Frontiers ofDevelopment Economics

frontmat.p65 12/12/2000, 8:26 AM1

Frontiers ofDevelopmentEconomics

Gerald M. MeierJoseph E. Stiglitz

editors

A COPUBLICATION OF THE WORLD BANK AND OXFORD UNIVERSITY PRESS

the futurein perspective

frontmat.p65 12/12/2000, 8:26 AM3

Oxford University Press

oxford new york athens auckland bangkok bogota buenos airescalcutta cape town chennai dar es salaam delhi florence hong kongistanbul karachi kuala lumpur madrid melbourne mexico citymumbai nairobi paris so paulo singapore taipei tokyo toronto warsaw

and associated companies in

berlin ibadan

2001 The International Bank for Reconstructionand Development / The World Bank1818 H Street, N.W., Washington, D.C. 20433, USA

Published by Oxford University Press, Inc.198 Madison Avenue, New York, N.Y. 10016

Oxford is a registered trademark of Oxford University Press.

All rights reserved. No part of this publication may be reproduced, stored in a retrievalsystem, or transmitted, in any form or by any means, electronic, mechanical, photo-copying, recording, or otherwise, without the prior permission of Oxford UniversityPress.

Manufactured in the United States of AmericaFirst printing December 2000

1 2 3 4 5 03 02 01 00

The findings, interpretations, and conclusions expressed in this study are entirely thoseof the authors and should not be attributed in any manner to the World Bank, to itsaffiliated organizations, or to members of its Board of Executive Directors or the coun-tries they represent. The boundaries, colors, denominations, and other information shownon any map in this volume do not imply on the part of the World Bank Group anyjudgment on the legal status of any territory or the endorsement or acceptance of suchboundaries.

Library of Congress Cataloging-in-Publication Data

Frontiers of development economics : the future in perspective / edited by Gerald M.Meier, Joseph E. Stiglitz.

p. cm.Revised papers and commentary from the symposium, The Future of Development

Economics, held in Dubrovnik in May 1999 and sponsored by the University ofZagreb and the World Bank.

Includes bibliographical references and index.ISBN 0-19-521592-31. Development economics. I. Meier, Gerald M. II. Stiglitz, Joseph E.

HD75 .F77 2000338.9dc21 00-020985

Text printed on paper that conforms to the American National Standardfor Permanence of Paper for Printed Library Materials, Z39.48-1984

frontmat.p65 12/12/2000, 8:26 AM4

vContents

Foreword Nicholas Stern vii

Preface Gerald M. Meier and Joseph E. Stiglitz ix

Introduction: Ideas for Development Gerald M. Meier 1

Gerald M. MeierThe Old Generation of Development Economists

and the New 13Comment Philippe Aghion 51Comment Hla Myint 58

Kaushik BasuOn the Goals of Development 61Comment Paul P. Streeten 87Comment Michael Lipton 94

Irma AdelmanFallacies in Development Theory and Their

Implications for Policy 103Comment David Vines 135Comment Sir Hans Singer 147

Vinod ThomasRevisiting the Challenge of Development 149

Ravi Kanbur and Lyn SquireThe Evolution of Thinking about Poverty:

Exploring the Interactions 183

Shahid Yusuf and Joseph E. StiglitzDevelopment Issues: Settled and Open 227

frontmat.p65 12/12/2000, 8:26 AM5

Pranab BardhanDistributive Conflicts, Collective Action,

and Institutional Economics 269Comment Irma Adelman 291Comment Paul Collier 296

Nicholas CraftsHistorical Perspectives on Development 301Comment Avner Greif 335Comment David Landes 340

Merilee S. GrindleIn Quest of the Political: The Political Economy

of Development Policymaking 345Comment Gustav Ranis 381Comment Timothy Besley 384

Karla Hoff and Joseph E. StiglitzModern Economic Theory and Development 389Comment Gustav Ranis 460Comment Abhijit V. Banerjee 464Comment Debraj Ray 478

Appendixes

Reflections by Nobel Prize Laureates 487A Research Agenda Lawrence R. Klein 489Needed: A Theory of Change Douglass C. North 491Sparks and Grit from the Anvil

of Growth Paul A. Samuelson 492What Is Development About? Amartya K. Sen 506Candidate Issues in Development

Economics Robert M. Solow 514

Reflections by Pioneers 515Pioneers Revisited Sir Hans Singer 517International Trade and the Domestic Institutional

Framework Hla Myint 520The Economics of a Stagnant Population

W. W. Rostow 529The View from the Trenches: Development Processes

and Policies as Seen by a Working ProfessionalArnold C. Harberger 541

Contributors and Commentators 563

Index 565

vi contents

frontmat.p65 12/12/2000, 8:26 AM6

Foreword

This volume is a valuable contribution to our understanding of theevolution of development thought and its relation to development policy.Over the past 50 years, two generations of development economistshave sought to analyze the process of development and to formulatepolicies that might reduce international poverty. Building on their as-sessment of what we doand do notknow about development, thecontributors to this volume emphasize issues that will challenge thenext generation of development academics and practitioners alike.

Development economics is about the big issues: how economies andsocieties grow and change. They are the issues that were at the heart ofthe work of classical economists in particular, Smith, Ricardo, andMarx. The pioneers of development economics, writing soon afterWorld War II, were firmly aware of these intellectual connections androots. They initially recognized the heritage of classical growth eco-nomics. The pioneers were also directly concerned with the role of chang-ing behavior and institutions in the process of development, issues thatthose working in development economics have been emphasizingstrongly in recent years. It is therefore now of great value for a newgeneration of development economists to interact with the earlier gen-erations. They have much to learn, not only in terms of ideas and con-cepts but also in terms of wise judgment on what is important. Of specialrelevance for future examination are the unsettled issues highlighted inthis volume.

Development economists have, throughout the last 50 years, beenstrongly involved with issues of policy. This involvement implies thatthe role of the state must be at center stage. On this subject there hasbeen a fundamental change in development thinking. In the early years,following World War II, there was, broadly speaking, a mistrust of mar-kets, including world markets, influenced in large part by the experi-ence of the Great Depression. There was also confidence in the abilityof government to take an effective and productive role in directing

vii

frontmat.p65 12/12/2000, 8:26 AM7

investment. The experience of the 1950s and 1960s did not supportthis confidence, and the 1970s and 1980s saw strong moves to liberal-ize and privatize.

By the end of the 1990s, we had seen many countries embark onmarket reform. Again, we have learned from experience. We see thatmarket reforms can be, and on the whole have been, an engine of growth.But we have also seen that if they are not supported by sound institu-tions and good governance, they can stall or fail. Hence, the focus nowis on the relations between institutions and markets.

The centrality of policy to development economics also requires speci-ficity on goals. Here the subject has broadened its perspectives. Whileincome distribution was an issue from the early days following WorldWar II, it began to be emphasized more strongly in the 1960s. And in1974, Robert McNamara, then president of the World Bank, in hisAnnual Meeting speech in Nairobi, made overcoming poverty a keygoal for the Bank. The understanding of well-being, and thus poverty,has gone beyond income, and now most of those working in the subjectwould place strong emphasis on improving health and education aspart of the development goals and as instrumental in generating growthof income. More broadly still, development is increasingly seen as ex-panding freedom of choice and action.

This broader perspective on the goals of development and the returnto the interests of some of the pioneers makes this millennial publica-tion particularly timely and significant. The broadening of the agenda,together with the emphasis on institutions, also reminds us that devel-opment economists need to be aware of contributions from the othersocial sciences: they have much to learn from economic historians, po-litical scientists, and anthropologists, for example. The potential fruit-fulness of some of these new interactions is well illustrated in this volume.

The contributors and commentators in the volume provide manyinsights and examine some new questions that make the subject bothmore interesting and more difficult. We have, it is true, learned much inthe last half century about the effectiveness of different kinds of policyinterventions and structures. Indeed, one of the more beneficial ad-vances is that the subject has become much more focused on evidence,in part in response to, and in part due to, the greater availability ofdata. The challenge now is to apply what we know with judgment andwisdom while taking on the very serious research challenges ahead ofus. In this task, the World Bank has a responsibility to work closelywith researchers throughout the developing and industrial world.

Nicholas SternChief Economist and Senior Vice President

for Development EconomicsWorld Bank

viii foreword

frontmat.p65 12/12/2000, 8:26 AM8

Preface

It is now timely and challenging to contemplate the future of de-velopment economics in light of the past half-centurys experience ofdevelopment thought and practice. This volume does so as an outgrowthof an intergenerational symposium on The Future of DevelopmentEconomics, held in Dubrovnik in May 1999 under the sponsorship ofthe University of Zagreb and the World Bank. The Bank had previouslyorganized a series of retrospective lectures by the first generation ofdevelopment economists (approximately 195075). These were pub-lished in Pioneers in Development (1984) and Pioneers in Develop-mentSecond Series (1987).

