to society, the resulting savings in transaction costs {TJP^YT) must exceed the resulting allocative and x-inefficiencies + {A'BJBC)]. For this conditiop to be met is a possibility, but it must ultimately be determined empirically, based upon a theoretical framework that takes into consider¬ ation the possibility of the existence of x-inefficiency and its increase as a consequence of integration. X-Inefficient Institutions One issue raised by Williamson (1975, 20&-212, 217-218, 232, 252; 1980, 197; 1985, 278, 288), although he himself pays little attention to it, concerns the relationship between law and overall institutional constraints and transac¬ tion cost economizing. The private cost-benefit calculus of cost-minimizing economic agents might presume that a particular institutional mode (for ex¬ ample, a particular firm size) is cost efficient in terms of facilitating the mini¬ mization of production and transaction costs. But this might be so only due to the particular institutional constraints under which the firm operates. There¬ fore, what appears to be a firm-specific, transaction cost-minimizing decision to integrate/merge might be only a second-best solution to the failure of gov¬ ernment to establish an institutional setting that minimizes the costs of draw¬ ing up, writing, and enforcing contracts among independent economic agents in the marketplace. Thus under a more efficient institutional setting, even firms whose objective functions were dominated by cost-minimizing arguments would not pursue integration. Transaction costs would be minimized with relatively smaller firms. The private economic calculus of economic agents would tend to result in smaller, as opposed to larger, firms. A particular institutional framework can generate overly large corpora¬ tions and thereby transaction costs that are higher than necessary. Moreover, by encouraging the formation of larger organizations, an inappropriate insti¬ tutional setting can contribute to the creation of market power and thus to the generation of allocative and x-inefficiency. The generation by inefficient in¬ stitutions of higher than necessary transaction costs yields higher than nec¬ essary unit costs that can be considered as one component of x-inefficiency insofar as unit costs are not being minimized.'*^ The remaining x-inefficiency and allocative inefficiency would simply be products of the market power generated by the development of a larger corporation. To the extent that in-

TRANSACTION COST-ECONOMIZING PARADIGM

257

stitutions are inappropriate, it is possible that what appears to be a transac¬ tion cost—economizing integration is in fact one that is not minimizing these costs. Whether size contributes to transaction cost economizing is, once again, an empirical question. It is also important to note that Chandler, to whom Williamson refers to strengthen his relatively optimistic view of corporate size, admits that the legal system of a nation (in his case, the United States of the late nineteenth and early twentieth centuries) contributes to the growth of the large corpora¬ tion (1977, 143, 144, 499; 1980, 17). Nevertheless, Chandler argues that for the most part, mergers were typically the result of firms attempting to mini¬ mize production costs. More recently, Lamoureux (1985,107-108,114—116) has cast a shadow of doubt over Chandler’s interpretation, arguing that con¬ solidations in the United States were established not largely for efficiency reasons but rather to reduce competitive pressures. She argues that the larger corporations were probably no more efficient than their rivals who did not engage in consolidation (1985, 153—154). Hence mergers did not typically mean more economic efficiency." Conclusion Williamson has made an important contribution to the economic analysis of integration by demonstrating that transaction cost economizing can be an im¬ portant factor driving the formation of the larger corporation or justifying its existence as an organizational mode that generates net economic benefits to society. However, there is a theoretical imperative to the logic of Williamson’s reasoning that presumes that integration is typically pursued to economize on transaction costs and that integration typically does economize on these and other costs of production. Critical to Williamson’s view of the firm as an economy-seeking entity is his belief that the creation of larger corporations does not significantly reduce competitive pressures because it is these pres¬ sures that discipline the firm into behaving in an economizing fashion. I have tried to show that integration might generate significant costs in ex¬ cess of any cost reductions to the firm that it might produce, particularly if one allows for the possibility of x-inefficiency as one consequence of integration. Moreover, integration need not take place for cost-economizing reasons. By increasing market power, integration can generate net private economic gains, which can be reason enough for the creation of larger films. But if integration does not significantly reduce competitive pressures, the larger corporation would probably not become increasingly cost inefficient. Laws, on the other hand, might encourage integrations that are not cost efficient. Therefore, even if com¬ petitive pressures remained severe and the larger corporation was a product of

258

CHAPTER 14

transaction cost economizing given a particular set of laws, transaetion costs would have been even lower under a different set of laws that encouraged smaller organizational modes. Ultimately, whether an economie organization is too large and is worthy of antitrust proceedings must be an empirical ques¬ tion, but it must be based upon a theoretical perspective that allows for the existence of what might be the most significant of all costs of corporate size: xinefficiency. One must determine what effects mergers/integrations have had or can be expected to have upon the average costs of firms. Finally, one important caveat must be added to this argument. Corporate size, which reduces competitive pressures, need not necessarily generate x-inefficiency. X-inefficieney is at least partly a funetion of the existence of an antagonistie labor-management—owner environment within the firm (chapter 9; Leibenstein 1987, ch. 5). Where a more cooperative work environment prevails, x-inefficiency need not be a significant by-product of corporate size. However, inereasing eorporate size, by sheltering the firm from competitive pressures, provides economic agents with a means of increasing their income or leisure without inereasing their effort inputs. The alternative to more x-efficient behavior is, in this ease, increasing product prices. For this reason, corporate size and x-ineffi¬ ciency are probably very strongly and positively eausally related, especially when the typical firms engaged in merger activity are lacking in the corporate culture most eonducive to x-efficiency in production. Notes 1. Demsetz (1973, 1—9), presents the view that integration is typically the result of superior economic performance. Williamson (1980, 197; 1981, 1564; 1985, 103, 120,273) in effect adds specificity to the more general argument of Demsetz. Refer to Adams and Brock (1991) for a comprehensive review of the literature on corporate bigness and antitrust economics. It must be emphasized, however, that Williamson clearly recognizes that it is possible that integration can yield net economic losses to society. For this reason, he advocates vigilant antitrust activity to uncover cases where corporate size generates net economic losses (1980, 197; 1981, 1537, 1564; 1985, 100). Finally, Shepherd (1986, 31-32, 41-42, 45-47) and Mueller (1986, 225-230) find, after critically reviewing the literature, that the larger corporations (those with the greater market share) are typically not more cost efficient. These authors point to strong evidence that suggests that such firms are at best no more efficient than the smaller firms. Indeed, most large corporations are much larger than they need to be to realize minimum economies of scale. 2. Williamson also writes: “Neoclassical economics maintains a maximizing ori¬ entation. That is unobjectionable, if all of the relevant costs are recognized” (1985, 45, see also Williamson 1981, 1540, 1551, n. 4, 1564; 1985, 1, 17). 3. Williamson (1980, 195) is critical of Chandler’s (1977) analysis of integration in the sense that Chandler emphasizes technological change as a determinant of inte¬ gration as opposed to transaction cost economizing.

TRANSACTION COST-ECONOMIZING PARADIGM

259

4. See Demsetz (1982,47—57) for a critical discussion of the concept of barriers to entry in economic theory. 5. An inefficient organizational mode can also evolve as a result of a firm’s work¬ ers receiving relatively low wages because low wages relieve the pressure on firm management to produce relatively x-efficiently. Low wages keep the firm’s unit cost from rising excessively even as its input productivity is relatively low. This theme is developed throughout this book. See, for example, chapter 1. A similar point is hinted at by Williamson (1985, 319). 6. Leibenstein (1966,392-415) argues that by focusing on allocative inefficiency, one ignores the more significant cost of monopoly or of imperfections on the product market that would be x-inefficient. See also Balibot, Frantz, and Green (1987). 7. Such a scenario would be more likely in Japan according to Williamson (1985,

122). 8.1 have introduced x-inefficiency into Williamson’s line of argument arbitrarily. Nevertheless, as already mentioned, the existence of x-inefficiency is consistent with Williamson’s behavioral assumptions of the economic agents of the firm—^that own¬ ers and managers need not be concerned entirely or largely with cost minimization. 9. The geometric conceptualization of allocative x-inefificiency presented here is similar to that developed in Balibot, Frantz, and Green (1987), except that their mea¬ sure of allocative and x-inefficiency pertains only to what is equivalent to my ABC and PcPxJB. Accordingly, their estimate of x-inefficiency is less than and their esti¬ mate of allocative inefficiency is greater than what 1 suggest. However, our estimates of the total social cost of monopoly and x-inefficiency are similar. 10. This point is further developed in chapter 7 above and in Thorbecke (1990). 11. Nevertheless, Lamoreaux (1985, 153-154) argues that the development of the larger economic organizations did not generate any significant market power to them, at least in the long run and therefore did not threaten the competitive stmcture of the American economy.

15

Culture as a Determinant of Material Welfare \

Introduction The potential role of values and norms and more generally of cultural factors as causal variables affecting the path and pattern of economic growth and development has become subject to considerable debate in academic circles and among public policy experts. Economic theory, however, especially the dominant neoclassical variant, does not well incorporate cultural factors as independent and causally substantive variables with respect to economic growth and development. Rather, strictly economic variables such as capital stock, technological change, and human capital are touted as the major ex¬ planatory variables affecting growth and development irrespective of cul¬ tural setting. Articulating the cultural setting of an economy, and more specifically of economic agents, is assumed not to add substantively to ei¬ ther the predictive or explanatory power of economic theory. For this reason, the conventional economic wisdom pays little heed to cultural factors in de¬ veloping explanations of economic problems. In this chapter culture is introduced into a behavioral model of the firm as a causal variable in the growth and development process and, more specifically, as a determinant of labor productivity and thereby of the level of gross national product (GNP). In this sense it introduces culture as an additional variable in the production function, not unlike what Max Weber in effect does in his clas¬ sic, The Protestant Ethic and the Spirit of Capitalism (1958). This does not deny the fundamental importance of economic variables to the process of eco¬ nomic growth and development. It is argued only that cultural variables must be treated as one of the potentially important causal variables, and the circum260

MATERIAL WELFARE

261

stances under which culture is of causal significance are articulated. Follow¬ ing from this, a fundamental question in economics of the long-run surviv¬ ability of inefficient economic entities in a competitive environment is addressed. Is it possible for low-productivity economic regimes that are low productivity for cultural reasons to survive in the marketplace? Under what circumstances can one expect such regimes to survive? Only if such circum¬ stances are not exceptional can it be concluded that cultural variables are substantively important to the development and growth process. In a behavioral model of the firm, inappropriate cultures can be a cause of economic inefficiency and appropriate cultural precepts become necessary but not sufficient conditions for economic efficiency. This raises the issue of how economic entities imbued with inefficient cultures and therefore them¬ selves inefficient can survive in the long run. Should not superior economic cultures drive out the inferior? I argue that this need not be the case even if economic agents are assumed to be rational utility maximizers in the Gary Becker sense that individuals do their best to maximize their utility as they conceive it in a consistent and forward-looking manner given the constraints that they face. ^ Although culture can be causally important to both the growth and development process, the adoption of cultural norms and mores most conducive to maximizing productivity and therefore material welfare is far from inevitable. If the appropriate cultural environment to maximize mate¬ rial welfare is inevitable, as is assumed in the conventional wisdom, then the cultural attributes of a society related to economic questions cannot be sub¬ ject to debate. If, on the other hand, the invisible hand does not inevitably and invariably produce efficient cultures, the question of the development and immutability of cultural attributes that either enhance or hinder the de¬ velopment process come front and center, for both analytical and public policy reasons. This would be true for scholars and public policy makers of differ¬ ing ideological perspectives. As Daniel Patrick Moynihan put it, “The cen¬ tral conservative truth is that culture, not politics, determines the success of a society. The central liberal truth is that politics can change a culture and save it from itself’ (cited in Harrison 1992, 1). What Is Culture? One economist examining the role of culture in the growth and development process argues that “It can define group or national value systems, attitudes, religious and other institutions, intellectual achievement, artistic expression, daily behavior customs, lifestyle, and many other circumstances” (Harrison 1992, 9). With respect to the relationship between culture and development, culture can be more narrowly defined as “a coherent system of values, atti-

262

CHAPTER 15

tudes, and institutions that influences individual and social behavior in all dimensions of human experience” (Harrison 1992, 9). Harrison (1992, 10) refers specifically to the sense of community; the rigor of the ethical system; the manner in which authority is exercised in a society; and individual atti¬ tudes toward work, innovation, saving, and profit as critical to an apprecia¬ tion of culture to the process of development. Cultural attributes, for Harrison, can change even in short shrift; therefore, being ahead or behind of the eco¬ nomic game for cultural reasons is no guarantee of future success or failure (see also Harrison and Huntington, eds. 2000). For Thomas Sowell, one of the foremost advocates for the importance of culture to the development process, the most relevant are “those aspects of culture which provide the material requirements of life itself—^the specific skills, general work habits, saving propensities, and attitudes towards educa¬ tion and entrepreneurship—^in short, what economists call ‘human capital’ ” (Sowell 1994, xii). Some cultural attributes are more conducive to economic development than others, and for Sowell, cultural attributes tend to be rela¬ tively immutable even in the long run. Therefore, little can be done to affect the economic standing of individuals or groups because it is a product of their particular cultural attributes. Cultural factors tend to override the economic-political environment in which individuals and groups are situ¬ ated. Individuals are products or prisoners of their cultural heritage, which changes, if at all, slowly over time. Both Harrison and Sowell attempt to explain persistent economic differences among individuals, groups, and na¬ tions that appear to be most strongly correlated to culture and apparently cannot be accounted for by economic factors alone.^ At a more general level Max Weber regarded his long-standing classic study of the potential linkage between religion and capitalism as a modest: “contribution to an understanding of the manner in which ideas become ef¬ fective forces in history” (1958, 90). This does not deny the importance of economic variables to the determination of development. Weber writes: Every such attempt at explanation must, recognizing the fundamental im¬ portance of the economic factor, above all take account of the economic conditions. But at the same time the opposite correlation must not be left out of consideration. For though the development of economic rationalism is partly dependent on rational technique and law, it is at the same time determined by the ability and disposition of men to adopt certain types of practical rational conduct. When these types have been obstructed by spiri¬ tual obstacles, the development of rational economic conduct has also met serious inner resistance. The magical and religious forces, and the ethical ideas of duty based upon them, have always been amongst the most impor¬ tant formative influences on conduct. (1958, 26—27)

MATERIAL WELFARE

263

Ideas or cultural attributes that are most conducive to capitalist development and ultimate improvements in material welfare are not immutable to Weber but are rather subject to change over time, education being a fundamental engine of cultural change (Weber 1958, ch. 2). A particular cultural heritage can either facilitate or impede certain behavior, but culture can change in response to changing circumstances, including the economic environment.^ Weber’s conception of the relationship between culture and economics is notable given this chapter’s focus on the relationship between culture and material welfare. Weber writes “Capitalism is identical with the pursuit of profit, and forever renewed profit, by means of continuous, rational, capital¬ istic enterprise. For it must be so: in a wholly capitalistic enterprise which did not take advantage of its opportunities for profit-making would be doomed to extinction [or at least must not rise]’’ (1958, 17). “We will define a capital¬ istic economic action as one which rests on the expectation of profit by the utilization of opportunities for exchange, that is on (formally) peaceful chances of profit’’ (1958, 17). Moreover, in capitalism everything is done with an eye to profit, and “So far as the transactions are rational, calculation underlie every single action of the partners” (1958, 18-19). Such calculations need not be accurate but simply the best possible given the circumstances, which include, in the language of modem economics, imperfect information, trans¬ action costs, and the level of human capital endowment. Also (Weber 1958, 21—22), rational industrial organization embedded in the marketplace requires the separation of business from the household plus rational bookkeeping. For Weber, rational capitalist behavior requires that individuals pursue the maximization of income or wealth as a “calling” in and of itself. In effect, individuals are instilled with a particular work ethic. Therefore, they contrib¬ ute toward maximizing labor productivity and material welfare. But wealth maximization as a calling is facilitated by certain cultures (or religious ori¬ entations) and deterred by others. According to Weber, though, as capitalism becomes the predominant mode of production, all economic agents are pres¬ sured into becoming wealth maximizers—that is, into opting for the domi¬ nant economic mentality.'* Inefficient cultures cannot survive the onslaught of market forces. The Weberian notion of economic rationality fits into the variant of con¬ ventional economic theory that presumes that rational, utility-maximizing economic agents are at all times in market economies working as hard and as well as possible by dint of market forces. This worldview assumes that eco¬ nomic agents, and therefore the economies of which they are part, are oper¬ ating along the production possibility frontier. They are, in effect, maximizing output per unit of labor by maximizing the quantity and quality of output per unit of time. In neoclassical theory, economic agents maximize output per

264

CHAPTER 15

unit of labor given the implieit behavioral assumption that the eultural pre¬ requisites for efficient output are in place a natura. Ultimately, economic agents who are not productivity maximizers are doomed to extinction, and the economy becomes dominated by the productivity-maximizing economic agents (chapter 3; Reder 1982). The notion that culture can serve to facilitate or impede economic efficiency (productivity maximization) is consistent with the conventional wisdom, al¬ though it pays no heed to the role of culture in generating economic efficiency. This is so since cost minimization (whose flip side is productivity maximiza¬ tion) is assumed to come naturally as a by-product of rational behavior. Ratio¬ nal behavior, joined with long-run competitive market forces, is assumed to force the economy to produce along the production possibility frontier. Cul¬ tural differences therefore cannot explain differences in productivity that are not expected to persist over time in a relatively competitive environment. Econo¬ mies are expected to converge in terms of productivity and wages over time as a consequence of competitive pressures. A critical problem with this analytical prediction is that it has not been realized. Convergence has not taken place in a fashion predicted by the conventional wisdom (Altman 1999c; Baumol 1986; Baumol and Wolff 1988; De Long 1988; Pritchett 1997). Modeling Culture Culture has been explicitly introduced into the modeling of the rational, util¬ ity-maximizing economic agent by Gary Becker as one component of an individual’s social capital, where social capital “incorporates the influence of past actions by peers and others in an individual’s social network and control system” (1998,4). Social capital can have either a positive or a nega¬ tive effect on utility, where the sign of the effect depends on the type of social capital. Culture is the component of social capital that changes at a relatively slow pace and is largely given to individuals over their lifetimes. Individuals have some choice over their social capital and even over its cul¬ tural component to the extent that they can choose the social network or control system of which they are part (Becker 1998, eh. 1). Moreover, social capital is introduced into the economic agent’s preference function and is thereby explicitly introduced as a possible determinant of behavior. Becker’s modeling therefore opens the door to modeling culture as a determinant of effort inputted into the production process, a move that Becker himself does not make. In the behavioral model presented in this book culture counts as a substantive determinant of material welfare. In a behavioral model of the firm, economic agents are assumed to be rational utility maximizers from the perspective of the conventional wisdom

MATERIAL WELFARE

265

(Stigler 1976; G. Becker 1998). In contrast to Leibenstein’s modeling, be¬ having in a manner that is consistent with x-efficient production need not be the standard for rational and utility-maximizing behavior from the perspec¬ tive of workers, managers, or owners given the constraints that they face and the arguments contained in their respective objective functions. Only under very specific industrial relations conditions would it be optimal or utility maximizing for rational economic agents, together, as members of a firm, to choose to perform x-efficiently.^ A fundamental underlying assumption of this model is that effort is a dis¬ cretionary variable in the production process and is affected specifically by the work environment, which includes working conditions and wage rates, and more generally by the larger cultural environment of which economic agents are part. In other words, effort inputted into the process of production is not assumed to be fixed at some maximum, independent of the industrial relations and cultural milieu of the economic agent. Indeed, x-efficiency in production is achieved only in a relatively cooperative work environment that is thought to be the most conducive to workers maximizing their effort inputs (chapter 9). In this model, labor costs, which incorporate wages plus all other costs related to the employment of labor and construction and maintenance of a particular work environment, are positively correlated with labor productivity—^an em¬ pirically based assumption rooted in the x-efficiency and efficiency wage lit¬ erature (chapter 9).^ In this case, lower x-inefficient levels of output need not result in higher vmit costs, nor need higher, more x-eflficient levels of produc¬ tion result in lower unit costs. In other words, higher labor costs offset higher productivity, whereas lower labor costs compensate for lower productivity. Under these assumptions it is possible for the x-inefficient firm to remain com¬ petitive, in terms of average costs, at relatively low wage rates when wages are low enough to compensate for relatively low levels of productivity. This point is highlighted in Equation 15.2, where for simplicity it is assumed that labor is the only input into the production function:

L where AC is average costs and w is rate of labor compensation; L is labor input; and Q is output. Average costs are determined by labor productivity and the wage rate. Ceteris paribus, the lower is labor productivity (the less xefficient is the firm) the higher are the average or unit costs of production. But if wages and productivity are both relatively low, there is no reason for x-inefficient firms to be uncompetitive.

266

CHAPTER 15

In other words, firms comprised of utility-maximizing economic agents can remain competitive in long-run equilibrium irrespective of their level of x-efficiency and even in the absence of protection proffered by monopolistic market structures, tariffs, subsidies, transportation costs, and the like. More¬ over, under these conditions, members of the firm hierarchy do not gain materially from increasing or reducing the level of x-efficiency since changes in labor benefits both motivate changes in the level of x-efficiency and neu¬ tralize any impact that these changes might otherwise have on unit costs and profits. Of course, the extent to which changes in labor costs neutralize op¬ posite changes in productivity is an empirical question and is related to the structure of the production function. This is illustrated in Figure 15.1, where a unique unit labor cost (OC*) is associated with an array of wage rate or labor compensation packages up to a point (W*), beyond which the quantity and quality of effort cannot be increased sufficiently to compensate for in¬ creased labor costs.^ The introduction of technical change into this model only serves to reinforce these results. When technological change is induced by changes in the cost of labor, the ability of high-wage firms to survive in a competitive market is enhanced by increasing the firm’s productivity so as to compensate for the increasing cost of labor (chapters 2 and 6; Altman 1998). In this behavioral model, coupling x-efficiency and culture, we are assuming that the relatively efficient cultural regime is causally related to the relatively high-wage economy, as illustrated in Figure 15.1. In this diagram an array of cultural regimes is coupled with an array of wage rates, and these yield a unique unit cost up to point W*. These various cultural regimes are cost competitive because of the differential levels of x-efficiency with which they are causally related. The more efficient cultural regime would dominate all others if it yields a relatively lower unit cost. In this specification of a behavioral model of the economic agent and the firm, cultural factors can be explicitly introduced into the production func¬ tion to help explain not only the productivity and differentials among econo¬ mies embodying different cultural regimes, but also those among firms, where size can range from one to a multiperson firm of a size consistent with a particular production function. Economies can comprise both relatively xefficient and x-inefficient firms predicated upon different sets of values and norms. These values and norms are part and parcel of an individual’s social capital and can be either firm-specific or part of an individual’s larger social envirorunent. As already mentioned, the importance of culture as a determi¬ nant of success and failure in the marketplace is quite consistent with the conventional wisdom. Only the conventional wisdom for the most part as¬ sumes that the Weberian work ethic is a natural attribute of men and women and therefore not worth much talking about.

MATERIAL WELFARE

267

Figure 15.1

Unit costs

Wage Rates and Related Cultural Regimes

In this model, cultural factors become important as either facilitators or impediments to x-efficient production since they affect the choices made by the economic agent with respect to the quantity and quality of work effort.^ X-efficient production would not be possible without the appropriate cul¬ tural environment in place. Under different sets of norms and mores utilitymaximizing economic agents choose to work at different levels of efficiency. X-efificient behavior, consistent with the Weberian work ethic, is only one possible long-run competitive result. To the extent that cultural variables are determinants of work effort, culture is important to economic well-being since it can contribute toward making economic agents more or less produc¬ tive and therefore the economy of which they are part. Culture thereby af¬ fects whether or not an economy operates along its production possibility frontier. In a word, culture can affect the level of per capita real output pro¬ duced in an economy and therefore the level of per capita material wealth as well as differences in the per capita wealth achieved by different economies at any given time. Efficient cultures serve to shift outward the familiar pro¬ duction possibility frontier. In this type of modeling, unlike what is analytically predicted by Weber

268

CHAPTER 15

or the conventional wisdom, the more efficient economies need not prevail over the less efficient economies. This point has been well made by Douglas North (1990; see also Olson 1996), who argues that efficient institutions need not dominate less efficient institutions for transaction cost reasons. The less efficient economies would include societies where wealth creation is built upon violence (Lane 1958) or slavery or serfdom (Domar 1970), as well as democratic societies built upon low wages and low productivity (chap¬ ters 1, 2, 6, and 8; Gordon 1996). In the model presented here, the x-inefFicient economies and corresponding cultures can remain competitive even under conditions of perfect product market competition. The x-inefficient economy, built upon a low-wage regime, passes the neoclassical survival test, as does the x-efficient firm. Market forces therefore do not necessarily result in economic agents adopting the Weberian work ethic.^ Culture is important when different cultural regimes, having differential and independent effects on the x-efhciency of their respective economies, can sur¬ vive simultaneously over time. Market forces need not drive all economic agents, and therefore the economies of which they are part, into adopting norms and mores that are maximally conducive to x-efficient production. One culture can overwhelm another in long-run competitive equilibrium only if it serves to in¬ crease productivity in one economy such that the unit cost falls below the unit cost for like output produced in a competing and relatively x-inefhcient economy under a different cultural regime. The latter approaches the world of the conven¬ tional economic wisdom. The x-inefficient economy can remain intact in the face of unbridled competition from the more efficient economy in the absence of cultural change only if labor costs can be reduced without a compensating de¬ crease in labor productivity. If one relaxes the assumption of competitive prod¬ uct markets, the x-inefficient economy along with its concomitant culture can survive if it is protected by monopolistic structures, tariffs, subsidies, and the like. In a competitive environment, the opportunity cost of maintaining economi¬ cally inefficient cultures is a lower level of per capita material welfare and more specifically lower rates of labor compensation broadly defined. The latter are required to keep the x-inefficient economy cost competitive. Conclusion Conventional economic reasoning has paid little heed to culture as a determi¬ nant of wealth creation because it is assumed that cultural regimes that are not consistent with economic efficiency will not survive in the market at least over the long haul. However, in the behavioral model presented in this chapter rea¬ sonable conditions were established whereby cultures conducive to either xefficiency or x-inefficiency in production persist over the long run. The more

MATERIAL WELFARE

269

efficient economic regime and the cultural infrastmcture within which it is embedded need not possess a cost advantage over the less efficient regime. For this reason, culture can affect the level of economic well-being achieved by an economy or a firm. Therefore, the behavioral model of the economic agent and the firm suggests that there are no good theoretical reasons to expect that the facets of contrasting cultures that are most closely related to work performance need converge toward representations consistent with x-efficient production, as predicted by the conventional wisdom. In such a world, culture is of sub¬ stantive importance, and therefore understanding the intricacies of different cultures can contribute towards a better understanding of the economy. If a particular regime generates a x-inefficient economy, the opportunity cost of not engaging in some form of cultural change that is more conducive to increasing the economy’s level of x-efficiency is the reduction of the average level of material well-being of members of such an economy. In other words, it might be possible to achieve higher levels of material well-being only at the expense of changing the values and norms of a firm or the larger society so that they conform to those of the more x-efficient economy. This follows if a par¬ ticular set of norms and values are necessary to the achievement of a particular level of x-efficiency. In this sense, increasing the efficiency of a firm or society requires choosing one set of norms and values as opposed to some other. This is not to say that a utility-maximizing individual will choose the more x-effi¬ cient society, especially if he or she places a relatively heavy weight on the cultural regime most consistent with the relatively x-inefficient economy. The preference for the x-inefficient cultural regime can be realized over the long term in a competitive environment, however, only if a low-wage regime is accepted as a direct complement of such a regime. A social dilemma arises if individuals want both the cultural milieu consistent with x-inefficiency and the material benefits that flow from an x-efficient economy. Such a choice is not globally possible given the trade-off between a desired cultural regime and the level of x-efficiency. For this reason alone it is fundamentally important to appreciate the opportunity cost of choosing one cultural regime over another. To the extent that culture matters, improvements in material well-being might very well come at the cost of cultural change. Notes 1. Becker’s (1998) concept of constrained utility maximization is consistent with the standard neoclassical fare, although his conceptualization of the question incor¬ porates into the objective function of the economic agent variables such as habits, cultures, peer pressures, childhood, altruism, addiction, and other forms of individual experience and social interaction typically ignored in the traditional modeling of the economic agent.