At Dubrovnik, representatives of the first and the second (approxi-mately 1975 to the present) generations of development economistspresented papers that have been revised for this volume. Most of thecontributors are now instructing the next generation of developmenteconomists. Viewing the past as prologueand as a sequel to the previ-ous two volumes by the Pioneersthey now look toward the unsettledissues that will confront the next generation. About 15 discussants com-ment on the main papers. Two appendixes offer reflections on the fu-ture by several Nobel laureates and first-generation pioneers.

The editors greatly appreciate the efforts of the large number ofcontributors. All who attended the Dubrovnik symposium are alsograteful for the splendid hospitality provided by Professor SoumitraSharma and the Faculty of Economics of the University of Zagreb.The World Bank furnished logistical support for the meeting, but theviews of the contributors are their own and should not be attributedto the World Bank.

The editors have benefited greatly from the assistance of DavidEllerman, Noemi Giszpenc, and Paola Scalabrin at the World Bank andKenneth MacLeod at Oxford University Press. Yuri Woo at StanfordUniversity was especially helpful in maintaining overall control of the

ix

frontmat.p65 12/12/2000, 8:26 AM9

manuscript and in bringing uniformity to a variety of bibliographicalstyles. Her patience and proficiency in dealing with all the details, fromthe first e-mail to the last footnote, were unsurpassed. Without herkeen attention to the needs of so many authors and the editors, themanuscript would not have reached its present form.

Gerald M. MeierStanford University

Joseph E. StiglitzStanford University

x preface

frontmat.p65 12/12/2000, 8:26 AM10

1Introduction:Ideas for Development

Gerald M. Meier

Over the past half-century, we have witnessed an unprecedentedeffort by the international community to accelerate the development ofpoor countries. This effort has been based on an evolution in thinkingabout economic developmentits nature, its causes, and the choice ofpolicies for improving the rate and quality of the development process.Although the development record exhibits many successes, there arealso failures and disappointed expectations. And while the first twogenerations of development economists brought about much progressin the evolution of the subject, many unsettled questions and centralissues remain to be resolved by the next generation. Accordingly, thecontributors to this volume consider the future of development eco-nomics from the perspective of the development record and develop-ment thought.

Ideas as Framework and as Productive Factors

Underlying all the papers collected here is the recognition that ideas arefundamental to the future progress of development. No formula existsfor development. Aid alone cannot yield development. As a former chiefeconomist of the World Bank observed,

[M]ore than ever before, the central priority for the World Bank .. . is to create and help implement improved strategies for eco-nomic development. These strategies must rely, to a greater ex-tent than before, on the transfer and transformation of knowledge,so as to compensate for the expected paucity of development as-sistance. . . . To put it bluntly, since there will not be much devel-opment money over the next decade, there had better be a lot ofgood ideas. (Summers 1991: 2)

chap1.p65 12/12/2000, 8:35 AM1

2 gerald m. meier

Working from the recognition that the knowledge gap between richand poor countries is as significant as the savings gap or the foreignexchange gap, the World Banks World Development Report 1998/99was devoted to the theme of knowledge for development. As JosephStiglitz, the chief economist of the Bank at the time of the Dubrovnikconference, observed,

Today the World Bank has shifted much of its emphasis to theintangibles of knowledge, institutions, and culture in an attemptto forge a more comprehensive New Development Frameworkfor our work. We want, for instance, to be a Knowledge Bank,not just a bank for infrastructure finance. We now see economicdevelopment as less like the construction business and more likeeducation in the broad and comprehensive sense that covers knowl-edge, institutions, and culture. (Stiglitz 1999a)

In a similar vein, the growth economist Paul Romer asserts that

Ideas should be our central concern. . . [I]deas are extremely im-portant economic goods, far more important than the objectsemphasized in most economic models. In a world with physicallimits, it is discoveries of big ideas, together with the discovery ofmillions of little ideas, that make persistent economic growth pos-sible. Ideas are the instructions that let us combine limited physi-cal resources in arrangements that are ever more valuable. (Romer1993b: 64)

Although the World Bank has been an intellectual actor (Stern 1997),ideas for development have come more naturally from university econo-mists and research institutes. This volume considers the future of devel-opment economics from the perspective of the advances in developmentthought made by both World Bank and academic economists. The evo-lution has been along several dimensions of analysis and policy impli-cations. In successive order, the focus has been as shown in Figure 1.

As Irma Adelman observes in this volume, there have been twistsand turns in the evolution of development thinking. Nonetheless, thesubject matter of development economics has evolved with increasinganalytical rigor, and policy implications have become more definitive.Yusuf and Stiglitz, in their chapter, can point to seven major issues assettled and as representing normal science and common wisdom.

The ultimate objective is for appropriate ideas on development to beabsorbed and implemented in developing countries. These ideas includeboth concepts of development policy, in a macro sense, and ideas abouttechnical progress, in a micro or enterprise sense.

The new growth theory emphasizes the role of ideas in promotinggrowth through the aggregation of advances at the micro level. Withina developing country, the implementation of ideas is essential for rais-

chap1.p65 12/12/2000, 8:35 AM2

introduction 3

Fig

ure

1.

The

Evo

luti

on o

f D

evel

opm

ent

Though

t

GO

AL

S O

F D

EV

EL

OPM

EN

T

Gro

ss d

omes

ticpr

oduc

t (g

dp)

Rea

l per

capi

ta g

dp

Non

mon

etar

yin

dica

tors

(Hum

anD

evel

opm

ent I

ndex

)

Miti

gatio

nof

pov

erty

Entit

lem

ents

and

capa

bilit

ies

Free

dom

Su

stai

nabl

ede

velo

pmen

t

MA

CR

OE

CO

NO

MIC

GR

OW

TH

TH

EO

RY

Har

rod-

Dom

aran

alys

isSo

low

sou

rces

of g

row

thN

ew g

row

thth

eory

CA

PIT

AL

AC

CU

MU

LA

TIO

N

Phys

ical

capi

tal

Hum

anca

pita

lK

now

ledg

eca

pita

lSo

cial

capi

tal

STA

TE

AN

D M

AR

KE

T

Mar

ket

failu

res

Non

mar

ket

failu

res

New

mar

ket

failu

res

Inst

itutio

nal

failu

res

GO

VE

RN

ME

NT

IN

TE

RV

EN

TIO

NS

Prog

ram

min

gan

d pl

anni

ngM

inim

alis

tgo

vern

men

t

Com

plem

enta

rity

of g

over

nmen

tan

d m

arke

t

PO

LIC

Y R

EF

OR

M

Poo

r be

caus

epo

or

Poor

bec

ause

poo

rpo

licie

s g

etpr

ices

rig

ht

Get

all

polic

ies

righ

tG

et in

stitu

tions

righ

t

chap1.p65 12/12/2000, 8:35 AM3

4 gerald m. meier

ing total factor productivity. (See the chapters by Meier and by Crafts,in this volume.)

Schumpeters emphasis on innovations is highly relevant for devel-opment. He distinguishes between inventionsthat is, ideas or con-ceptsand innovations (Schumpeter 1939). For development to occur,ideas have to produce new combinations of productive meansthatis, innovations. These include the introduction of a new good or a newquality of a good, the introduction of a new method of production, theopening of a new market, the introduction of a new source of supply, orthe carrying out of new organization of an industry (Schumpeter 1949).Such innovations can offset diminishing returns.

The Schumpeterian type of competition, based on innovations, de-pends on entrepreneurial performance. To accelerate development,the supply of entrepreneurship has to be increased. This depends onideas being accepted as individual knowledge and implemented throughhuman capabilities. Entrepreneurial ability is thus a form of humancapital.1

Beyond contributing to technical change and raising the growth rate,the absorption of ideas may also facilitate the structural transforma-tion of the economy, allow better control of demographic changes, andimprove the distribution of income. In an even deeper sense, scientificideas and rationality can change a societys values and can give supportto modernization.

The backwardness of individuals as economic agents is an unfortu-nate cause and result of poverty. Schooling and training are commonlyadvocated as means of raising creative capacity and inspiring achieve-ment. Beyond this, a facilitating environment can be promoted throughthe establishment and protection (but not overprotection) of intellec-tual property rights, through use of regulatory and tax measures toreward enterprises that innovate, and through intensified competitionamong decentralized enterprises.2

There is certainly an important role for the production of ideas,knowledge, and information by the developing country itself. But inthe earlier phases of the development process, reliance may have to beplaced on the transmission of ideas via international trade, foreign di-rect investment, and technology transfer. Perhaps of even greater valuethan the importation of material goods is the fundamental educativeeffect of trade (Myint 1971). A deficiency of knowledge is a morepervasive handicap to development than is the scarcity of any otherfactor. Knowledge, however, is a global public good (Stiglitz 1999b),and contact with more advanced economies provides an expeditiousway of overcoming this deficiency.3 The importation of technical know-how and skills is an indispensable source of technical progress, and theimportation of ideas in general is a potent stimulus to developmentvital not only for economic change but also for political and socio-

chap1.p65 12/12/2000, 8:35 AM4

introduction 5

cultural advances that may be necessary preconditions of economicprogress. By providing the opportunity to learn from the achievementsand failures of the more advanced economies, and by facilitating selec-tive borrowing and adaptation, foreign trade can help considerably inaccelerating a countrys development.

In the mid-19th century, J. S. Mill observed, It is hardly possible tooverrate the value in the present low state of human improvement, ofplacing human beings in contact with persons dissimilar to themselves,and with modes of thought and action unlike those with which they arefamiliar . . . Such communication has always been and is peculiarly inthe present age, one of the primary sources of progress.4 In the 21stcentury, the pattern and pace of change still differ among countries,thereby allowing trade in ideas to yield dynamic gains from trade. In-deed, Romer (1993a: 543) can assert, Nations are poor because theircitizens do not have access to the ideas that are used in industrial na-tions to generate economic value.5

Although the creation of ideas is a necessary condition for develop-ment, it is not a sufficient condition. The absorptive capacity of thedeveloping country is crucial. If development economists and visitingmissions are not listened to, their ideas will come to naught. The sameis true if ideas on policy reform require political conditions for theirimplementation and these conditions do not exist, or if the absorptivecapacity depends on institutional change that is not forthcoming. Inthe organization of government and the design of information and anincentives system, the preconditions must be in place for the accep-tance and implementation of ideas. The chapter by Grindle in this vol-ume throws additional light on the problem.

The rejection of bad ideas is as important as the acceptance of goodideas. In the 1960s a too-ready importation of the Harrod-Domar analy-sis overemphasized physical capital accumulation and misinterpretedan idea that was meant in a special way for industrial, not developing,countries. The acceptance of ideas about import-substitution indus-trialization also turned out to have adverse consequences. So too, atthe micro level there have been mistaken ideas. Foreign direct invest-ment may bring benefits, but it may also be overly capital intensivewhen there is surplus labor, or multinationals may present costs tothe host country that mount over time and alter the cost-benefit ratiounfavorably.

The absorption of wrong ideas may make it necessary to reverse orend policiesa difficult task. In general, ideas that become embeddedin human capital as knowledge need to be appropriate knowledge,analogous to appropriate technology. Indeed, inappropriate humancapital may be more of a handicap than inappropriate physical capitalbecause human capital cannot be scrapped. A bad practice is not con-veniently ended, and a bad idea may drive out a good idea.

chap1.p65 12/12/2000, 8:35 AM5

6 gerald m. meier

Whether ideas are imported from abroad or produced within thedeveloping countries, they should avoid the biases of ideology. Ideo-logical beliefs have only too easily permeated development thought.With postwar decolonization, development economics was initially of-ten viewed as the economics of resentment or discontent. Center andperiphery were emotivenot logicalcategories. As a policy-orientedand problem-solving subject, development economics was also suscep-tible to the ideologies of both the left and the right. Disciplined think-ing on the proper balance between state and market has only too oftenbeen neglected (see Stiglitz 1999a).

If ideas are to be more influential, they will have to evolve fromrigorous analysis and empirical testing. To this end, multiple sources ofanalytical arguments and empirical evidence should be promoted inboth industrial and developing countries. As in the past, so too in thefuture will ideas for development be improved by learning from experi-ence and subjecting ideas to open debate. The emotive and ideologicalmay then be reduced in favor of disciplined analysis that will strengthendevelopment economics.

Contents of This Volume

With the above premises in mind, we turn to the next chapters, whichlook to future ideas for development. In so doing, we may heedSamuelsons observation (1996: 27): It is a mistake in science to thinkthat any generation arrives at the banquet table late, after the feast hasbeen consumed. Sciences work is never done. Science is a movable feast.One solved problem fans out into many new open questions that beg inturn for solutions. Not only will new issues confront the next genera-tion of development economists: more elegant analytical techniquesshould allow them to refine and extend some of the insights of earliergenerations.

In the next chapter, The Old Generation of Development Econo-mists and the New, Gerald M. Meier summarizes the past and futureof development economics from the viewpoint of the older genera-tion. To put the future of development economics in perspective, Meierreviews the development ideas of the first (approximately 195075)and second (approximately 1975present) generations of developmenteconomists. Against this background, he then considers the unsettledquestions and unfinished tasks for the next generation. These involvethe recognition of an expanded meaning of economic development;more attention to the residual (total factor productivity) in the produc-tion-function approach to the sources of growth; refinement and ex-tension of new growth theories in relation to the economics of ideasand knowledge; interpretation of the right institutions; determina-tion of the sources and consequences of social capital; undertaking of

chap1.p65 12/12/2000, 8:35 AM6

introduction 7

multidisciplinary analysis; recognition of historical lessons; examina-tion of the opportunities and problems being created by globalization;and attention to new perspectives on the interdependence of the stateand the market in the development process.

Meier distinguishes between ordinary neoclassical economic analy-sis of development and a more comprehensive approach that looks tothe operation of large, innovative changes and to political-economyissues in development policymaking. All these issues are subsumed inthe general question of whether development economics is to be re-garded simply as applied economics or whether there is a need for aspecial development theory to supplement general economic theory.

In On the Goals of Development, Kaushik Basu maintains thatnew goals for developmentbeyond simply increasing the rate of eco-nomic growthare implied by the movement toward human devel-opment or comprehensive development. But can these larger socialand political goals be given more precise meaning, let alone be sub-jected to measurement and some operational metric for purposes ofevaluation? This question receives prime attention. To the extent thatincome growth is relevant, Basu suggests that the focus should be onhow the poorest people are faring and on the growth rate of the percapita income of the poorest quintile of the population. Of special in-terest is the relatively ignored subject of the strategic interaction be-tween the goals of different countries and the issue of conditionalmorality that they present. Such an analysis is relevant for the designof coordinated actions by nations to achieve developmental objectives.

Irma Adelman, in Fallacies in Development Theory and Their Im-plications for Policy, identifies three major misconceptions: (a) under-development has but a single cause (whether it be low physical capital,missing entrepreneurship, incorrect relative prices, barriers to interna-tional trade, hyperactive government, inadequate human capital, orineffective government); (b) a single criterion suffices to evaluate devel-opment performance; and (c) development is a log-linear process.Adelman maintains that development should be analyzed as a highlymultifaceted, nonlinear, path-dependent, dynamic process involvingsystematically shifting interaction patterns that require changes in poli-cies and institutions over time.

Three chapters look at how some of the development strategies pro-posed in the World Banks World Development Reports of the early1990s have fared. In Revisiting the Challenge of Development, VinodThomas argues that development outcomes in the past decade confirmthe essential contribution of market-friendly actions but also highlightmissing or underemphasized ingredients. Foremost among the latterare the distribution of human development, the protection of the envi-ronment, globalization and financial regulation, and the quality of gov-ernance. Giving top priority to these issues would mean integrating the

chap1.p65 12/12/2000, 8:35 AM7

8 gerald m. meier

quality dimension into development approaches instead of striving tomaximize short-term growth. It would also replace one-track efforts tohasten the pace of market liberalization and would expand the atten-tion given to consensus building in civil society, along with the concernfor policy changes.

In The Evolution of Thinking about Poverty: Exploring the Inter-actions, Ravi Kanbur and Lyn Squire describe the progressive broad-ening of the definition and measurement of poverty, from commandover market-purchased goods (income) to other dimensions of livingstandards such as longevity, literacy, and health and, most recently, toconcern about risk and vulnerability, and about powerlessness and lackof voice. Kanbur and Squire argue that although there are some corre-lations among these different dimensions, the broadening of the defini-tion significantly changes our thinking about how to reduce poverty.The broader concept expands the set of relevant policies, but it alsoemphasizes that poverty-reducing strategies must recognize interactionsamong policies: the impact of appropriately designed combinations willbe greater than the sum of the individual parts. The authors maintainthat additional research is required to increase our understanding ofthose interactions; that in-depth country case studies are needed to ex-plore the best policy combinations for countries with different prob-lems and different capacities; and that institutional innovations designedto overcome information failures and knowledge gaps need to be care-fully evaluated.