270

CHAPTER 15

2. See also Ayal (1963), Hampden-Tumer and Trompenaar (1993), Hofstede (1984), Landes (1998), and Maital and Sharabani (1997) for notable efforts to docu¬ ment and assess the importance of culture to the process of economic development. 3. Himmelfarb discusses the importance of the individual as an independent determinant of history in the context of the various deterministic methodological approaches to history where individual behavior is simply a product of circumstance, be it economic or sociological. This obliterates or at least denigrates free will as a causal factor of historical events. The same can be said of the other disciplines, in¬ cluding economics. Himmelfarb points to Alexis de Tocqueville who, writing in the nineteenth century, comments: “Once the trace of the influence of individuals on the nations has been lost, we are often left with the sight of the world moving without anyone moving it. As it becomes extremely difficult to discern and analyze the rea¬ sons which, acting separately on the will of each citizen, concur in the end to produce movement in the whole mass, one is tempted to believe that this movement is not voluntary and that societies unconsciously obey some superior dominating force.... Not content to show how events have occurred, they [democratic historians in America] pride themselves on proving that they could not have happened differently. They see a nation that has reached a certain point in its history, and they assert that it was bound to have followed the path that led it there. That is easier than demonstrating that it might have taken a better road” (Himmelfarb 1994a, 42). 4. See Fumham (1990) for a detailed discussion of the psychological treatment of the Weberian notion of the work ethic and its implications for economic processes. 5. Leibenstein (1978, 1987) assumes that behavior that deviates from the neo¬ classical ideal of x-efficient production is only quasi-rational, while the neoclassical ideal typically assumes that effort discretion does not exist and that effort is fixed at some maximum level in terms of its quantity and quality. 6. Unlike the efficiency wage literature, I do not assume that a unique wage rate generates a unique and maximum level of labor productivity that in turn yields a unique and minimum average cost of production (Akerlof and Yellen 1986). A lead¬ ing efficiency wage scholar, J.E. Stiglitz (1987; see also Akerlof and Yellen 1990), however, recognizes that it is possible for there to be an array of wage rates consistent with a unique unit cost when effort and therefore productivity change sufficiently to just compensate for changes in the wage rate. There is no evidence to suggest that the latter is not the most appropriate assumption. The founding fathers of economics, led by Adam Smith, also argue that wage rates and productivity are positively correlated (N.G. Marshall 1998). 7. Stiglitz (1987) further elaborates upon this point. 8. Buchanan (1994) argues that societies with a stronger work ethic yield benefits that go beyond those achieved by the individual. He argues that if an individual works more hours, this increases his own output and thereby market size. This in turn yields more specialization and thereby increases productivity throughout the economy. This assumes increasing returns to market size and thereby externalities to an individual’s income-leisure choice decision. In terms of the model presented here, a more x-eflficient work ethic yields a greater productivity impact than would be predicted in a world of constant returns. 9. See chapter 3 for a discussion of the survival principle and for a reconstruction of this principle that allows for the long-run equilibrium existence of inefficient firms even in a competitive environment.

References

Abramovitz, M. 1952. “Economics of Growth.” In Haley, ed. 1952, 132—178. -. 1986. “Catching Up, Forging Ahead, and Falling Behind.” Journal of Eco¬ nomic History 46: 385-406. Adams, W., and J.W. Brock. 1991. Antitrust Economics on Trial: A Dialogue on the New Laissez-Faire. Princeton, N.J.: Princeton University Press. Aghion, P., and P. Howitt. 1998. Endogeneous Growth Theory. Cambridge, Mass.: MIT Press. Aghion, R, E. Caroli, and C. Garcia-Penalosa. 1999. “Inequality and Economic Growth: The Perspective of the New Growth Theories.” Journal of Economic Literature 37:1615-1660. Akerlof, G.A. 1982. “Labor Contracts as Partial Gift Exchange.” Quarterly Journal of Economics 97: 543—569. -. 1984. “Gift Exchange and Efficiency—Wage Theory: Four Views.” 4/nmcan Economic Review, Papers and Proceedings 74: 79—83. Akerlof, G.A., and J.L. Yellen. 1986. “Introduction.” In Akerlof and Yellen, eds. 1986, 1-21. -. 1990. “The Fair Wage Hypothesis and Unemployment.” Quarterly Journal of Economics 105: 255—283. -, eds. 1986. Efficiency Wage Models of the Labor Market. Cambridge and New York: Cambridge University Press. Albert, M. 1993. Capitalism V5. Capitalism: How America’s Obsession with Indi¬ vidual Achievement and Short-Term Profit Has Led It to the Brink of Collapse.

New York: Four Walls Eight Windows. Albrecht, J.W., and S.B. Vroman. 1998. “Nash Equilibrium Efficiency Wage Distri¬ butions.” International Economic Review 39: 183—203. Alcaly, R.E. 1997. “Reinventing the Corporation.” New York Review of Books 40: 38-45. Alchian, A.A. 1950. “Uncertainty, Evolution, and Economic Theory.” Journa/of Political Economy 58:211—221. Alchian, A.A., and H. Demsetz. 1972. “Production, Information Costs, and Economic Organization.” American Economic Review 62: 777—795. Altman, M. 1988. “Economic Development with High Wages: An Historical Per¬ spective.” Explorations in Economic History 25: 198—224.

271

272

REFERENCES

-. 1990. “A Critical Appraisal of Corporate Bigness and the Transactions Cost Economizing Paradigm.” In Frantz, Singh, and Gerber, eds. 1990, 217—232. -. 1992. “The Economics of Exogenous Increases in Wage Rates in a Behavioral/X-Efficiency Model of the Firm.” Review of Social Economy 50: 163—192. -. 1993. “Human Agency as a Determinant of Material Welfare.” Journal of Socio-Economics 22: 199—218. -. 1996a. Human Agency and Material Welfare: Revisions in Microeconomics and Their Implications for Public Policy. Boston, Dordrecht, and London: Kluwer Academic Press. -. 1996b. “Labor Productivity in Late Nineteenth Century Quebec and Ontario Manufacturing: Explaining Ontario’s Advantage.” American Review of Canadian Studies 25:219-249.

-. 1997. “Labor Rights and Labor Power and Welfare Maximization in a Mar¬ ket Economy: Revising the Conventional Wisdom.” Paper presented at the Indus¬ trial Relations Research Association/Allied Social Science Association annual meetings. New Orleans, January. -. 1998. “High Path to Economic Growth and Development.” Challenge: The Magazine of Economic Affairs 41:91—104. -. 1999a. “A Theory of Population Growth When Women Really Count.” 52 (Fasc. 1): 27-43. -. 1999b. “The Methodology of Economics and the Survival Principle Revisited and Revised: Some Welfare and Public Policy Implications of Modeling the Economic Agent.” Review of Social Economics 57: 427-449. -. 1999c. “Social Capacity and Convergence.” In O’Hara, ed. 1999,1033—1035. -. 2000a. “A Behavioral Model of Path Dependency: The Economics of Profit¬ able Ineffieiency and Market Failure.” JoMr«a/ of Socio-Economics 29: 127—145. -. 2000b. “Behavioural Economics.” In Michie, ed. 2000. Altman, M., and L. Lamontagne. 1996. “Aspects of Female Employment and Pay Inequality in Canada, 1900-1930.” Journal of Socio-Economics 25: 285—309. Amsden, A.A. 1994. “Macro-Sweating Policies and Labour Standards.” In Sengenberger and Campbell, eds. 1994a, 185—193. Appelbaum, E., and R. Batt. 1994. The New American Workplace: Transforming Work Systems in the United States. Ithaca, N.Y.: Cornell University Press. Arrow, K. 1962. “The Economic Implications of Learning by Doing.” Review ofEco¬ nomic Studies 29: 155—173. -. 1973. “The Theory of Discrimination.” In Ashenfelter and Rees, eds. 1973,3-33. -. 1980. “Some Models of Racial Discrimination in the Labor Market.” In King, ed. 1980,117-136. Arrow, K., B. Bolin, R. Costanza, P. Dasgupta, C. Folke, C.S. Holling, B.O. Jansson, S. Levin, K.G. Maler, C. Perrings, and D. Pimentel. 1995. “Economic Growth, Carrying Capacity, and the Environment.” Science 268: 520—521. Arthur, B.W. 1989. “Competing Technologies, Increasing Returns, and Lock-In by Historical Events.” Economic Journal 99: 116-131. -. 1990. “Positive Feedbacks in the Economy.” Scientific American 204: 92— 99. Ashenfelter, O., and R. Layard, eds. 1986. Handbook of Labor Economics, vol. 1. Amsterdam: Elsevier Science. Ashenfelter, O., and A. Rees, eds. 1973. Labor Markets. Princeton, N.J.: Princeton University Press.

REFERENCES

273

Aslanbeigui, N., and Y. Back Choi, eds. 1997. Borderline of Economics: Essays in Honor of Daniel R. Fusfeld. London: Routledge. Ayal, E.B. 1963. “Value Systems and Economic Development in Japan and Thai¬ land.” Journal of Social Issues 19: 35—51. -, ed. 1973. Micro-Aspects of Development. New York: Praeger. Balibot, G., R. Frantz, and L. Green. 1987. “Natural Monopolies and Rents: A Georgist Remedy for X-lnefficiency among Publicly-Regulated Virms." American Journal of Economics and Sociology 46: 205—217. Barney, J.B. 1995. “Looking Inside for Competitive Advantage.” Academy of Man¬ agement Executive 9: 49-61. Basu, K. 1999a. “Child Labor: Cause, Consequence, and Cure, with Remarks on International Labor Standards.” Journal of Economic Literature 37:1083—1119. -. 1999b. “International Labor Standards and Child Labor.” Challenge: The Magazine of Economic Affairs 42: 80—93. Bator, F.M. 1957. “The Simple Analytics of Welfare Economics.” American Eco¬ nomic Review 47: 22—59. Baumol, W.J. 1977. Economic Theory and Operations Analysis. Englewood Cliffs, N.J.: Prentice Hall. -. 1982. “Contestable Markets: An Uprising in the Theory of Industry Struc¬ ture.” American Economic Review 72: 1—15. -. 1986. “Productivity Growth, Convergence, and Welfare: What the LongRun Data Show.” American Economic Review 76: 1072—1085. Baumol, W.J., and E.N. Wolff. 1988. “Productivity Growth, Convergence, and Wel¬ fare: Reply.” American Economic Review 78: 1155—1159. Becker, B., and G. Barry. 1996. “The Impact of Human Resource Management on Organizational Performance: Progress and Prospects.” Academy of Management Journal 39-. 119-^0\.

Becker, B., and M.E. Huselid. 1998. “High Performance Work Systems and Firm Performance: A Synthesis of Research and Managerial Implications.” In Ferris, ed. 1998,53-101. Becker, G.S. 1971. The Economics of Discrimination. Chicago: University of Chi¬ cago Press. First published 1957. -. 1985. “Human Capital, Effort, and the Sexual Division of Labor.” Jourwa/ of Labor Economics 3: S33-S58. --. 1998. Accounting for Tastes. Cambridge, Mass.: Harvard University Press. Bell, D., and 1. Kristol, eds. 1981. The Crisis in Economic Theory. New York: Basic Books. Bemmels, B. 1987. “How Unions Affect Productivity in Manufacturing Plants.” In¬ dustrial and Labor Relations Review 41: 241—253. Bergmann, B.R. 1971. “The Effect on White Incomes of Discrimination in Employ¬ ment.” yowrna/ of Political Economy 79: 294—313. -. 1986. The Economic Emergence of Women. New York: Basic Books. -. 1989. “Does the Market for Women’s Labor Need Fixing?” Journal of Eco¬ nomic Perspectives 3: 43—52. Bewley, T.F. 1999. Why Wages Don’t Fall During a Recession. Cambridge, Mass., London: Harvard University Press. Bielby, D.D., and W.T. Bielby. 1988. “She Works Hard for the Money: Household Responsibilities and the Allocation of Work Efiori." American Journal of Sociol¬ ogy 93: 1031-1059.

274

REFERENCES

Bielby, W.T. 1991. “The Structure and Process of Sex Segregation.” In Cornwall and Wunnava, eds. 1991,97—112. Binswanger, H.P. 1978. “Induced Technical Change: Evolution of Thought.” In Binswanger and Ruttan, eds. 1978, 13-43. Binswanger, H.P., and V.W. Ruttan, eds. 1978. Induced Innovation: Technology, Insti¬ tutions, and Development. Baltimore and London: Johns Hopkins University Press. Blaug, M. 1998. “Disturbing Currents in Modem Economics.” Challenge: The Maga¬ zine of Economic Affairs 41:11—34. Blinder, A.S., ed. 1990. Paying for Productivity: Looking at the Evidence. Washing¬ ton, D.C.: Brookings Institution. Bliss C., and N. Stem. 1978. “Productivity, Wages and Nutrition. Part I: The Theory, and Part II: Some Observations.” Joi/rnaZ of Development Economics 5: 331—398. Blois, K. J. 1974. “Some Comments on the Theory of Inert Areas and the Definition of X-Efficiency.” Quarterly Journal of Economics 88: 681—686. Bluestone, B., and B. Harrison. 1988. The Great U-Turn: Corporate Restructuring and the Polarizing of America. New York: Basic Books. -. 1990. “Wage Polarization in the U.S. and the ‘Flexibility’ Debate.” Cam¬ bridge Journal of Economics 14: 351—377. Blum, J. 1957. “The Rise of Serfdom in Eastern Europe.” American Historical Re¬ view 62:807—836. Bomstein, M.H., ed. 1984. Psychology and Its Allied Disciplines. Hillsdale, N.J.: L. Erlbaum Associates. Bosemp, E. 1965. The Condition of Agricultural Growth: The Economics of Agrarian Growth under Population Pressure. Chicago: Aldine. -. 1981. Population and Technical Change: A Study of Long-Term Trends. Chi¬ cago: University of Chicago Press. Bowles, S. 1985. “The Production Process in a Competitive Economy: Walrasian, Neo-Hobbesian, and Marxian MoAeX?,." American Economic Review 75: 16-36. Bowles, S., and H. Gintis. 1998. “Efficient Redistribution: New Rules for Markets, States and Communities.” In Bowles and Gintis, eds. 1998, 3-71. -, eds. 1998. Recasting Egalitarianism: New Rules for Communities, States and Markets. London: Verso. Boyer, R. 1997. “The Variety and Unequal Performance of Really Existing Markets: Farewell to Doctor Pangloss.” In Hollingsworth and Boyer, eds., 1997, 55—93. Boyer, R., and J.R. Hollingsworth 1997. “Coordination of Economic Actors and So¬ cial Systems of Production.” In Hollingsworth and Boyer, eds. 1997, 1—47. Brenner, R. \9%'i. History: The Human Gamble. Chicago: University of Chicago Press. Brown, C. 1988. “Minimum Wage Laws: Are They Overrated?” Jo«r«a/ of Economic Perspectives I: 133—145. Brown, C., and J. Medoff. 1978. “Trade Unions in the Production Process.” Jowrna/ of Political Economy 86: 355—378. Brown, C., C. Gilroy, and A. Kohen. 1982. “The Effect of the Minimum Wage on Employment and Unemployment.” Jowrwa/ of Economic Literature 20: 487-528. Buchanan, J.M. 1994. Ethics and Economic Progress. Norman and London: Univer¬ sity of Oklahoma Press. Buchele, R., and J. Christiansen. 1995. “Worker Rights Promote Productivity Growth.” Challenge: The Magazine of Economic Affairs 38: 32—37. --. 1999. “Labor Relations and Productivity Growth in Advanced Capitalist Economies.” Review of Radical Political Economics 31: 87—110.