Shahid Yusuf and Joseph E. Stiglitz, in Development Issues: Settledand Open, consider which issues in development economics appearsettled and which require future attention. Settled issues have to dowith the following questions: What are the sources of growth? Doesmacroeconomic stability matter, and how can it be sustained? Shoulddeveloping countries liberalize trade? How crucial are property rights?Is poverty reduction a function of growth and asset accumulation, orare poverty safety nets required? Can developing countries defer ordownplay environmental problems? How closely should the state man-age and regulate development?

Current trends reveal a range of issues that are likely to call forfuture analysis and action. These trends relate to globalization, local-ization, environmental degradation, demographic change, food andwater security, and urbanization. The issues can be grouped under twoheadings: (a) multilevel governance and regulation issues (participa-tory politics, organizational capability, decentralization, inequality, andurban governance) and (b) issues related to managing human capitaland natural resources (cross-border migration, aging and capital sup-plies, management of the global commons, and food and water secu-rity). Convergence of both income levels and human development levelscould be accelerated by responding to these issues. Fresh thinking on

chap1.p65 12/12/2000, 8:35 AM8

introduction 9

governance, institutions, regulatory policies, and measures for manag-ing resources will lead to the highest payoff.

Institutions and incentive structures are also central in PranabBardhans discussion of Distributive Conflicts, Collective Action, andInstitutional Economics. Drawing the connections between the newinstitutional economics and development economics, Bardhan givesparticular attention to some issues that have been neglected in the theo-retical institutional economics literature, in particular, (a) the persis-tence of dysfunctional institutions in poor countries, (b) institutionalimpediments as outcomes of distributive conflicts, (c) the collective ac-tion problems these conflicts exacerbate, and (d) a more complex andnuanced role of the state, to deal with the need for coordination. Theanalysis focuses on the effects of distributive conflicts among differentsocial groups and asymmetries in their bargaining power. In this light,Bardhan explains institutional failures and draws attention to the in-evitable collective action problems at both state and local levels.

In Historical Perspectives on Development, Nicholas Crafts askswhether development economics has much to gain from resuming acloser relationship with economic history. To answer this question, Craftsappraises the legacy of the two older generations of economic histori-ans: the postwar pioneers who linked economic history and develop-ment economics, and the more recent practitioners of the new economichistory of the later 1960s and 1970s. Moving beyond growth regres-sions, the analysis is focusing less on production and more on livingstandardsan area, Crafts suggests, in which development economistsand economic historians can interact fruitfully.

Central to both long-run economic history and development are en-dogenous institutional and technological changes. Crafts emphasizesthe importance of solving agency and appropriation problems in creat-ing an environment conducive to innovation and productivity improve-ment. Looking to future collaboration between historians anddevelopment economists, he advises against forcing patterns of eco-nomic growth and development into the framework of the augmented-Solow neoclassical growth model. Crafts emphasizes institutions butrecognizes that different countries can be expected to diverge in theirinstitutional arrangements. He notes the changing relationship betweengrowth in real wages and in gross domestic product (gdp) per capitaand improvement of living standards. The next generation of develop-ment economists, Crafts observes, should be able to gain more insightsfrom economic history than did the second generation.

Although progress has been made in understanding the intersectionof policies and politics, Merilee S. Grindle (In Quest of the Political:The Political Economy of Development Policymaking) asserts that thereis still much to explain about development policymaking from a politi-cal-economy perspective. Grindle explicates two divergent traditions

chap1.p65 12/12/2000, 8:35 AM9

10 gerald m. meier

of political economy, based on economics and on sociological theory,that offer different interpretations of decisionmaking and the processof policy reform. Neither of these contending paradigms, Grindle ar-gues, is adequate for understanding four real-world puzzles: Why andwhen are politicians interested in supporting policy change? How dopolitical institutions affect the choices made by politicians? How arenew institutions created or transformed? What are the consequences ofnew rules of the game for economic and political interaction?

Grindle asserts that political-economy analyses should be able tomodel reality by reflecting the dynamics of policy interactions in thedesign and implementation of development policy and in the creationor transformation of institutions. Political-economy analysis can thenprovide helpful ideas about what might be done to improve practice,not only for macroeconomic management but also as regards a numberof new issues related to institutional change, social policies, and decen-tralized and participatory forms of governance. The author emphasizesthat if development economists are to understand political decision-making, inquiry into political processes is especially needed. In addi-tion, leadership, ideas, and improved institutions are importantdeterminants of successful policy outcomes that are not yet sufficientlyexplained in political-economy theory.

Throughout this volume, there are two pivotal questions: What arethe forces that can explain the divergence in incomes across countries?What interventions are most likely to promote development? In thefinal chapter, Modern Economic Theory and Development, KarlaHoff and Joseph E. Stiglitz focus directly on these questions. Their an-swers reflect recent advances in the economics of imperfect informa-tion and the economics of coordination failures. Instead of assumingthat information costs are negligible and the capacity to contract islimitless, Hoff and Stiglitz make explicit assumptions about individual-specific information constraints and the set of feasible transactions.Ultimately, development economists have to endogenize the institutionsthat affect information and enforcement costs. This means that institu-tions, history, and distributional considerations do matter and that theanalyst must go beyond the usual fundamentals of resources, technol-ogy, and preferences. Neither government-induced distortions nor lowcapital accumulation have proved adequate for explaining underdevel-opment. Rather than concentrate on differences between industrial anddeveloping countries in levels of physical capital, human capital, orgovernment-induced distortions, Hoff and Stiglitz emphasize that thetwo groups are on different production functions and are organized indifferent ways.

In contrast to earlier models in development economics, the authorsprovide many examples of models with multiple equilibria that canexplain why poor countries may be caught in a low-level equilibrium

chap1.p65 12/12/2000, 8:35 AM10

introduction 11

trap from which they cannot be freed by market forces. They presentthree complementary hypotheses relating to gaps in knowledge, insti-tutions that may be dysfunctional, and an ecological perspective onsocial, economic, and political institutions. They formulate a numberof models with development traps that arise from coordination fail-ures as a result of interaction effects among agents that are not fullymediated through prices. The models indicate multiple Pareto-rankedequilibria.

Of special interest from the standpoint of the chapters emphasis onmodern theory is the demonstration that the modeling of coordinationfailures may provide insights into economic policies for resolving them.From this new perspective, more appropriate policies can be recom-mended that will coordinate good equilibria, affect information, orchange incentives and organizational structures.

The volume closes with two sections that present reflections on thefuture of development economics by Nobel laureates and some originalPioneers.

Notes

1. The central role of entrepreneurship has long been emphasized in devel-opment literature. See Leibenstein (1968).

2. For a discussion of the wide variety of approaches to encouraging theproduction and use of ideas, see Romer (1993b).

3. To the extent that an idea is nonrivalrous but excludable, it is an impurepublic good.

4. Mill (1848): vol. 2, book 3, ch. 17, sect. 5.5. For an analysis of ideas as nonrival goods that give rise to increasing

returns and nonconvexities, see Romer (1993a) and the discussion of newgrowth theory in the chapters by Meier and by Adelman in this volume. Forthe links between trade, innovation, and growth, see Grossman and Helpman(1991).

References

Grossman, Gene, and Elhanan Helpman. 1991. Innovation and Growth in theGlobal Economy. Cambridge, Mass.: mit Press.

Leibenstein, Harvey. 1968. Entrepreneurship and Development. AmericanEconomic Review 58 (2, May): 7275.

Mill, J. S. 1848. Principles of Political Economy, vol. 2, book 3, ch. 17, sect. 5.London.

Myint, Hla. 1971. Economic Theory and Underdeveloped Countries. NewYork: Oxford University Press.

chap1.p65 12/12/2000, 8:35 AM11

12 gerald m. meier

Romer, Paul M. 1993a. Idea Gaps and Object Gaps in Economic Develop-ment. Journal of Monetary Economics 32 (December): 54373.

. 1993b. Two Strategies for Economic Development: Using Ideas andProducing Ideas. In Proceedings of the World Bank Annual Conferenceon Development Economics 1992, 6391. Washington, D.C.: World Bank.

Samuelson, Paul A. 1996. The Age of Bhagwati et al. In Robert C. Feenstra,Gene M. Grossman, and Douglas A. Irwin, eds., The Political Economy ofTrade Policy. Cambridge, Mass.: mit Press.