REFERENCES

275

Bulow, J.I., and L.H. Summers. 1986. “A Theory of Dual Labor Markets with Appli¬ cation to Industrial Policy, Discrimination, and Keynesian Unemployment.” Jowrnal of Labor Economics A\ 376-414. Button, K.J., ed. 1989. The Collected Essays of Harvey Leibenstein, vol. 2, X-Efficiency and Micro-Micro Theory. New York: New York University Press. Cain, G.G. 1986. “The Economic Analysis of Labor Market Discrimination: A Survey.” In Ashenfelter and Layard, eds. 1986, 693—785. Caldwell, B.J. 1980. “A Critique of Friedman’s Methodological Instrumentalism.” Southern Economic JoumaL Al\ 366-374. Campbell, D., and W. Sengenberger. 1994. “Labour Standards, Economic Efficiency and Development: Lessons from Experience with Industrial Restructuring.” In Sengenberger and Campbell, eds. 1994a, 421^39. Card, D., and A.B. Krueger. 1995. Myth and Measurement: The New Economics of the Minimum Wage. Princeton, N.J.: Princeton University Press. Chandler, A.D., Jr. 1977. The Visible Hand: The Managerial Revolution in American Business. Cambridge, Mass.: Belknap Press of Harvard University Press. -. 1980. “The United States: Seedbed of Managerial Capitalism.” In Chandler and Daems, eds. 1980, 9—40. Chandler, A.D., Jr. and H. Daems, eds. 1980. Managerial Hierarchies: Comparative Perspectives on the Rise of the Modern Industrial Enterprise. Cambridge, Mass.: Harvard University Press. Coase, R.H. 1994a. “How Should Economists Choose?” In Coase 1994b, 15—33. -. 1994b. Essays on Economics and Economists. Chicago and London: Univer¬ sity of Chicago Press. Cohen, D. 1998. The Wealth of the World and the Poverty of Nations. Cambridge, Mass.: MIT Press. Cooler, R., and P. Rappoport. 1984. “Were the Ordinalists Wrong about Welfare Eco¬ nomics?” Journal of Economic Literature 22: 507—530. Cornwall, R.R. and P.V. Wunnava, eds. 1991. New Approaches to Economic and So¬ cial Analyses of Discrimination. New York: Praeger. Cropper, M.L. and W.E. Oates. 1992. “Environmental Economics: A Survey.” Jour¬ nal of Economic Literature 30: 675—740. Cummingham, H. and P. Paolo Viazzo. 1996. “Some Issues in the Historical Study of Child Labour.” In Cummingham and Paolo Viazzo, eds. 1996, 11—22. -, eds. 1996. Child Labor in Historical Perspective, 1800—1985: Case Studies from Europe, Japan and Colombia. Florence: UNICEF. Cyert, R.M. 1988a. “Competition, Growth, and Efficiency.” In Cyert 1988b, 36-56. -. 1988b. The Economic Theory of Organization and the Firm. New York: New York University Press. Cyert, R.M. with C.L. Hedrick. 1988. “Theory of the Firm: Past, Present, and Future: An Interpretation.” In Cyert 1988b, 59-77. Cyert, R.M. with G. Pottinger. 1988. “Towards a Better Microeconomic Theory.” In Cyert 1988b, 201-219. Darity, W., Jr. 1989. “What’s Left of the Economic Theory of Discrimination?” In Shulman and Darity, eds. 1989, 335—374. Dasgupta, P. and D. Ray. 1986. “Inequality as a Determinant of Malnutrition and 'Unemployment.'" Economic Journal 96: 1011—1034. David, P.A. 1985. “Clio and the Economics of Qwerty.” American Economic Review, Papers and Proceedings 75: 332—337.

276 REFERENCES Davis, J.B. 1999. “Justice.” In O’Hara, ed. 1999, 596-599. De Alessi, L. 1983. “Property Rights, Transaction Costs, and X-Efficiency: An Essay in Economic Theory.” American Economic Review 73: 64—81. Dean, J.W. and M. Perlman. 1998. “Harvey Leibenstein as a Pioneer of Our Times.” Economic Journal 108: 132—152. De Jong, H.W. and W.G. Shepherd, eds. 1986. Mainstreams in Industrial Organiza¬ tion, vol. 2. Boston: Kluwer Academic Press. De Long, B.J. 1988. “Productivity Growth, Convergence, and Welfare: Comment.” American Economic Review 78: 1138—1154. Demsetz, H. 1973. “Industry Structure, Market Rivalry, and Public Policy." Journal of Law ahd Economics 16: 1—9. —-. 1982. “Barriers to Entry." American Economic Review 72: 47—57. De Tocqueville, A. 1966. Democracy in America. Translated by G. Lawrence and edited by J.P. Mayer and M. Lemer. New York: Harper and Row. First published 1838. Domar, E.D. 1970. “The Causes of Slavery or Serfdom: A Hypothesis.” Journal of Economic History 30: 18—32. Dore, R. 1973. British Factory—Japanese Factory: The Origins of National Di¬ versity in Industrial Relations. Berkeley and Los Angeles: University of Cali¬ fornia Press. DoAvrick, S. and N. Gemmell. 1991. “Industrialization, Cathing Up and Economic Growth: A Comparative Study across the World’s Capitalist Economies.” Eco¬ nomic Journal 101:263—275. Drago, R. and J.S. Heywood. 1992. “Is Worker Behaviour Consistent with Efficiency Wages?” Scottish Journal of Political Economy 40: 141—153. Eatwell, J., M. Millgate and P. Newman, eds. 1987a. The New Palgrave: A Dictionary of Economics, vol. 1. London and Basingstoke: Macmillan. -. 1987b. The New Palgrave: A Dictionary of Economics, vol. 3. London and Basingstoke: Macmillan. -. 1987c. The New Palgrave: A Dictionary of Economics, vol. 4. London and Basingstoke: Macmillan. Eccles, M. and R.B. Freeman. 1982. “What! Another Minimum Wage SX\xd.y." Ameri¬ can Economic Review, Papers and Proceedings 72: 226-232. Ehrenberg, R. G. and G. T Milkovich. 1987. “Compensation and Firm Performance.” In Kleiner et al., eds. 1987, 87—122. Ehrlich, P.R. and A.H. Ehrlich. 1996. Betrayal of Science and Reason: How AntiEnvironmental Rhetoric Threatens Our Future. Washington, D.C./Covelo, Calif.: Island Press/Shearwater Books. EIkington,J. 1998. Cannibals with Forks: The Triple Bottom Line of 21st Century Busi¬ ness. British Columbia, Canada, and Stony Creek, Conn.: New Society Publishers. Engerman, S. and R.E. Gallman, eds. 1986. Long-Term Factors in American Eco¬ nomic Growth. Chicago and London: University of Chicago Press. National Bu¬ reau of Economic Research Studies in Income and Wealth, vol. 51. Fagerberg, J. 1994. “Technology and International Differences in Growth Rates.” Journal of Economic Literature 32: 1147—1175. Fallon, P. and Z. Tzannatos. 1998. Child Labor: Issues and Directions for the World Bank. Washington, D.C.: World Bank. Faucheux, S. and 1. Nicholai. 1998. “Environmental Technological Change and Goverance in Sustainable Development Policy." Ecological Economics 27: 243-256.

REFERENCES 111 Fehr, E. and S. Gachter. 2000. “Reciprocal Fairness, Cooperation, and Limits to Co¬ operation.” In Hozl, ed. 2000, 1-7. Fehr, E. and M.K. Schmidt. 1999. “A Theory of Fairness, Competition, and Coopera¬ tion.” Quarterly Journal of Economics 114: 817-S68. Feis, H. 1994. “International Labour Legislation in the Light of Economic Theory.” In Sengenberger and Campbell, eds. 1994a, 29-55. Feldman, A.M. 1987. “Welfare Economics.” In Eatwell, Millgate, and Newman, eds. 1987c, 889-895. Ferguson, C.E. 1969. The Neoclassical Theory of Distribution. Cambridge: Cam¬ bridge University Phess. Ferguson, C.E. and J.R Gould. 1975. Microeconomic Theory, 4th ed. Homewood, Ill.: Richard D. Irwin. Ferris, G.R., ed. 1998. Research in Personnel and Human Resources Management. Stanford, Conn.: JAI Press. Fischer, S., ed. 19^6. NBER Macroeconomics Annual, 1986. Cambridge, Mass.: MIT Press. Fogel, R. and S. Engerman. 1971. “The Relative Efficiency of Slavery: A Compari¬ son of Northern and Southern Agriculture in 1860.” Explorations in Economic History 8: 353—361. Frantz, R. 1988. X-Efficiency: Theory, Evidence, Applications. Boston, Dordrecht, and London: Kluwer Academic Press. -. 1997. X-Efficiency: Theory, Evidence and Applications, 2d ed. Boston, Dordrecht, and London: Kluwer Academic Press. Frantz, R., H. Singh, and J. Gerber, eds. 1990. Handbook of Behavioral Economics, vol. 2A. Greenwich, Conn.: JAI Press. Freeman, R.B. 1994. “A Hard-Headed Look at Labour Standards.” In Sengenberger and Campbell, eds. 1994a, 79-92. -. 1995. “Are Your Wages Set in Beijing?” Journal of Economic Perspectives 9:15-32. Freeman, R.B., and J.L. MedofT. 1984. What Do Unions Do? New York: Basic Books. Friedman, M. 1951. “Some Comments on the Significance of Labor Unions for Eco¬ nomic Policy.” In D.M. Wright, ed. 1951,204—234. -. 1953a. “The Methodology of Positive Economics.” In Friedman 1953b, 3-43. -. 1953b. Essays in Positive Economics. Chicago: University of Chicago Press. Fuchs, V.R. 1988. Women’s Quest for Economic Equality. Cambridge, Mass, and London: Harvard University Press. Fukuyama, F. 1995. Trust: The Social Virtues and the Creation of Prosperity. New York: Free Press. Fumham, A. 1990. The Protestant Work Ethic: The Psychology of Work-Related Be¬ liefs and Behaviours. London and New York: Routledge. Fyfe,A. 1993. Child Labour: A Guide to Project Design. Geneva: International Labour Office. Galbraith, J.K. 1998. Created Unequal: The Crisis in American Pay. Chicago: Uni¬ versity of Chicago Press. George, D. 1989. “Social Evolution and the Role of Knowledge.” Review of Social Economy Al: 55—73. Goldin, C. 1986. “The Female Labor Force and American Economic Growth.” In Engerman and Gallman, eds. 1986, 557-604.

278 REFERENCES Goldin, C. 1990. Understanding the Gender Gap: An Economic History of American Women. New York and Oxford: Oxford University Press. Gordon, D.M. 1996. Fat and Mean: The Corporate Squeeze of Working Americans and the Myth of Managerial ‘Downsizing’. New York: Free Press. -. 1998. “Conflict and Cooperation: An Empirical Glimpse of the Imperatives of Efficiency and Redistribution.” In Bowles and Gintis, eds. 1998, 181—207. Gordon, W. 1980. Institutional Economics: The Changing System. Austin and Lon¬ don: University of Texas Press. Graaf, J. de V. 1957. Theoretical Welfare Economics. London: Cambridge University Press. Grootaert, C. and R. Kanbur. 1995. “Child Labour: An Economic Perspective.” Inter¬ national Labour Review \2>A\ 187—203. Groshen, E. L. 1991. “The Structure of the Female/Male Wage Differential: Is It Who You Are, What You Do, or Where You Work?” Journal of Human Resources 26: 457-472. Grossman, G.M. and E. Helpman. 1994. “Endogenous Innovation in the Theory of Growth.” Journal of Economic Perspectives 8: 23-44. Gunderson, M. 1989. “Male-Female Wage Differentials and Policy Responses.” Jowrnal of Economic Literature 27: 46-72. Habbakkuk, H.J. 1962. American and British Technology in the Nineteenth Century: The Search for Labour-Saving Inventions. Cambridge: Cambridge University Press. Haley, B., ed. 1952. Survey of Contemporary Economics, Vol. II. Homewood, IL: Richard D. Irwin. -, ed. 2000. Survey of Contemporary Economics, vol. 2. Homewood, Ill.: Rich¬ ard D. Irwin. Hampden-Tumer, C. and A. Trompenaar. 1993. Seven Cultures of Capitalism. New York: Doubleday. Harrison, L.E. 1992. Who Prospers: How Cultural Values Shape Economic and Po¬ litical Success. New York: Basic Books. Harrison, L.E. and S.P. Huntington, eds. 2000. Culture Matters: How Values Shape Human Progress. New York: Basic Books. Harrod, R.F. 1939. “An Essay in Dynamic Vaeory." Economic Journal A9-. 14—33. Hashimoto, M. 1990. “Employment and Wage Systems in Japan and Their Implica¬ tions for Productivity.” In Blinder, ed. 1990, 245—295. Hayami, Y. and V.W. Ruttan. 1971. Agricultural Development: An International Per¬ spective. Baltimore: Johns Hopkins University Press. Helpman, E. 1992. “Endogenous Macroeconomic Growth Theory.” European Eco¬ nomic Review 36: 237—267. Hicks, J. 1932. The Theory of Wages. London: Macmillan. Himmelfarb, G. 1994a. “Of Heroes, Villains, and Valets.” In Himmelfarb 1994b, 27-49. -. 1994b. On Looking in the Abyss: Untimely Thoughts on Culture and Society. New York: Alfred A. Knopf. Hodgson, G. 1989. “Institutional Rigidities and Economic Growth.” Cambridge Jour¬ nal of Economics 13:79-101. Hofstede, G.H. 1984. Culture's Consequences: International Differences in WorkRelated Values. Beverly Hills, Calif: Sage. Hollingsworth, J.R. and R. Boyer, eds. 1997. Contemporary Capitalism: The Embed¬ dedness of Institutions. Cambridge and New York: Cambridge University Press.

REFERENCES 279 Holzer, H J. 1990. “Wages, Employer Costs, and Employee Performance in the Firm.” Industrial and Labor Relations Review, Special Issue 44: S147-S164. Hozl, E. ed. 2000. Conference Proceedings: Fairness and Cooperation, XXVAnnual Colloquium on Research in Economic Psychology and SABE 2000 Conference.

Baden and Vienna: lAREP. Human Rights Watch. 1996. The Small Hands of Slavery: Bonded Labor In India. New York: Human Rights Watch. Hume, D. \993. An Enquiry Concerning Human Understanding. Indianapolis: Hackett. First published 1777. Ichniowski, C., T.A. Kochan, D. Levine, C. Olson, and G. Strauss. 1996. “What Works at Work: Overview and Assessment.” Industrial Relations 35: 299-333. International Labour Organisation. 1993. Towards the Abolition of Child Labour: ILO Policy and Its Implications for ILO Technical Cooperation Activities. Geneva: In¬ ternational Labour Office. -. 2000a. Statistical Information and Monitoring Programme on Child Labour (SIMPOC). Web address: http://www.ilo.org/public/englishystandards/ipec/simpoc/ simpocOO/index.htm. -. 2000b. Cl 82 Worst Forms of Child Labour Convention, 1999. Web address: http://www.ilo.org/public/english/standards/ipec/ratification/convention/text.htm. Jaffe, A.B., S.R. Peterson, and P.R. Portney. 1995. “Environmental Regulation and the Competitiveness of U.S. Manufacturing: What Does the Evidence Tell Us?” Journal of Economic Literature 33: 132—163. Jorgenson, D.W. and P. J. Wilcoxen. 1990. “Environmental Regulation and U.S. Eco¬ nomic Growth.” RAND Journal of Economics 2\: 314—340. Kahneman, D. and A. Tversky. 1979. “Prospect Theory: An Analysis of Decisions under Bask." Econometrica 47: 363—391. Kaldor, N. 1957. “A Model of Economic Growth.” Economic Journal 67: 591-624. --—. 1970. “The Case for Regional Policies. ” Scottish Journal of Political Economy 18:337-348. Katz, L.F. 1986. “Efficiency Wage Theories: A Partial Evaluation.” In Fischer, ed. 1986,235-276. Kaufman, B.E. 1989. “Models of Man in Industrial Relations Research.” Industrial and Labor Relations Review 43: 72—88. -. 1990. “A New Theory of Satisficing.” Jourwa/ of Behavioral Economics 19: 35-51. Kendrick, J.W., ed. 1984. International Comparisons of Productivity and Causes of the Slowdown. Washington, D.C., and London: American Enterprise Institute/ Ballinger. Kilby, P. 1962. “Organization and Productivity in Backward Economies.” Quarterly Journal of Economics 76: 303—310. -. 1971. “Hunting the Heffalump.” In Kilby, ed. 1971, 1-40. -, ed. 1971. Entrepreneurship and Economic Development. New York: Free Press. King, J. 1995. “High Performance Work Systems and Firm Performance.” Monthly Labor Review 118: 29-36. King, J.E., ed. 1980. Readings in Labor Economics. Oxford: Oxford University Press. Klein, Janice. 1984. “Why Supervisors Resist Employee Involvement.” Harvard Busi¬ ness Review 62: 87—95.

280

REFERENCES

Kleiner, M.M., et al., eds. 1987. Human Resources and the Performance of the Firm. Madison: Industrial Relations Research Association Series. Kochan, T.A., H.C. Katz, and R. B. McKersie. 1986. The Transformation of Ameri¬ can Industrial Relations. New York: Basic Books. Krueger, A.B. and L.H. Summers. 1988. “Efficiency Wages and the Inter-Industry Wage Structure.” Econometrica 56: 259-293. Krugman, R 1991. “Increasing Returns and Economic Geography.” Jowr«a/ of Politi¬ cal Economy 99: 483^99. -. 1994. “Does Third World Growth Hurt First World Prosperity?” Harvard Business Review 72: 113—121. Kuznet, S. 1955. “Economic Growth and Income Inequality.” American Economic Review A5: 1—23. Lamond, A.M. 1977. “Economic Theories of Employment Discrimination.” In Wallace and Lamond, eds. 1977, 1—11. Lamontagne, L. Forthcoming. “Waiting for Elijah: The Structuring of the Firm in the Community in the New York Men’s Garment Industry, 1900-1920.” Ph.D. Disser¬ tation, Department of History, University of Toronto. Lamoreaux, N. 1985. The Great Merger Movement in American Business: 1895— 1904. Cambridge: Cambridge University Press. Lamoreaux, N., D.M.G. Raff, and P. Temin, eds. 1999. Learning by Doing in Mar¬ kets, Firms, and Countries. Chicago: University of Chicago Press. Landes, D.S. 1998. The Wealth and Poverty of Nations: Why Some Are So Rich and Some So Poor. New York and London: W.W. Norton. Lane, F.C. 1958. “The Economic Consequences of Organized Violence.” Journal of Economic History 18: 401^17. -. 1975. “The Role of Governments in Economic Growth in Early Modem Times.” Journal of Economic History 35: 8—17. Lansky, M. 1997. “Child Labor: How the Challenge Is Being Met.” International Labour Review 136: 233—257. Lawrence, R.Z., D. Rodrick, and J. Whalley, eds. 1996. Emerging Agenda for Global Trade: High Stakes for Developing Countries. Washington, D.C.: Overseas De¬ velopment Council. Leibenstein, H. 1957. Economic Backwardness and Economic Growth. New York: J. Wiley and Sons. -. 1966. “Allocative Efficiency vs. ‘X-Efficiency’.” American Economic Re¬ view 56: 392-415. -. 1969. “Organizational or Frictional Equilibria, X-Efficiency, and the Rate of Innovation.” Quarterly Journal of Economics 83: 600-623. -. 1973a. “Competition andEfficiency: ARep\y.'' Journal of Political Economy 81:765-777. -. 1973b. “Notes on X-Efficiency and Technical Change.” In Ayal, ed. 1973, 18-38. -. 1974. “Efficiency Wages, X-Efficiency, and Urban Employment.” In Sellekaerts, ed. 1974, 168-815. -. 1978a. “X-ineflficiency Xists: A Reply to an Xorcist.” American Economic Review 68: 203—211. -. 1978b. General X-Efficiency and Economic Development. New York: Ox¬ ford University Press.

REFERENCES 281 -. 1979. “A Branch of Economics Is Missing: Micro-Micro Theory.” JowrA?a/o/ Economic Literature 17: 477—502. -. 1980. Beyond Economic Man: A New Foundation for Microeconomics. Cam¬ bridge, Mass.: Harvard University Press. -. 1981. “Microeconomics and X-Efficiency Theory: If There Is No Crisis, There Ought to Be.” In Bell and Kristol, eds. 1981, 97—122. -. 1982. “The Prisoner’s Dilemma in the Invisible Hand: An Analysis of Intrafirm Productivity.” American Economic Review 12: 92—97. -. 1983a. “Intrafirm Productivity: Reply.” American Economic Review 13: 822— 823. -. 1983b. “Property Rights and Efficiency: Comment.” American Economic Review 73: 831—842. -. 1984. “On the Economics of Conventions and Institutions: An Exploratory Essay.” Journal of Institutional and Theoretical Economics 140: 74—89. -. 1985. “On Relaxing the Maximization Hypothesis.” Jourua/ of Behavioral Economics 14: 5—20. -. 1986. “Leibenstein on the Maximization Postulate: Comments and Responses to Contributors.” Journa/ of Behavioral Economics 15: 57—63. -. 1987. Inside the Firm: The Inefficiencies of Hierarchy. Cambridge, Mass., and London: Harvard University Press. Leibenstein, H. and Maital, S. 1994. “The Organizational Foundations of X-Inefficiency: A Game-Theoretic Interpretation of Argyris’ Model of Organizational Learn¬ ing.” Jbwr/ia/ of Economic Behavior and Organization 23: 251—268. Leontief, W. 1971. “Theoretical Assumptions and Nonobserved Facts.” American Economic Review 6\: 1—7. Lester, R. 1946. “Shortcomings of Marginal Analysis for Wage-Employment Prob¬ lems.” American Economic Review 36: 62—82. -. 1947. “Marginalism, Minimum Wages, and Labor Nlarkets.” American Eco¬ nomic Review 36: 135—148. Levine, D.I. 1995. Reinventing the Workplace: How Business and Employees Can Both Win. Washington, D.C.: Brookings Institution. Levine, D.I. and L. D’Andrea Tyson. 1990. “Participation, Productivity, and the Firm’s Environment.” In Blinder, ed. 1990, 183—243. Lewis, W.A. 1954. “Economic Development with Unlimited Supplies of Labour.” Manchester School of Economic and Social Studies 26: 139-191. Liebowitz, S.J. and S.E. Margolis. 1990. “The Fable of the Keys.” Journal of Law and Economics 33: 1—25. -. 1994. “Network Externalities: An Uncommon Tragedy.” Journal of Eco¬ nomic Perspectives 8: 133—150. Lockwood, B. 1987. “Pareto Efficiency.” In Eatwell, Millgate, and Newman, eds. 1987b,811-813. Logue, J. and J.S. Yates. 1999. “Worker Ownership American Style: Pluralism, Par¬ ticipation and Performance.” Economic and Industrial Democracy 20: 225—252. Lucas, R.E., Jr. 1988. “On the Mechanics of Economic Development.” Jowrna/ of Monetary Economics 22: 3^2. Maddison, A. 1995. Monitoring the World Economy, 1820-1992. Paris: OECD. Maital, S. 1982. Minds, Markets and Money: Psychological Foundations ofEconomic Behavior. New York: Basic Books.

282

REFERENCES

Maital, S. and S. Sharabani. 1997. “Culture as a Growth Resource.” In Aslanbeigui and Back Choi, eds. 1997, 171-193. Maital, S. and S.L. Maital. 1984. “Psychology and Economics.” In Bomstein, ed. 1984,55-87. -, eds. 1993. Economics and Psychology. Aldershot, England: Edward Elgar. Mankiw, G.M., D. Romer, and D.N. Weil. 1992. “A Contribution to the Theory of Economic Growth.” Quarterly Journal of Economics 107: 407—437. Marsh, E. 1993. “Is Environmental Technology a Key to a Healthy Economy?” Sci¬ ence 260\ 1886-1888. Marshall, M.G. 1998. “Scottish Economic Thought and the High Wage Economy: Hume, Smith and McCulloch on Wages and Work Motivation.” Scottish Journal of Political Economy 45: 309-328. Marshall, R. 1974. “The Economics of Racial Discrimination: A Survey.” Jowrna/ of Economic Literature 12: 849—871. -. 1987. Unheard Voices: Labor and Economic Policy in a Competitive World. New York: Basic Books. -. 1994. “The Importance of International Labour Standards in a More Competitive Global Economy.” In Sengenberger and Campbell, eds. 1994a, 65— 78. McCloskey, D.N. 1990. If You’re So Smart: The Narrative of Economic Expertise. Chicago: University of Chicago Press. -. 1996. The Vices of Economists—The Virtues of the Bourgeoisie. Amsterdam: Amsterdam University Press. Mendelievich, E., ed. 1979. Children at Work. Geneva: International Labour Office. Michie, J., ed. 2000. Reader’s Guide to the Social Sciences. London: Fitzroy Dearborn Publishers. Miller, G.J. 1992. Managerial Dilemmas: The Political Economy of Hierarchy. New York: Cambridge University Press. Mitchell, D.J.B. 1990. “Alternative Pay Systems, Firm Performance, and Productiv¬ ity.” In Blinder, ed. 1990, 15-94. Mokyr, J. 1990. The Lever of Riches: Technological Creativity and Economic Progress. New York: Oxford University Press. Mueller, D.C. 1986. “United States’ Antitrust: At the Crossroads.” In De Jong and W.G. Shepherd, eds. 1986, 215—241. Muller, J.Z. 1993. Adam Smith in His Times and Ours. Princeton, N.J.: Princeton University Press. Musgrave, A. 1981. “Unreal Assumptions in Economic Theory: The F-Twist Un¬ twisted.” Kyklos 34 (Fasc. 3): 377—387. Myrdal, G. 1963. Economic Theory and Underdeveloped Regions. London: Methuen. Neal, J.A. and C.L. Tromley. 1995. “From Incremental Change to Retrofit: Creating High-Performance Work Systems.” Academy of Management Executive 9:42-53. Nelson, R.N. and G. Wright. 1992. “The Rise and Fall of American Technological Leadership: The Postwar Era in Historical Perspective.” Journal of Economic Lit¬ erature hW. 1931—1964. Nelson, R. R. 1968. “A ‘Diffusion’ Model of International Productivity Differences in Manufacturing Industry.”/Iwencaw Economic Review 58: 1219-1248. -. 1984. “Where Are We in the Discussion? Retrospect and Prospect.” In Kendrick, ed. 1984, 397-409.