Schumpeter, Joseph A. 1939. Business Cycles. New York: McGraw-Hill.. 1949. The Theory of Economic Development. Cambridge, Mass.:

Harvard University Press.Stern, Nicholas. 1997. The World Bank as an Intellectual Actor. In Davesh

Kapur, John P. Lewis, and Richard Webb, eds., The World Bank: Its FirstHalf Century, vol. 2, ch. 12. Washington, D.C.: Brookings Institution.

Stiglitz, Joseph E. 1999a. Knowledge for Development: Economic Science,Economic Policy, and Economic Advice. In Boris Plesovic and Joseph E.Stiglitz, eds., Annual World Bank Conference on Development Economics1998. Washington, D.C.: World Bank.

. 1999b. Knowledge as a Global Public Good. In Inge Kaul, IsabelleGrunberg, and Marc A. Stern, eds., Global Public Goods: InternationalCooperation in the 21st Century, 30825. New York: Oxford UniversityPress.

. 1999c. Public Policy for a Knowledge Economy. Remarks at theDepartment for Trade and Industry and Center for Economic Policy Re-search, London, January 27. At .

Summers, Lawrence. 1991. Research Challenges for Development Econo-mists. Finance and Development 28 (3, September): 25.

World Bank. 1999. World Development Report 1998/99: Knowledge for De-velopment. New York: Oxford University Press.

chap1.p65 12/12/2000, 8:35 AM12

13

The Old Generation of DevelopmentEconomists and the New

Gerald M. Meier

Near the end of the 19th century, in a retrospective review ofeconomic theory entitled The Old Generation of Economists and theNew, Alfred Marshall wrote of the relation to which the older gen-eration of economists, which is coming near the close of its activity,stands to the work which appears to lie before the coming generation(Marshall 1897: 115). Now, at the end of the 20th century, we mayattempt a similar exercise for development economists.

This chapter does so by appraising the progress in the evolution ofideas made by the first two generations of development economistsover the past 50 years. The next two sections are devoted to the firstgeneration (roughly 195075) and the second generation (roughly 1975present). The final section outlines the unsettled questions and unfin-ished tasks that the next generation will face. The intention is not topresent yet another survey of the literature but rather to offer a subjec-tive summary appraisal of the past and future of the subject from theviewpoint of the older generations.1

The First Generation

After World War II the subject of development was thrust upon econo-mists as newly independent governments in emerging countries soughtadvice for the acceleration of their development. Political independencecould be obtained from Whitehall, but for economic independence thenew governments of Asia and Africa turned to economists in the UnitedKingdom and America. As a discipline, however, development economicshad to be rediscovered or newly founded.2

At the outset in the 1950s, development economists were more con-fident than now. They formulated grand models of development strat-

chap2.p65 12/12/2000, 8:50 AM13

14 gerald m. meier

egy that involved structural transformation and a correlative role forextensive government involvement in development programming orplanning.3 The models were visionary, looking to the requirements foran increase in per capita real income. Because populationthe denomi-natorwas increasing, the emphasis had to be on a rapid rate of growthof the numerator, gross domestic product (gdp). Capital accumulation,as the necessary requirement, was the central focus of the models. TheHarrod-Domar equation, although originally formulated for conditionsof full growth in an industrial economy, was applied to estimate capitalrequirements in developing countries.4

Growth accounting also emphasized the contribution of capital. Thesimple Solow (1957) decomposition of growth into factor contribu-tions and a residual was based on a differentiation of a productionfunction, Y = F(K, L, t ), where Y is output, K is capital, L is labor, andt is time, to form

Y

Y

.

=FKK

Y( (KK.

+FLL

Y( (LL.

+FtY

.

(Subscripts denote partial derivatives.)The contribution of capital accumulation to growth is measured by

(K./K) multiplied by the share of capital in national income (Stern 1991).

The residualgrowth in total factor productivity (tfp)was left to beexplained exogenously by technical progress.

Other early models of development strategy also featured capitalaccumulation: Rostows stages of growth, Nurkses balancedgrowth, Rosenstein-Rodans external economies and big push,Lewiss unlimited supply of labor and dual-sector model, the Prebisch-Myrdal-Singer hypotheses about terms of trade and import substitu-tion, Leibensteins critical minimum effort thesis, and Chenerystwo-gap model.5

The models and hypotheses had policy implications that involvedstrong state action. To many of the early development economists, aless-developed economy was characterized by pervasive market fail-ures. To correct or avoid market failure, they advocated central coordi-nation of the allocation of resources. The newly expanding subject ofwelfare economics also provided considerable rationale for governmentaction to correct market failures. Furthermore, the structuralist schoolcriticized the market price system by emphasizing rigidities, lags, short-ages and surpluses, low elasticities of supply and demand, structuralinflation, and export pessimism.6

Believing that a developing country did not have a reliable marketprice system, that the supply of entrepreneurship was limited, and thatlarge structural changesnot merely marginal adjustmentswere

chap2.p65 12/12/2000, 8:50 AM14

the old generation and the new 15

needed, the first generation of development advisers turned to the stateas the major agent of change. The government of a developmental statewas to promote capital accumulation, utilize reserves of surplus labor,undertake policies of deliberate industrialization, relax the foreign ex-change constraint through import substitution, and coordinate the al-location of resources through programming and planning.

A growing number of visiting missions and foreign advisers cooper-ated with local planning agencies and industrial development corpora-tions in producing the analyses and policy recommendations underlyingnational development plans. To provide tests for the consistency, bal-ance, and feasibility of plans, they utilized modern techniques of eco-nomic analysis, especially input-output analysis, dynamic programming,and simulation of growth models.

The advocacy of inward-looking policies derived from a belief thatexport earnings were inelastic. This gave support to the two-gap modelof savings-and-investment and the balance of trade (Bruno and Chenery1962). According to this model, extra savings cannot be converted intoimports of capital goods and is therefore frustrated, but foreign aid canallow investment to expand by relaxing the constraint in foreign ex-change. There was also belief in structural inflation, in which the mar-ginal propensity to import exceeds the marginal propensity to export,and in the need for balanced growth (Lewis, in Meier and Seers 1984).

At the same time as pessimistic conclusions were reached about de-veloping countries capacity to export primary products and to pursueexport-led development, optimistic conclusions were expressed aboutcapacity to accelerate development through the extension of the publicsector and through wide-ranging governmental policies. This combina-tion of external pessimism and internal optimism dominated the think-ing of the first generation.

With these macro-strategies it was believed that government couldaccomplish a structural transformation in the developing economy.Government would give reality to the slogans of the first generation bybreaking Nurkses vicious circle of poverty via Rosenstein-Rodansbig push and through balanced growth that would establishcomplementarity in demand, achieve Leibensteins critical minimumeffort, break out of the low-level equilibrium trap, and fulfill theconditions of Rostows takeoff.7

Both the models and the policy advocacy of the first generation weresubsequently criticized.8 The models lacked sufficient empirical con-tent. Moreover, as Krugman observes, the development theorists of the1950s were

at first unable, and later unwilling, to codify [their insights] inclear, internally consistent models. At the same time the expectedstandard of rigor in economic thinking was steadily rising. The

chap2.p65 12/12/2000, 8:50 AM15

16 gerald m. meier

result was that development economics as a distinctive field wascrowded out of the mainstream of economics. Indeed, the ideasof high development theory (of the 1950s) came to seem not somuch wrong as incomprehensible. (Krugman 1993: 29)

Furthermore, in the 1960s the initial concentration on physical capi-tal accumulation was giving way to the concept of investment in hu-man capital and its implications for development. It was increasinglyrecognized that development depended on productive human agentswho, through their acquisition of knowledge, better health and nutri-tion, and increase in skills, could raise total factor productivity.

Above all, criticisms of the early models were reinforced by experi-ence with the adverse effects of government intervention. Economistsbecame increasingly disenchanted with development programming orplanning. Despite the optimism of the earlier generation and the delib-erate efforts of governments to accelerate development, it became onlytoo painfully evident in many countries that mass poverty still existed,that more people were unemployed or underemployed, that the num-bers in absolute poverty were increasing, and that the distribution ofincome and assets was becoming more unequal.

To explain these disappointments, many blamed the policy-induceddistortions and the nonmarket failures resulting from public policies.Particular criticisms were levied at the neglect of agriculture, the ineffi-ciency of state-owned enterprises, the adverse effects of import-substi-tution industrialization, and balance of payments deficits.