REFERENCES 283 -. 1995. “Recent Evolutionary Theorizing about Economic Change.” Jonrna/ of Economic Literature 33: 48—90. -. 1998. “The Agenda for Growth Theory: A Different Point of View (Short¬ comings of Neoclassical Economics in Explaining).” Cambridge Journal of Eco¬ nomics 22:497—520. North, D.C. 1990. Institutions, Institutional Change, and Economic Performance. Cambridge: Cambridge University Press. OECD. 1983. Environmental Policies and Industrial Competitiveness. Paris: OECD. -. 1985. Environmental Policy and Technical Change. Paris: OECD. -. 1996. Trade, Employment and Labour Standards: A Study of Core Workers' Rights and International Trade. Paris: OECD. O’Hara, R, ed. 1999. Encyclopaedia of Political Economy. London: Routledge. Okun, A.M. 1975. Equality and Efficiency: The Big Trade-Off. Washington, D.C.: Brookings Institution. Olson, M. 1996. “Big Bills Left on the Sidewalk: Why Some Nations Are Rich, and Others Voov." Journal of Economic Perspectives 10: 3—24. Olson, M. and S. Kahkonen, eds. 2000. A Not-So-Dismal Science. Oxford: Oxford University Press. Osberg, L. 1995. “The Equity/Efficiency Trade-Off in Retrospect.” Canadian Busi¬ ness Economics 3: 5—19. Pack, H. 1994. “Endogenous Growth Theory: Intellectual Appeal and Empirical Short¬ comings.” Journal of Economic Perspectives 8: 55—72. Palmer, K., W.E. Oates, and P.R. Portney. 1995. “Tightening Environmental Stand¬ ards: The Benefit—Cost or the No-Cost Paradigm?” Jour«a/ of Economic Perspec¬ tives 9: 119—132. Parente, L. and E.C. Prescott. 2000. Barriers to Riches. Cambridge, Mass, and Lon¬ don: MIT Press. Pencavel, J. and C.E. Hartsog. 1984. “A Reconsideration of the Effects of Unionism on Relative Wages and Employment in the United States, 1920-1980.” Jouma/ of Labor Economics 2: 193—232. Perelman, M. 1993. The Pathology of the U.S. Economy: The Costs of a Low Wage Economy. New York: St. Martin’s Press. Pfeffer, J. 1995. “Producing Sustainable Competitive Advantage through the Effec¬ tive Management of Veofie." Academy of Management Executive 9: 55—69. Pfeffer, J. and R.I. Sutton. 2000. The Knowing-Doing Gap: How Smart Companies Turn Knowledge into Action. Boston, Mass.: Harvard Business School Press. Pigou, A.C. 1951. “Some Aspects of Welfare Economics.” American Economic Review 41:287-302. -. 1952. The Economics of Welfare. London: Macmillan. Porter, M.E. 1991. “America’s Green SVcaXogy.” Scientific American 264: 168. Porter, M.E., and C. Van der Linde. 1995a. “Toward a New Conception of the Environment-HZompetitiveness Relationship.” Journal of Economic Perspectives 9: 97-118. -. 1995b. “Green and Competitive: Breaking the Stalemate.” Harvard Busi¬ ness Review 13: 120—134. Poundstone, W. 1992. Prisoner’s Dilemma: John Von Neumann, Game Theory, and the Puzzle of the Bomb. New York: Anchor Books. Pritchett, L. 1997. “Divergence, Big Time.” Journal of Economic Perspectives 11:3-17.

284

REFERENCES

Reder, M.W. 1982 “Chicago Economics: Permanence and OnangQ." Journal of Eco¬ nomic Literature 20: 1—38. Reich, M. 1981. Racial Inequality: A Political-Economic Analysis. Princeton, N.J.: Princeton University Press. Reich, R.B. 1992. The Work of Nations: Preparing Ourselves for 21st Century Capi¬ talism. New York: Vintage Books. Rodrick, D. 1996. “Labor Standards in International Trade: Do They Matter and What Do We Do about Them?” In Lawrence, Rodrick, and Whalley, eds. 1996, 35—79. Romer, P.M. 1986. “Increasing Returns and Long-Run Growth. "Journal of Political Economy 94: 1002—1037. -. 1990. “Endogenous Technical Change.” Journal of Political Economy 98: S71-S102. -. 1994. “The Origins of Endogenous Growth.” Jowr«a/ of Economic Perspec¬ tives 8: 3—22. Rosenberg, A. 1992. Economics: Mathematical Politics or Science of Diminishing Returns? Chicago and London: University of Chicago Press. Rosenberg, N. 1982a. “The Historiography of Technical Progress.” In Rosenberg 1982b, 3-33. -. 1982b. Inside the Black Box: Technology and Economics. Cambridge: Cam¬ bridge University Press. Rozen, M.E. 1985. “Maximizing Behavior: Reconciling Neoclassical and X-Efficiency Approaches.” Jowrna/ of Economic Issues 19: 661-685. -. 1990. “X-Efficiency, Implicit Contracting, and the Theory of the Firm.” In Weirmair and Perlman, eds. 1990, 95—125. -. 1991. The Economics of Organizational Choice: Workers, Jobs, Labor Mar¬ kets, and Implicit Contracting. Ann Arbor: University of Michigan Press. Rutherford, M. 1996. Institutions in Economics: The Old and the New Institutional¬ ism. Cambridge and New York: Cambridge University Press. Ruttan, V.W. 1997. “Induced Innovation, Evolutionary Theory and Path Dependence: Sources of Technical CbmgQ." Economic Journal 107: 1520—1529. Salter, W.E.G. 1969. Productivity and Technical Change, 2d ed. London: Cambridge University Press. Schumpeter, J.A. 1974. The Theory of Economic Development: An Inquiry into Prof¬ its, Capital, Credit, Interest, and the Business Cycle. London: Oxford University Press. First published 1934. Schwartz, H.H. 1998. Rationality Gone Awry: Decision Making Inconsistent with Economic and Financial Theory. Westport, Conn.: Praeger. Scitovsky, T. 1943-44. “A Note on Profit Maximization and Its Implications.” Re¬ view of Economic Studies 11: 57-60. Sellekaerts, W., ed. \914. Economic Development and Planning. London: Macmillan. Sengenberger, W. 1994a. “International Labour Standards in a Globalized Economy: The Issues.” In Sengenberger and Campbell, eds. 1994a, 3—15. -. 1994b. “Labour Standards: An Institutional Framework for Restructuring and Development.” In Sengenberger and Campbell, eds. 1994a, 3—41. Sengenberger, W. and D. Campbell, eds. 1994a. International Labour Standards and Economic Interdependence. Geneva: International Labour Organization. -. 1994b. Creating Economic Opportunities: The Role of Labour Standards in Industrial Restructuring. Geneva: International Labour Organization.

REFERENCES 285 Shapiro, C. and J.E. Stiglitz. 1984. “Equilibrium Unemployment as a Worker Disci¬ pline T)qV\cq." American Economic Review lA: 433-444. Shen, T.Y. 1984. “The Estimation of X-Efficiency in Eighteen Countries.” Review of Economics and Statistics 66: 98—104. -. 1991. “Towards a General Theory of X-Efficiency.” Jowr«a/ of Socio-Eco¬ nomics. 20: 277—295. Shepherd, W.G. 1984. “Contestability vs. Competition.” 4 Economic Review lA: 572-587. -. 1986. “On the Core Concepts of Industrial Economics.” In De Jong and Shepherd, eds. 1986,23—67. Shiller, R.J. 1991. Market Volatility. Cambridge and Cambridge, Mass.: MIT Press. Shimada, H. 1985. “The Perceptions and the Realities of Japanese Industrial Rela¬ tions.” In Thurow, ed. 1985,44—66. Shulman, S. and Darity, W., Jr., eds. 1989. The Question of Discrimination: Racial Inequality in the US. Labor Market. Middletown, Conn.: Wesleyan University Press. Siddiqi, F. and H.A. Patrinos. 2000. “Child Labor: Issues, Causes and Interventions.” HCO Working Papers, no. 56. Web address: http://www.worldbank.org/html/extdr/ hnp/hddflash/workp/wp_00056.html. Simon, H.A. 1959. “Theories of Decision Making in Economics and Behavioral Sci¬ ence.” American Economic Review 49: 252—283. -. 1978. “Rationality as a Process and as a Product of Thought.” Economic Review 19'. 1—16. -. 1987. “Behavioral Economics.” In Eatwell, Millgate, and Newman, eds. 1987a, 221-225. Simon, J.L. 1991. “A Theory of Effort as an Economic Variable.” of SocioEconomics lO: 105—123. Smith, A. \92>1. An Inquiry into the Nature and Causes of the Wealth of Nations.New York: Modem Library. Smith, H. 1995. Rethinking America. New York: Random House. Solow, R.M. 1956. “A Contribution to the Theory of Economic Growth.” Quarterly Journal of Economics 70: 65—94. --. 1957. “Technical Change and the Aggregate Production Function.” Review of Economics and Statistics 39: 312—320. -. 1986. “Another Possible Source of Wage Stickiness.” In Akerlof and Yellen, eds. 1986,41^4. -. 1994. “Perspectives on Growth Theory.Journal of Economic Perspectives 8: 45-54. Sowell, T. 1994. Race and Culture: A World View. New York: Basic Books. Statman, M. 2000. “Socially Responsible Mutual Funds.” Financial Analysts Journal 56: 30-39. Stewart, R.B. 1993. “Environmental Regulation and International Competitiveness.” Yale Law Journal 102: 2039—2108. Stigler, G. 1946. “The Economics of Minimum Wage Legislation.” Eco¬ nomic Review 56: 358—365. -. 1976. “The X-istence of X-Efficiency.” 4/«enca« Economic Review 66:2\5216. Stiglitz, J.E. 1976. “The Efficiency Wage Hypothesis, Surplus Labour and the Dis¬ tribution of Income in LDC’s.” Oxford Economic Papers 28: 185—207.

286

REFERENCES

-. 1987. “The Causes and Consequences of the Dependence of Quantity on Price.” Journal of Economic Literature 25: 1^8. -. 2000. “The Insider: What I Learned at the World Economic Crisis.” New Republic {NpxW 17). Summers, L. 2000. “Interview With Larry Summers.” Red Herring Magazine, Sep¬ tember. Sutcliffe, R.B. \91\. Industry and Underdevelopment. London: Addison-Wesley. Szostak, R. 1999. Econ-art: Divorcing Art from Science in Modern Economics. Lon¬ don and Sterling, Va.: Pluto Press. Takeuchi, H. 1985. “Motivation and Production.” In Thurow, ed. 1985, 18—28. Thaler, R.H. 1992a. “Interindustry Wage Differentials,” in Thaler 1992b, 36-49. -. 1992b. The Winner’s Curse: Paradoxes and Anomalies of Economic Life. New York: Free Press. Thorbecke, E. 1990. “Institutions, X-Efficiency, Transaction Costs and Socioeconomic Development.” In Weiermair and Perlman, eds. 1990, 295—313. Thurow, L. 1975. Generating Inequality: Mechanisms of Distribution in the U.S. Economy. New York: Basic Books. -, ed. 1985. The Management Challenge: Japanese Views. Cambridge, Mass.: MIT Press. Tobey, J.A. 1990. “The Effects of Domestic Environmental Policies on Patters of Trade: An Empirical Test.” Kyklos 43: 191—209. Todaro, M.P. 1989. Economic Development in the Third World, 4th ed. New York and London: Longman. Tomer, J.F. 1987. Organizational Capital: The Path to Higher Productivity and WellBeing. New York: Praeger Publishers. -. 1997. The Human Firm: A Socio-Economic Analysis of Its Behavior and Potential in a New Economic Age. London and New York: Routledge. -. 1999. “Economics and Human Resource Management: Understanding the True Potential of the Employer—Employee Relationship.” Paper presented at the biannual meeting of the Society for the Advancement of Behavioral Economics, San Diego, July. Tversky, A. and D. Kahneman. 1981. “The Framing of Decisions and the Psychology of Choice.” SciencelX 1: 453—458. Van der Linde, C. 1993. “Micro-Economic Aspects of Environmental Policies and Competitiveness: The Micro-Economic Implications of Environmental Policies and Competitiveness.” In OECD 1983, 69-77. Von Weizsacker, E., A.B. Lovins, and L.H. Lovins. 1997. Factor Four, Doubling Wealth, Halving Resource Use: The New Report to the Club of Rome. Lon¬ don: Earthscan. Wallace, P.A. and A.M. Lamond, eds. 1977. Women, Minorities, and Employment Discrimination. Lexington, Mass.: Lexington Books. Weber, M. 1958. The Protestant Ethic and the Spirit of Capitalism: The Relationship between Religion and the Economic and Social Life in Modern Culture. New York: Charles Scribner’s Sons. Weiermair, K., and M. Perlman, eds. 1990. Studies in Economic Rationality: X-Effi¬ ciency Examined and Extolled. Ann Arbor: University of Michigan Press. Weitzman, M.L. and D.L. Kruse. 1990. “Profit Sharing and Productivity.” In Blinder, ed. 1990,95-141.

REFERENCES 287 Wellington, AJ. 1991. “Effects of the Minimum Wage on the Employment Status of Youths; An Update.” Journal of Human Resources 26: 27-46. Wessels, W. 1985. “The Effeets of Unions on Employment and Productivity: An Un¬ resolved Contradiction.” of Labor Economics 3: 101—108. Wilkinson, F. 1994. “Equality, Efficiency and Economic Progress: the Case for Uni¬ versally Applied Equitable Standards for Wages and Conditions of Work.” In Sengenbeger and Campbell, eds. 1994b, 61—86. Williamson, O.E. 1975. Markets and Hierarchies: Analysis and Antitrust Implica¬ tions. New York: Free Press. -. 1980. “Emergence of the Visible Hand: Implications for Industrial Organiza¬ tion.” In Chandler and Daems, eds. 1980, 182—202. -. 1981. “The Modem Corporation: Origins, Evolutions, Attributes.” Jowr«a/ of Economic Literature 19: 1537—1568. -. 1985. The Economic Institutions of Capitalism: Firms, Markets, Relational Contracting. New York: Free Press. -. 1986. Economic Organization: Firms, Markets and Policy Control. New York; New York University Press. Winther, G. and R. Marens. 1997. “Participatory Democracy May Go a Long Way: Comparative Growth Performance of Employee Ownership Firms in New York and Washington States.” Economic and Industrial Democracy 18: 393-422. Witt, U. 1997. “‘Lock-in’ vs. ‘Critical Masses’; Industrial Change under Network Externalities.” InternationalJournal of Industrial Organization 15: 753—773. World Bank. 1999. World Development Report 1999/2000. New York: Oxford Uni¬ versity Press. -. 2000. World Development Indicators 2000. Washington, D.C.: World Bank. Wright, D.M., ed. 1951. The Impact of the Union: Eight Economic Theorists Evalu¬ ate the Labor Union Movement. New York: Harcourt, Brace and Company. Wright, G. 1975. “Slavery and the Cotton Boom.” Explorations in Economic History 12:439-451. Yang, B. and D. Lester. 2000. “An Exploration of the Impact of Culture on the Economy: an Empirical Study of Unemployment.” Journal of Socio-Economics 29:281-290. Yankelovich, D. and J. Immerwahr. 1983. Putting the Work Ethic to Work: A Public Agenda Report on Restoring America’s Competitive Vitality. New York: Public Agenda Foundation. Yellen, J.L. 1984. “Efficiency Wage Models of Unemployment.” 4 wencan Economic Review 74: 200—205. You, Jong-Il. 1994. “Macroeconomic Stmcture, Endogenous Technical Change and Growth.” Cambridge Journal of Economics 18: 213—233. Young, A.A. 1928. “Increasing Returns and Economic Progress.” Economic Journal 38:527-542.

.

■ «tT6;2 i«^/!ij|firftivV< ?m v5,AV^;ii»^,M!ff.^ >

-■ .

-flV‘■rjA ^/C* J^rri ,*,.T >■••. • ./'B l'A7.a>ir-Uu \ ■ '

’*■ * - .

to ■' '

.-'. t»/ifl ^\uU loS »^Jii(‘>jrtJ2

'

;'*5 ■’

,

t^rtjjffA

d^'-

’

'

-

•,?r3i5--^5t

wr!-^-of wtjJr

.

' i-'

.

■/

arfnSilf * -Or^r‘. .4ai-«f\ :jf,^tAkur*V-tW .**.#rWr :. . ,^ -01 - J ^ C.;.:.^ ^..f>

,

.'' tv

•yd" -irvjf) >'?y.V}>ftcaa

_' .' " 'r'~ ' ''■

:

^

-*. Vv-,-.

/{ ’ U »>'. r^; ^ rV;, * ’. kjO^IafO-l- UltJ *' . ••.«l»2J;r.h^ .J.*.'.|.n’‘ iJ:,.

■.. .Jl»

^r

• ■’1 \t/^.\rdRVnfiVfin5(r^«/tyvoiO

L

t

»

. •■•».^*it^«l*^ftllf!P*«^^i'■t

< .■Nn»>ti-*i4' < i* • i ■ ■ . *• il’ •»‘

■^ ■'• - '

■ ■

' -

4*^-^ #E . . ... n. ; J!

.■/-4»^-,- ’"rrTn^ifwy ^^•'/iVCJj V'J i. •'iv.^.-ti»u- * ;c»i.'^

A. Vf-'J •y'\»^i

.'I

j ^j

Index Abramovitz, Moses, 133, 196 Africa, x, 199 Age restrictions, 182, 183, 200-201 Aghion, R, 97 Akerlof, G.A., 14, 116/J.2, 221 Allen, Woody, xv Allocative efficiency, 69-71, 74, 75, 94 and economic welfare, 72, 85 and integration, 251—52 and Pareto Efficiency, 72 and Pigouvian Optimality, 83 Altruism, 195, 214 American Question see Axiom of Modest Greed Antitrust activity, 248, 258, 258/j. 1 Apprenticeships, 201 “Are Your Wages Set in Beijing?” (Freeman), 192 Arrow, Kenneth, xii, 119, 217, 226/1.3 Arthur, Brian, xiii, 19, 150, 152, 154 “As if’ hypothesis, 54 Asia, 199 Asset specificity, 252—53 Association, freedom of, 197/1.2 Average costs, 59-60 Average factor productivity, 32 Axiom of Modest Greed, 5, 85, 86-88, 91, 171,231,244 Basu, K., 214 Baumol, W.J., 5 Becker, Gary, xiv, 21, 216, 217, 222, 227-28/1/1.4, 7, 8, 228-29/1.12, 261, 264, 269/1.1 Behavioral economics, 26, 66/i. 1, 94,

101 2 1.2 -

/

Bemmels, B., 117/1.4 Bergmann, Barbara, 226/1.3 Berlin Wall, ix Blackboard economics, 52 Blaug, M., 66/1.2 Bluestone, B., 135 Boserup, Ester, 44

Bounded rationality, 66/1.1, 252-53 Brenner, Reuven, 44, 49/1.12 Buchanan, J.M., 270/1.8 Buchele, R., 172 Canada, 173 Capital investments, 48/1.6, 126 Capital mobility, 133 Capital-labor ratio, 127, 137, 146-47, 149/1.7 Capitalism, ix-x, 262-63 Cardinalists see Material Welfare School Caroli, E., 97 Chandler, A.D., 257 Chicago School, x, 4, 55, 56, 94 Child care, 227/1.4 Child labor, 21, 197/1.2, 199-214 downward trend in, 202 eliminating, 212—14, 215/1.6 extent of, 199, 200/ and family income, 201,210-12 government intervention in, 213 household decision makers and, 213—14 impact of reducing, 201, 203, 205, 206/ and per capita income, 202 per capita output and, 210-12 and unit costs, 201, 203-10 working conditions of, 208—9 Children, 82 Choice, xii—xiii, 4, 15, 23, 24, 26, 47 see also Consumer choice; Effort discretion Christiansen, J., 172 Cleanup, 232 Coefficient of discrimination, 217—19, 225 Collective bargaining, 197/1.2 Community, sense of, 262 Commutative justice, 65-66 Competition, 55—56, 57, 249-50 Consumer choice, 57, 186 Contestable markets, theory of, 5-6 289

290

INDEX

Contracts, 30, 36-31, 59, 72, 107, 127, 139, 187, 236, 253, 256 Convention Number 5, 200 Convention Number 138, 200-201 Convergence, 19, 25n.2, 120-21, 126, 128, 264 in labor productivity, 137-48 Cooler, R., 79 Corporations, 22—23, 60 traditional critique of, 250 see also Transaction cost-economizing paradigm Correlation analysis, 54 Cost minimization, 56, 264 Cost-minimizing behavior, 48/7.3 Culture, xiv-xv, 260-69 defined, 261—64 and economic development, 23, 261-64 model of, 264—68 and wage rates, 267/ see also Work culture Cumulative causation, 119-20 Darwin, Charles, x David, Paul, xiii, 19, 150, 152, 154 De Alessi, Louis, 48/7.3 De Tocqueville, Alexis, 270/7.3 Densetz, H., 258/7.1 Diminishing utility, 80 Discrimination, xiv, 21—22, 197/7.2 see also Pay inequality Distributive justice, 65^6, 79 Divergence, 126 Domain assumption, 58 Domar, E.D., 169/7.14 Dresher, Melvin, 180-81/7.9 Dynamic efficiencies, 71, 85, 91 Eastern Europe, 202 Economic analysis, 50 Economic efficiency, 5-6, 48/7.3 and corporate bigness, 22—23, 60, 249 and firm organization, 20 and labor costs, 18 persistence of, 20 and worker satisfaction, 20-21 Economic imperative, 5-6 Economic justice, 65-66 behavioral theory of, 69-101 and Pigouvian Optimality, 83-84 Economic models, 50-51, 57-58

Economic policy, 50 Economic theory behavioral assumptions and, 53—54, 58, 64 billiard player example in, 54 and culture, 260 and data sets, 52 efficacy of, 51 empirical instrumentalism, 51—52, 64-65 survival principle see Survival principle validity of, 66^7/7.3 Economic welfare, 3—24 Economic welfare see Welfare Economics, methodology of, 50-66 Economics of Welfare, The (Pigou), 79 Education, 81, 82, 85, 94, 200, 201, 263 Efficiency wage, 13—14, 140-41, 159, 160/ 168/7.8, 198/7.5, 228/7.10 and child labor, 207 and conventional wisdom, 58-64 and technical change, 161-62 Efficiency wage theory, 9, 13, 25/7.5, 48/7.2, 61, 67/7.11, 116/7.2, 118/7.9, 149/7.6, 204 Effort bundles, 30, 34 Effort discretion, 17, 24, 29, 30, 57, 62, 63, 72, 85, 107, 118/7.10, 127, 138-39, 174, 179, 256 and child labor, 205—6 and environmental regulation, 236-37, 244-45 and labor rights, 186-92 and path dependency, 155-56, 163-64 and pay inequality, 220-21 and productivity, 85—86 and x-efficiency theory, 58-60 Effort inputs, xii-xiii, 16, 62, 118/7.10, 138,225 diminishing returns to, 14 Effort intensity, 221-25 Effort maximization, 237 Empirical instrumentalism, 51—52 Employment, 103-4 long-term, 170 Employment growth, 62, 103-4, 115 and wage rates, 113 Entrepreneurs, 27, 40, 45, 46, 48/7.5, 110, 154 and technology, 43-44 Entry barriers, 250

INDEX

Environmental policy, 22, 230-45 and capital and labor, 241, 242/ conventional wisdom on, 244-^5 government intervention in, 234—35 opportunity costs of, 230, 233-35 productivity increases from, 232—33 and technical change, 240-44 and x-inefficiency, 236-39 Equilibrium wage, 78 Equity, 18 Ethics, 262 Factor inputs, 87 Fair wage hypothesis, 221—22 Family farms, 44 Family income, 210-13, 215«.5 Firm hierarchy see Management Firms behavioral model of, 26 and child labor, 2Ql-% motivational structures and culture of, 27 objective function of, 248 profit maximizing and x-inefficient, 33-34 First-mover advantage, 120, 150, 152—53, 168n.3 Flood, Merrill, 180-81«.9 Forbes Magazine, xi Free labor, 166, 169n. 14 Free will, 46—47 Freeman, Richard, 184—85, 192, 193, 196 Friedman, Milton, 4, 16-17, 50, 64—65, 66-67«.3, 72 methodology of, 51—57 Fuchs, V.R., 227«.4 Fukuyama, Francis, 23 Galbraith, J.K., 16 Garcia-Penalosa, C., 97 General Motors (GM), 173-74 George, D., 168/2.2 Germany, 173 Globalization, x—xi, 5-7, 136/2.7 GM see General Motors Gordon, D.M., 172, 193 Government participation, 6, 185, 190, 192, 195 Governments, 143, 256 Graaf, J. de V., 70

291

Great Britain, 126 effect of wages in, 129 Grossman, Gene, 119, 126 Growth behavioral model of, 119-35 conventional model for, 121—26 high-wage path to, 119, 121 and income inequality, 95—101, 98-100/ low-wage path to, 121 objective of, 133—35 patterns of, 127—33 sources of, 135/2.3 sustainable paths to, 121 Growth theory, 18-19, 126 Habbakkuk, H.J., 129 Harrison, B., 135, 262 Harrod, R.F., 122 Hashimoto, M., 118/2.11 Health and safety regulations, 182, 183 Health services, 94 Helpman, Elhanen, 119, 126 Heuristic assumption, 58 Hicks, John, 240 High-wage regimes, xii—xiii, 14, 15, 33, 196-97 economic costs of, 195 and growth, 127—133 and technology, 42—43 Himmelfarb, G., 270/2.3 Household work, 228—29/2.12 Housing, 94 Human agency, 23, 26-47, 186 alternative model of, 27-40 objective functions of, 27—28, 29 and technical change, 40-45 and utility, leisure, and welfare, 45-46 Human capital, 82, 85, 126, 137, 200, 229/2.12, 262 Human Development Index, xi—xii Human poverty, xi—xii Human resource management, 170 Hume, D., 56 Hunger, x-xi Immerwahr, J., 118/2.10 Imperfect product markets, 28 inceincetives, 253 Income see Wages Income, comparative vs. absolute levels of, 80-81

292

INDEX

Income inequality, and GNP, 95-101,98—

100/ Income redistribution, xiii, 17—18, 65-66, 69-71 and economic welfare, 71, 74, 80-81, 91-95 equity-efficiency trade off, 95—101 and investment incentives, 84 and output drop, 84 and productivity, 85 and real output, 84, 92-94, 93/ 94 in Smithian theory, 95 and social justice, 76-83, 84—93 through taxation, 82 and x-efficiency, 74 Indirect incentives, 187 Industrial Relations (IR), 4, 10, 20, 26-27, 61, 117«.4, 156, 168/J.6 convergence of, 171 cooperative systems of, 25/1.4, 59, 77, 105, 170, 187, 195, 265 and path dependency, 158-59 see also Work culture Inefficiency see Economic efficiency; X-inefficiency Information, 185, 253 Information costs, 24n. 1 Innovation, 40, 44-45, 126, 233 Input productivity, 149/J.3 Institutionalist instrumentalism, 67/1.4 Institutions, 46-47 Integration, 247—58, 258/z.l by-products of, 251 competitive pressures and, 250, 252 costs of, 253, 254-56, 255/ institutional framework for, 248-49 and market pressure, 254 societal benefits of, 254 and x-inefficient institutions, 256-57 Internal governance structures, 253 International Labour Organisation, 199-201 Investment, 193 Israel, xiii Italy, 173 Jaffe, A.B., 235 Japan, 173 Job security, 187

Kaldor, Nicholas, 120, 135/1.2 Kibbutz system, xiii Knowing-doing gap, xiv Knowledge, 119, 126 Krueger, A.B., 149/1.6 Krugman, Paul, 193 Kuhn, Thomas, 7—8 Kuznets, S., 96-97 Labor, 35 capacity of, 82 costs, 11, 16, 38, 62, 102/1.5, 109, 265 convergence in, 19 and economic efficiency, 18 and effort inputs, 14 demand, 78, 108—14, 118/1.12 demand curve, 63 and environmental regulation, 241, 242/ free vs. unfree, 169n. 14 management relations, 16 markets, 10, 134—35 discrimination in, 216, 217—20, 225-26 imperfections in, 192 segmentation of, 226/1.3, 228/1.9 structures of, 166 mobility, 129 participation by, 170 participation of, 187 preferences of, 36-37 productive capacities of, 91 rights and power of, 182—97 alternative view of, 186-92 conventional view of, 183—86 welfare enhancing effects of, 196-97 shortages, 15 standards, 20-21, 33-34, 62, 63, 85, 90, 139, 182, 196-97 core, 197/1.2 implementation of, 185 international, 198/1.8 market failure in, 192-96 and unit costs, 184—86, 185/ surpluses, 15 well-being of, 94 see also Child labor; Labor productivity Labor productivity, 30, 61, 61 n. 12, 81, 103, 115, 127, 265 and average costs, 60 child vs. adult, 204—10 and convergence, 137—48