In 1952 W. Arthur Lewis could say:

Planning in backward countries imposes much bigger tasks ongovernments than does planning in advanced countries. The gov-ernment has too many things which can in advanced countries beleft to entrepreneurs. It has to create industrial centers, to putthrough an agricultural revolution, to control the foreign exchangesmore strictly, and in addition to make up a great leeway of publicservices and/or ordinary economic legislation. And all this has tobe done through a civil service that is usually much inferior tothat of an advanced country. Why then do backward countriestake more readily to planning? Because their need is also so obvi-ously much greater. And it is also this that enables them to carry itthrough in spite of error and incompetence. For, if the people areon their side, nationalistic, conscious of their backwardness, andanxious to progress, they willingly bear great hardships and toler-ate many mistakes, and they throw themselves with enthusiasminto the job of regenerating their country. Popular enthusiasm isboth the lubricating oil of planning, and the petrol of economicdevelopmenta dynamic force that almost makes all things pos-sible. (Lewis 1952: 128)

chap2.p65 12/12/2000, 8:50 AM16

the old generation and the new 17

By the late 1960s and early 1970s, however, deficiencies in indus-trial programming and comprehensive planning had become acute.Former supporters of development planning began to lament the cri-sis in planning (Streeten and Lipton 1969; Faber and Seers 1972).Critics now pointed to the causes of government failure: deficiencies inthe plans, inadequate information and resources, unanticipated dislo-cations of domestic economic activity, institutional weaknesses, andfailings on the part of the administrative civil service (Killick 1976:164; Chakravarty 1991).

Although the rationale for government interventions had been toremedy market failure, the perverse result was only too often govern-ment failure. This was increasingly evident in the adverse effects ofprice distortionsdistortions that were especially prevalent in wagerates, interest rates, and foreign exchange rates. The policy challengenow became to get prices right. As Timmer (1973) expressed it, get-ting prices right does not guarantee economic development, butgetting prices wrong frequently is the end of development. The logicof choice was again reasserting itself in economic analysis. The secondgeneration of development economists was to give support to a resur-gence of neoclassical economics (Little 1982: chs. 910).

The Second Generation

If the first generation of development economists was visionary anddedicated to grand theories and general strategies, the second genera-tion was almost moralistic, dedicated to a somber realism grounded onfundamental principles of neoclassical economics. Harberger could sayto the governments of developing countries, Economics is good foryouand by economics, he meant neoclassical analysis as the basisfor policymaking (Harberger 1993).

Governments were admonished not only to remove price distortionsbut also to get all policies right. Not differences in initial conditionsbut differences in policies were now thought to explain the disparateperformances of developing countries. A country was not poor becauseof the vicious circle of poverty but because of poor policies. Markets,prices, and incentives should be of central concern in policymaking.

It had become evident that economic rationality characterizes agentsin developing countries as well as in the more developed countries.Claiming that the usual postulates of rationality and the principles ofmaximization or minimization have general applicability, some econo-mists emphasized the universality of neoclassical economics and dis-missed the claim of the first generation that development economicswas a special subdiscipline in its own right. Krueger, for example,maintained:

chap2.p65 12/12/2000, 8:50 AM17

18 gerald m. meier

Once it is recognized that individuals respond to incentives, andthat market failure is the result of inappropriate incentives ratherthan of nonresponsiveness, the separateness of development eco-nomics as a field largely disappears. Instead, it becomes an appliedfield, in which the tools and insights of labor economics, agricul-tural economics, international economics, public finance and otherfields are addressed to the special questions and policy issues thatarise in the context of development. (Krueger 1986: 62 f)9

In accordance with neoclassical economic theory, the second genera-tion moved from highly aggregative models to disaggregatedmicrostudies in which the units of analysis were production units andhouseholds. For offering policy advice, grand theories came to beviewed as less useful than highly specific applications. Microstudies,rather than the broader visionary models of the earlier period, couldprovide more direct policy implications for specific policies such as achange in tariffs or agricultural subsidies (Arrow 1988). There was amarked change from a focus on the process of development to an em-phasis on particular features of underdevelopment. Quantitative ana-lytical tools were used more extensively, especially for empirical analysisof microphenomena that were country specific, sector specific, or projectspecific. The greater availability of microdata sets allowed the model-ing of household behavior and of human capital investments in educa-tion and health.

Studies concluded that how capital is allocated is more importantthan the level of capital accumulation. Growth could be slow despitehigh rates of saving, as in India; high rates of saving were seen to beneither necessary nor sufficient for success. Recognizing the importanceof the allocation of capital, analysts gave more attention to refinementsin the techniques of cost-benefit analysis and shadow pricing that laybehind project appraisal.

In earlier concepts of the aggregate production function, the residualwas thought of as a coefficient of technical advance. The second gen-eration, however, looked at the growth process in a more microeconomicfashion. The residual was recognized to be a composite of the effectsof many different forces: (i) improvements in the quality of labor througheducation, experience and on-the-job training; (ii) reallocation of re-sources from low-productivity to higher-productivity uses, either throughnormal market forces, or through the reduction of barriers or distor-tions; (iii) exploitation of economies of scale; (iv) improved ways ofcombining resources to produce goods and services, not just at the levelof new machines or processes, but also by relatively mundane adjust-ments at the level of the factory or the farm (Harberger 1983: 864 ff).

Numerous studies criticized price distortions, high effective rates ofprotection, and rent-seeking. Not adverse external conditions but inap-

chap2.p65 12/12/2000, 8:50 AM18

the old generation and the new 19

propriate domestic policies explained why some countries were not tak-ing advantage of their external economic opportunities. In contrast,the East Asian newly industrializing economies were viewed as the suc-cess stories of development.10 The correct policies were to move frominward-looking strategies toward liberalization of the foreign trade re-gime and export promotion; to submit to stabilization programs; toprivatize state-owned enterprises; and to follow the dictates of the mar-ket price system. Through its guidance toward the correct policies, neo-classical economics was believed to be the safeguard againstpolicy-induced distortions and nonmarket failures.

Human Capital and the Power of Innovation

With the questioning of how strategic the role of physical capital is,more weight has been given to human capitalto creating agents whocan become more productive through their acquisition of knowledge,better health and nutrition, and increased skills. The focus was on knowl-edge as a source of increasing returns. Near the end of the 19th century,Marshall (1890) had stated that although nature is subject to dimin-ishing returns, man is subject to increasing returns . . . Knowledge is themost powerful engine of production; it enables us to subdue nature andsatisfy our wants. A little later, J. M. Clark observed that knowledgeis the only instrument of production that is not subject to diminishingreturns (1923: 120). Now, at the beginning of the 21st century, thisview is reiterated in the new or newly rediscovered growth theory thattreats knowledge as a nonrival good and emphasizes aggregatenonconvexities associated with investment in knowledge capital(Romer 1986). The theory explains technical progress as being deter-mined by the accumulation of knowledge by forward-looking, profit-maximizing agents (Romer 1986: 1003). It thus supplements Solowsmodel, which emphasized capital accumulation. Although the newgrowth theory is not literally new, it does emphasize newness inproduction functions and in the goods produced (Romer 1994a, 1994b).The introduction of new goods is important to development and raisesthe problem of fulfilling total conditions (the introduction of an indus-try) rather than neoclassical marginal conditions (the production ofadditional units).11

By emphasizing knowledge and ideas, the new endogenous growththeory of the 1980s and 1990s brought about a marked change in theanalysis of aggregate production functions (Romer 1986, 1989, 1990;Lucas 1988). Instead of the early Solow version of diminishing mar-ginal returns to physical capital and to labor separately and constantreturns to both inputs jointly, with technological progress as a residual,the new growth theory examines production functions that show in-creasing returns because of an expanding stock of human capital and

chap2.p65 12/12/2000, 8:50 AM19

20 gerald m. meier

as a result of specialization and investment in knowledge capital.12

Technological progress and human capital formation are endogenizedwithin general equilibrium models of growth. New knowledge is gener-ated by investment in the research sector, and the technological progressresidual is accounted for by endogenous human capital formation andthe increases in the public stock of knowledge. With imperfect compe-tition, the prospect of monopoly profits induces innovation. Knowl-edge or information, once obtained, can be used repeatedly at noadditional cost. The new production processes and products then cre-ate spillover benefits to other firms. [T]he creation of new knowledgeby one firm is assumed to have a positive external effect on the produc-tion possibilities of other firms . . . [so that] production of consumptiongoods as a function of the stock of knowledge exhibits increasing re-turns; more precisely, knowledge may have an increasing marginal prod-uct (Romer 1986: 1003). This allows aggregate investment in the publicstock of knowledge to exhibit increasing returns to scale, to persistindefinitely, and to sustain long-run growth in per capita income.

For developing countries, the new growth theory implies a greateremphasis on human capital (including learning), even more than onphysical capital, and recognition of the benefits from the internationalexchange of ideas that accompanies an open economy integrated intothe world economy. The new growth theory is also relevant for thequestion of convergence.13 Convergence occurs as the technology gapbetween countries is overcome and poor countries catch up with richones by growing faster. Free mobility of capital among countries willspeed this convergence as the rate of diffusion of knowledge increases.