INDEX

Labor productivity (continued) and direct investment, 82 environmental variables in, 81 and free labor, 169«. 14 and GDP, 123-24 and human capital investment, 126 increasing, 109-10 and labor rights, 188 lack of convergence in, 19 maximum, 105 and path dependency, 155, 157—58 and pay inequality, 223—25 variations in, 158—59 and wages, 77—78, 106-7, 113, 116n.2, 118/3.9, 128-33, 138-48, 270/3.5 and work environment, 77, 172—74 and x-efficiency, 32 Lamontagne, L., 25/3.6 Lamoureux, N., 257, 259/311 Land to labor ratio, 44 Landes, David, 23 Latin America, 199 Law, 257-58 Law of one price, 138 Least cost behavior, 56 Leibenstien, Harvey, xiii—xiv, 10, 13, 27, 29, 32, 34, 38, 59, 60, 68/3.14, 77, 85-86, 101/3.1, 103-5, 108, 115, 117/3.6, 168/3.6, 187-88, 241, 250-51 Leisure, 45-46, 223 Lester, Richard, 110 Levine, D.I., 180 Liebowitz, S.J., 153—54, 164 Lock-in, 165 Long-run industry supply curve, 37—38, 37/ Low-wage regimes, xii-xiii, 6, 14, 15, 16, 33,194 and growth, 127—33 and market failure, 133-35 and technology, 42—43 McCloskey, Deirdre, 5 Management, xiii, 20, 188, 266 American, 117/3.4 and cooperative work cultures, 107-8, 172,175,178-79, 195 employee ratios, 87 employee relations, 16 and environmental regulation, 236, 245

293

Management (continued) and high- vs. low-wage regimes, 134-35 human resource, 87, 170, 188 and increased wages, 114 and labor standards, 194—95 and path dependency, 156-57 preferences of, 12—13, 14, 38-40, 249 reorganization of, 11 and x-efficiency improvements, 143-44 and x-inefficiency, 34—36, 36/ Marginal costs, 29, 30 Marginal revenue, 29 Marginal revenue product curves, 109,

110 Margolis, S.E., 153—54, 164 Market forces, 15, 90 Married women, 227—28/3/3. 4, 8 Marshall, Ray, 190, 192 Material welfare see Welfare Material Welfare School (MWS), 76-83 Maximization-returns hypothesis, 54-55 Mergers, 22, 60 justifications of, 247 see also Integration Metering costs, 29, 187, 253 Miller, G.J., 25/3.4 Minimum wages, xiii, 18, 62, 63, 103, 108, 109, 116/3.4, 144, 182, 183, 186,190 net effect of, 183 Monitoring costs, 29, 187, 253 Monopoly, 85—86 Mueller, D.C., 258/3.1 Multinational enterprises, 193 Musgrave, Alan, 58 MWS see Material Welfare School Myrdal, Gunnar, 119-20 Natural selection, 55 Negligibility assumption, 58 Nelson, Richard R., 135/3.3, 136/3.7, 137 New United Motor Manufacturing, Inc. (NUMMI), 173-74 Nonallocative inefficiencies, 86 Norms, 260, 266, 269 North, D., 25/3.2, 165, 268 Norway, 101 NUMMI see New United Motor Manufacturing, Inc.

294

INDEX

Oates, W.E., 234, 235, 237 Obesity, x-xi Opportunism with guile, 252—53 Optimal economic behavior, 29 Ordinalist School, 79, 81 Organizational capital, 87—88, 156-57, 161, 174-76, 188, 194, 195, 236 Organizational entropy, 108—9 Origin of Species, The (Darwin), x Ought/is fallacy, 56 Output, 87, 129-30, 175, 237 effect, 112' and income redistribution, 84, 94, 92-94, 93/ maximization, 56 and minimum wage legislation, 183 Output-price effect, 112 Palmer, K., 234, 235, 237 Pareto Efficiency, 72, 75, 84 Pareto Optimality 17-18, 65, 69-72, 73f, 74, 75, 78, 79, 81,83,91,94 Chicago approach to, 94 Path dependency, xiii, 19-20, 150-67, 168/1.2 behavioral approach to, 155-64, 167 standard approaches and criticisms of, 151-55 Pay inequality, 216-26 basic model of, 223—25 discrimination and, 217—20 and effort discretion, 220-21 elimination of, 220 and fair wage hypothesis, 221—22 neoclassical model of, 220-21 persistence of, 224—25, 227/1.7 and x-efficiency theory, 222—23 Peterson, S.R., 235 Pfeffer, Jeffrey, xiii Pigou, Arthur Cecil, 71, 77, 79-85 Pigouvian Optimality, 83-84, 91 Plant reorganization, 11 Pollution, xiv, 230 abatement, 232, 241—44 true costs of, 233 see also Environmental policy Population growth, 44 Porter Hypothesis, 230-31, 232—38, 240, 244 Porter, Michael, 230, 232

Portney, P.R., 234, 235, 237 Poverty, xi—xii, 81 Power, 87, 179 Predictions, 52, 53, 57, 67/1.3 Preference see Choice Pressure, 48/1.7 Prestige, 87 Prices, 143 Prisoner’s Dilemma, 10, 180-81/1.9 Product demand, 180/i. 2 Product labeling, 185, 196 Production and economic welfare, 75—76 minimum cost of, 105 possibility frontier, 4, 73/, 74, 75, 78, 83, 87, 186, 263, 267 Productivity, 83 and culture, 23 and effort discretion, 85—86 and income redistribution, 85 and integration, 248 sustainable differentials in, 138 team component of, 15 and unit costs, 88-89 and x-efficiency, 87-88 see also Labor productivity Profit opportunities, 153/ 154 Profit-maximizing behavior, 48/1.3 Profit-maximizing effect, 112 Profits, 30, 39, 142 Protection see Sheltering strategies Protestant Ethic and the Spirit of Capitalism, The (Weber), 260 Public policy, 6, 104 and child labor, 212—14 and economic theory, 50-51 and incorrect behavior assumptions, 63 mainstream theory and, 3 and survival principle, 65—66 and work cultures, 171—72 and x-efficiency theory, 60 Puritan work ethic, 27, 40,46 Quality, 175 Racism, 228/1.9 Rappoport, R, 79 Rationality, 28, 66/1.1, 180/1.6, 264 Real income transfers, 82—83 Reder, Melvin, 55 Reich, M., 228/1.9

INDEX

Religion, 262-63 Research and development, 126 Risk, 44-45, 46, 154 Romer, Paul, 119, 126 Rosenberg, Nathan, 243 Rubin, Robert, xi Russia, xii Safety standards, 62 Savings, 122, 125 Scale economies, xiv, 247, 253 Schumpeter, Joseph, 27 Scitovsky, T., 48«.5 Scope economies, 253 Search costs, \ \ln.6 Serfdom, 166, 169«.12 Sharabani, Shoshana, xiv Sheltering strategies, 40, 45, 47, 59-60, 103, 108, 143, 165 Shen, T.Y., 149«. 3 Shepherd, W.G., 258«.l Shock therapy, xii Simon, Herbert A., 29, 66/j. 1, 101—2n.2 Skill upgrading, 81 Slavery, 166, 169/1.12 Smith, Adam, 71, 76-83, 85, 91 Social capabilities, 133, 135, 196 Social capital, 264, 266 Solow, Robert, 18-19, 119, 125, 127 Sowell, Thomas, 23, 262 Specialization, 270/2.8 Stability, 122, 132-33, 165 Statman, M., xiv Stigler, George, 48/2.3, 68/2.14, 104, 109-10, 114, 183, 245/2.3 Stiglitz, J.E., xi, xii, 14, 67/2.11 Stress, 48/2.7, 223 Subsidies, 40, 86, 143 Substitution effect, 112 Summers, Larry H., x, 149/2.6 Supply and demand, law of, 120 Survival principle, 17, 50-66, 69, 72, 87 and conventional wisdom, 54—57 and material welfare, 65-66 revisited and revised, 57—58 strong form of, 55—56 weak form of, 55 and x-inefficiency, 89 Sutcliffe, R.B., 149/2.7 Sutton, Robert I., xiii Switzerland, 101

295

Tariffs, 86, 133, 143 Tax breaks, 40 Taxation, 82, 91 Team effects, 15 Technical change, 49/2.12, 89, 102/2.5, 136/2.7, 266 alternatives to, 44 and child labor, 209-10, 215/2.5 and efficiency wage, 161-62 and environmental regulation, 231, 240-44 and growth, 124—25 individuals as agents of, 40-41, 46 and labor standards, 189-90 pace of, 161 and path dependency, 154, 163-64 threshold level of, 243 unit costs and, 41 and wages, 130-32 and x-inefficiency, 43, 46, 161-62 Technology, 5, 11—12, 19, 27, 67/2.4, 88, 129, 137 environmental, 22 and human agency, 40-45 progress effect, 130-32 Teenagers, 116/2.4 Thurow, Lester, 226/2.3 Todaro, M.P., 96-97 Tomer, J.F., 48/2.6 Toyota, 173—74 Trade barriers, 122 and child labor, 202 international, 193 policy, 182 protection, 190, 192 Training, 82, 85 Transaction cost-economizing paradigm, 247-58 and x-eflficiency theory, 250-52 and x-inefficiency, 252—57 Transaction costs, 22, 24/2.1, 25/2.2, 29, 60, 107, 117/2.6, 118/2.11, 253-54 Trust, 67/2.10, 105, 178-79, 187, 252-53 Tyson, L.D., 180/2.2 Unemployment, 62, 110, 118/2.12, 180/2.2, 183, 215/2.5 Unions, 18, 40, 62, 63, 103, 108, 109, 116-17/2.4, 144, 182, 183, 186, 190, 195

296

INDEX

Unit costs, 34, 61, 62, 175-76 and child labor, 201,203—10 and integration, 250-52, 256 and labor standards, 184—86, 185/, 187-92 and level of x-inefFiciency, 105 and technical change, 41 and wages, 88-89 United Kingdom, 173 United States, 173 effect of wages in, 129 income inequality in, xi—xii, 101 slavery in, 166 Utility, 45—46 Utility maximization, 28, 134, 237, 263-64 constrained, 114, 269/1.1 in work cultures, 171 Values, 260, 266, 269 Van der Linde, C., 230 Verdoon’s Law, 135/1.2 Wage differentials, 108—14 Wage effects, 130-33 Wage rates, 11, 16, 34, 40, 41, 47, 61, 82,

88 and average costs, 60, 105-6, 106/ Boserup’s model of, 44 and child labor, 21, 200, 204-10 and culture, 267/ determinants of, 16 and efficiency, 121 and employment growth, 62 exogenous increases in, 103—16 and inefficient organizations, 259/1.5 inflexibility of, 14 and labor demand, 108—14, 111/ and labor productivity, 77—78, 106-7, 116/1.2, 118/1.9, 128-33, 138-48, 270/1.6 and labor rights, 188—92 marginal utility of, 70, 74, 79, 81, 84, 91,92 and market failure, 192-96 microeconomic implications of higher, 77 and path dependency, 157—58 subsistence, 149/2.4 and technology, 43-44, 46, 49/i. 12 and trade, 193—94

Wage rates (continued) unique, 67/1.11, 168/1.8, 180/1.8 and unit costs, 88—89 and work cultures, 89/ worker/employer struggle over, 78-79 and x-inefficiency, 67—68/2.13 see also Minimum wages; Pay inequality Wage shocks, 103-4, 108—15 Wealth maximization, 263 Wealth of Nations, The (Smith), 76 Weber, Max, xiv, 23, 27, 260, 262—63 Welfare, 45-46, 76-83, 237 behavioral theory of optimization, 84-93 conventional view of, 71—76 and income redistribution, 71, 74, 80-81, 91-93 and labor rights, 196-97 Pareto criteria for maximization of, 70 perspectives on, 94, 95/ Pigou’s approach to, 79-84 and Pigouvian Optimality, 8^-84 and superior work cultures, 177—79 Well-being, 17 Wessels, W., 118/1.12 Williamson, O.E., 247—53 Women see Pay inequality Work culture, 20-21, 33, 47, 77, 94, 144, 251, 265 challenges to innovative, 170-79 cooperative, 29, 34, 35, 107, 172—73 and efficiency, 87—88 and incentives, 167 and labor rights, 188—92 multiple, 175-77 neoclassical theory of, 179 and path dependency, 157 significance of external environments on, 180/1. 2 superior, 172—74, 177—79 utility maximization in, 171 and wages, 89/ and x-inefficiency, 174—75 see also Culture Work ethic, 263, 270/1.8 Work practices see Work culture Workers see Labor Working hours, 182 World Development Indicators 2000, xi Wright, G., 136/1.7

INDEX

X-efFiciency, 9, 10, 27, 28, 72 and child labor, 206-7 and culture, 269 defined, 29-30, 32/ 104-5, 127 effect, 112-13, 130 incentives to improve, 34, 45-46 increasing levels of, 11 and integration, 248-49 and labor costs, 61, 62/ and labor productivity, 32 and neoclassical economics, 114—16 and pay inequality, 222—23 and productivity, 87—88 reorganizing for, 40 and superior work eultures, 177—79 and survival principle, 57—58 and trust, 61 n. 10 and wages and working conditions, 33, 38 worker preferences for, 36-37 X-inefficiency, 10-11, 58-64 and corporate bigness, 60, 250-52, 258 and environmental regulation, 231, 236-39, 244-^5

297

X-inefficiency (continued) examples of, 164—66 existence of, 104-8 fundamental critique of, 68«. 14 implications of, 59-60 and increasing competition, 32—33 and labor costs, 61, 62/ management preference for, 34—36, 36/ and market forces, 59 and monopolies, 23, 85—86 persistence of, 17, 33-40, 39, 60, 63, 85-86, 90, 152-53, 168«.2, 187-88 and survival principle, 89 and teehnical change, 43, 46, 161—62 and transaction costs, 252—56 and wage differentials, 138-48 and wage rates, 67-68/7.13, 108-14 and work cultures, 174—75, 179 Yankelovich, D., 118/7.10 Yellen, J.L., 14, 221 Yerkes-Dodson Law, 48/7.7 Young, A. A., 119