Learning by doing (Arrow 1962) and learning by watching (Kingand Robson 1989) are also knowledge-producing activities and sourcesof scale economies. Unlike growth theory that relies on discrete innova-tions, however, the learning by doing model emphasizes the increasein productivity from the important process of continuous improve-ment (Solow 1997: 40 f, 66 f).

The New Political Economy and the State

The second generation, which was able to reflect on two or three decadesof development experience, recognized the increasing heterogeneity ofdeveloping countries and gave more attention to an explanation of dif-ferential rates of country performance. Cross-country econometric stud-ies of the determinants of economic growth multiplied. A comparativeapproach was adopted in an attempt to understand why certain poli-cies were effective in a given country while others were not, and whythe same type of policy was effective in one country but not in another.

The inquiry into the causes of differential development performanceled to more attention to the politics of policymaking. Elements of a

chap2.p65 12/12/2000, 8:50 AM20

the old generation and the new 21

new political economya positive theory of politicswere formu-lated. In this view, the analytical concepts and principles for interpret-ing why governments do what they do are analogous to those ofneoclassical economic analysis. Postulates of rationality, the concept ofself-interest or self-goal choice, and the techniques of marginal analysisand equilibrium outcomes have been applied to political markets andpolitical objective functions. Whereas the first generation followed theusual approach of normative economic analysis, which assumed thatthe government is composed of Platonic guardians and that the stateacts benevolently in seeking the public interest, proponents of the newpolitical economy now focus on other types of statesthe leviathanstate, the bureaucratic state, or the factional state. To the first genera-tion the government was an exogenous force, but the new politicaleconomy attempts to endogenize the decisions of politicians, bureau-crats, and administrators. It seeks to open windows in the black box ofthe state by using various strands of thought: public choice, collec-tive choice, transaction costs, property rights, rent-seeking, and directlyunproductive profit-seeking activities.

Whether a leviathan, bureaucratic, or factional model of the state isused, the thrust of the new political economy is that an underdevelopedeconomy has commonly given rise to an overextended state and to anegative or exploitative state. This implication appears in writings onprice distortions (rent-seeking and directly unproductive profit-seekingactivities), state-owned enterprises (patronage and bureaucracy), finan-cial repression (politicized credit allocation and cheap credit to sup-porters), agricultural markets (pro-urban bias), inflation (populism),and tariffs and quotas (lobbying).

New Market Failures

A major modification of neoclassical analysis occurred in the 1980sand 1990s, when new market failures were analyzed. The recogni-tion of the existence of imperfect and costly information, incompletemarkets, and transaction costs and of the absence of futures marketsextended the range of market failures beyond the earlier attention topublic goods and externalities that required only selective governmentintervention (Stiglitz 1989b). Risk and information imperfections inthe economy became highly relevant for development analysis. Correc-tion of the new market failures provided a basis for a potential role formore pervasive government intervention. Nonetheless, in the 1990s moreemphasis was actually given to government failures than to marketfailures, and concern about policy reform dominated (Krueger 1990).The recognition of risk and information imperfections did, however,improve analysis of two sectors that had been relatively neglected bythe first generation: agriculture and finance.

chap2.p65 12/12/2000, 8:50 AM21

22 gerald m. meier

Countering the first generations emphasis on industrialization, thesecond generation evaluated policies that would promote rural devel-opment. The effects of government intervention in agricultural pricingbecame a major concern. Numerous studies presented evidence thatmisguided agricultural pricing policies were having an adverse effecton the gap between urban and rural incomes, the incentives to producefood and export crops, the ability of governments to establish foodreserves, and employment opportunities in farming, processing, andrural industries. The theory of rural organization was advanced throughthe use of information, risk, and contract analyses (Binswanger andRosenzweig 1981; Braverman and Guasch 1986; Stiglitz 1986). Themicroeconomics of the rural sector examined the organization and link-ages of labor, land, and credit markets. The recognition of informationconstraints and transaction costs helped explain how rural institutionsare a response to missing markets and also clarified the situations inwhich the potential benefits of government intervention are greatest.14

Decisionmaking by members of rural households was studied from theperspective of the maximization behavior of a household-firm (Singh,Squire, and Strauss 1986).

Financial institutions and financial markets had been neglected bythe first generation. A too facile approach had been taken, in the spiritof Joan Robinsons comment that where enterprise leads, finance fol-lows. Experience with financial bottlenecks and financial repression,however, led the second generation to become more concerned with thedesign of financial systems that would allow the banking system andmoney and capital markets to perform their proper functions of effi-cient investment allocation and financial intermediation between sav-ers and investors. New market failures also gave due weight totransaction costs, adverse selection, and moral hazard in an analysis ofcapital market imperfections and the requirements for more effectivefinancial policies (Stiglitz 1989a).

First-Generation Ideas Revisited

The second generations recognition of new market failures has alsorenewed interest in the first generations models that were concernedwith issues of investment allocation and coordination activities. So,too, have elements of the new growth theory (knowledge, externalities,dynamic increasing returns), the new institutional economics (informa-tion, contracts, response to missing markets), and the new internationaleconomics (imperfect competition, strategic trade theory). This new orextended neoclassical analysis provides a basis for increasing returnsand coordination of externalities resulting from capital accumulation.There is, accordingly, a return to the first generations emphasis on the

chap2.p65 12/12/2000, 8:50 AM22

the old generation and the new 23

importance of increasing returns and pecuniary external economies aris-ing from the effects of market size.

As Krugman now concludes, intellectual credibility can be restoredto a useful set of core ideas from the early analysis of the 1950s.

What was ironic was that a competitive neoclassical orthodoxysettled in on the development front just as the orthodoxy wasbreaking up in other fields. We can now see that whatever badpolicies may have been implemented in the name of high develop-ment theory, the theory itself makes quite a lot of sense. Indeed, insome ways it was a remarkable anticipation of ideas that wouldcome to analytical fruition thirty years later in the field, for ex-ample, of international trade and economic growth. (Krugman1993: 29)

Bardhan (1993: 139) notes that as economic theory has turned moretoward the study of information-based market failures, coordinationfailures, multiple roles of prices and the general idea of the potentialcomplexity of market interactions, it has inevitably turned to questionsthat have long exercised development economists.

Although a more meaningful case can now be made for a big pushor for balanced growth (Murphy, Shleifer, and Vishny 1989), theexperience with government failure has remained dominant in weigh-ing against government intervention. The common consensus in the1990s was for the promotion of policy reform. The state was believedto be overextended. A market price system was needed to get pricesright. And to get policies right, there was a need for stabilization, liber-alization, deregulation, and privatization. Supporting these policies werethe International Monetary Funds (imfs) conditionality for loans andthe World Banks structural adjustment lending.

It is now, finally, recognized that to get prices right and to get poli-cies right, it is also necessary to get institutions right. As North (1997)observes, we now know a good deal about what makes for successfuldevelopment, but we still know very little about how to get thereandespecially how to establish the institutional and organizational structurethat will support the desired rate and composition of economic change.

Tasks for the New Generation

Although, at the end of the 20th century, the second generation of de-velopment economists leaves the subject in a far more advanced statethan at midcentury, there is clearly much unfinished business and manyunsettled questions to be considered by the new generation. It would bepresumptuous and unrealistic to dictate a future research agenda, butwe may suggest some central topics that deserve consideration.

chap2.p65 12/12/2000, 8:50 AM23

24 gerald m. meier

Beyond Income Growth: Patterns of Growthand Income Distribution

The new generation must still begin with an understanding of the mean-ing of economic development. To the old generation, the objective ofdevelopment was an increase in per capita real income (or in a purchas-ing power parity index of per capita income), to be attained by growthof gdp. But it was increasingly realized that development meantgrowth plus change and that change implied other objectives beyondsimple gdp growth. Emphasis on quality growth, or a desired pat-tern of growth, incorporates broader criteria of development such aspoverty reduction, distributional equity, environmental protection, orSens emphasis on entitlements and enhancement of human capa-bilities (Sen 1983) and, more recently, development as freedom (Sen1999). In this broader view the growth of real income and output hasto be put in its place as ultimately an instrumental concerndeeplyconditional on its causal role in enhancing more intrinsically valuedobjects (Sen 1994: 367). Successful development policies need to de-termine not only how more rapid growth of real income can be gener-ated but also how real income can be used to achieve the other valuesincorporated in development.

Not only the rate of growth but also the pattern of growth is rel-evant, especially for better understanding of the role of income distri-bution in the process of development. The persistence of povertyevenwith creditable rates of growthis the shame of inadequate develop-ment policy. The World Bank estimates (World Bank 2000: 25) thatapproximately 1.5 billion people in the developing world are consum-ing less than US$1 a day (at 1985 prices). If poverty is to be reduced,future analysis will have to give more attention to how the pattern ofgrowth determines who are the beneficiaries of growth. Patterns ofgrowth will have to be designed that avoid urban bias, displacement ofunskilled labor, alteration of relative prices to the disadvantage of thepoor, gender gaps, deterioration of child welfare, and the erosion oftraditional entitlements that have served as safety nets. Moreover, inso-far as experience indicates that economic growth does not always leadto widespread improvement in standards of health and education, spe-cial policies that differ from those for simply increasing income willhave to be devised to improve the health and educational attainment ofthe poor (Squire 1993: 379).

Employment Creation

When forces in certain types of growth regimes can plunge some groupsinto poverty, it becomes all the more essential to devise governmentpolicies that are adequate to lift them out of poverty. A central prob-

chap2.p65 12/12/2000, 8:50 AM24

the old generation and the new 25

lem of development will remain surplus labor. The need to create jobswill be especially pressing, given that the worlds labor force will in-crease by 40 percent over the next two decades, with 95 percent ofthe increase in developing countries, where less than 15 percent of theworlds capital investment will occur (Summers 1991: 5). To reducepoverty by increasing productivity and earnings, governments willhave to devise appropriate policies in four crucial sectors of theeconomy: the rural sector, the urban informal sector, the export sec-tor, and the social sector.

Understanding the Sources of Growth

In the future, the criteria of development that should guide policy mayacquire even wider meaning, incorporating for the purpose of bettergovernance such political objectives as the attainment of civil liberties,popular participation, and democracy. Although specific strategies willbe necessary to achieve the nonmonetary objectives, growth and changewill continue to be central to any explanation of the determinants ofdevelopment.

Advancement in determining the sources of growth has been no-table. Because of the importance of total factor productivity, however,future research will have to increase our understanding of the unex-plained residual factor in aggregate production functions. Abramovitz(1956) originally called the residual some sort of measure of igno-rance, but he advised that it was also some sort of indication of wherewe need to concentrate our attention. This is still true. Now, as Sternobserves, We seem to have too many theories claiming property rightsin the unexplained residual, and have no reassurance that any of them,separately or together, really capture what is going on. Just as worryingis that they omit many issues which are probably crucial to growth inthe medium run, including economic organization and the social andphysical infrastructure (Stern 1991: 131).

In the future search for the sources of growth, more attention shouldalso be directed to the joint and interdependent action of the causes ofgrowth (Abramovitz 1993: 237 f). There needs to be analysis of thedependence of both tangible and human capital accumulation on thepace and character of technological progress, as well as in the otherdirectionon how capital accumulation influences technologicalprogress. How do capital accumulation and the advance of knowledge,arising in part from independent or poorly understood sources of growth,work together to produce joint effects (Abramovitz 1993: 236 f)? Moregenerally, growth accounting still has to establish the interactions inthe residual among technological progress, economies of scale and scope,tangible capital accumulation, human capital, knowledge capital, andinstitutional change.

chap2.p65 12/12/2000, 8:50 AM25

26 gerald m. meier

Beyond disaggregating the residual into recognizable elements andcoming to understand the joint and interdependent action of the mainsources of growth, the ultimate task is to determine how these elementsare to yield to policies. Many policies that economists have consideredbear on the supply of inputs, but it will be a more difficult challenge todevise and implement policies to promote the income-raising forces con-stituting the residual.

To explain the residual, or total factor productivity, in growth ac-counting, refinement and extension of the new growth theories will bea major task for the next generation of development economists. Toproceed from capital accumulation to technological progress, as in thenew growth theories, is only part of the story. In the first place, asAbramovitz observes, there is still far too much that is poorly under-stood about the influence of relative factor costs, about the evolutionof science and technology, and about the political and economic insti-tutions and modes of organization on which the discovery or acquisi-tion of new knowledge depends. We cannot reduce the actual advanceof technology, its direction as well as its pace, to a stable function of thesupply of savings and costs of finance alone (Abramovitz 1993: 237).Second, as urged above, there must also be the reverse concern, aboutthe effect of technical progress on the rate of capital accumulation.

The economics of ideas and knowledge require extension. Centralpolicy questions remain to be answered by endogenous innovationmodels. If deliberate technical progress is to be achieved, what is theinstitutional design that will motivate behavior for the creation of knowl-edge? Can governments encourage the production and use of new knowl-edge? What is the search mechanism for the discovery of the mostproductivity-enhancing ideas? Can the sectors or locations be identi-fied in which spillover effects may be large? Given the forces of global-ization, what are the extensions of endogenous growth theory tointernational trade, international capital flows, and the internationaldiffusion of ideas? The developing economies, given their increasingopenness, will be most affected by the international aspects of the newgrowth theory. What are the best institutional arrangements for gain-ing access to the knowledge that already exists in the world? Moreextensive analysis will have to be given to the nonconvexities involvedin the process of diffusion and adoption of new goods and techniquesin a developing economy.15 All of this will have to be integrated withtheories of imperfect competition.

The Influence of Institutions

To understand the heterogeneous experience of countries in achievingdevelopment in the broad sense, it will be necessary to appreciatemore fully the role of organizations and institutions. It is common to

chap2.p65 12/12/2000, 8:50 AM26

the old generation and the new 27

say that institutions matter, and, as advice for overcoming dualism andestablishing a robust market price system, it is now common to sayget institutions right. But what is the meaning of right? And howare the right institutions to be established? The model of the competi-tive ideal world is essentially institutionless and provides little guidancefor the actual establishment of efficient markets. These are importantquestions for the new generations research agenda.

Some preliminary insights have been offered by Douglass North(1994, 1997), who emphasizes that the incentive structure of societywhich is fundamental for the process of changeis a function of theinstitutional structure of that society.16 Institutions are the rules of thegame in society or . . . the humanly devised constraints that shape hu-man interaction (North 1990: 3)not only formal rules (constitu-tions, laws, and regulations) but also informal constraints (norms ofbehavior, conventions, and self-imposed codes of conduct). It is theadmixture of rules, norms, and enforcement characteristics that deter-mines economic performance.

Oliver Williamson (1995) also interprets the new institutional eco-nomics from the perspective of the institutional environmentthemacroanalytics of the political and legal rules of the gameand fromthe microanalytical perspective of the firm and market modes of con-tract and organization. Starting from the objective of economizing trans-action costs, firms and markets establish institutions for governance ofcontracts, investment, and private ordering. There are alternative modesof organization: markets, hybrids, hierarchies, and public bureaus. Eachmode establishes different incentives and controls that lead to differentdegrees of cooperation or competition, credible investment conditions,and credible contracting.17

Neoclassical institutional economists have tended to focus on insti-tutions that improve allocative efficiency and have considered relativeprice changes to be the main motive force for institutional change(Bardhan 1989: 139193). Institutional change, however, also involvesa redistributive change. This raises issues of collective action, bargain-ing power, state capacity (Evans 1992), and political processes that neo-classical institutional economics has ignored.

The future concern with incomplete or missing institutions may alsolead to revision and extension of the first generations dual-sector model.Early on, Myint (1985) suggested that dualism is preeminently a phe-nomenon of an underdeveloped organizational framework, character-ized by the incomplete development not only of the market networkbut also of the governments administrative and fiscal system. In con-trast to the first generations reliance on the limited analysis of a two-sector model, the concept of organizational dualism moves the policyimplications away from getting prices right to an examination ofwhat constitutes the development of appropriate institutions.

chap2.p65 12/12/2000, 8:50 AM27

28 gerald m. meier

So too does the new approach to institutional development in termsof game theory, especially evolutionary and repeated games. As ana-lyzed by Aoki and others in the field of comparative institutional analy-sis, an institution is an equilibrium outcome of a game. Rather thanbeing determined by the polity or by culture, an institution originatesas an endogenously created solution of a game in the economic, social,or political exchange domain. This type of new analysis of comparativeinstitutional issues calls for application of game theory, contract theory,and information economics in comparative and historical contexts (seeAoki forthcoming).

A type of constitutional economics will be required to determinewhat reforms of the political-institutional parameters will provide ap-propriate incentives and rules of behavior for persons entering into ex-change relationships. In a developing economy it is especially importantthat institutions both facilitate and adapt to change (Stern and Stiglitz1997: 273). To do this, and to realize its potential for development, acountry must be capable of much social inventionchanges in ar-rangements by which people are induced to cooperate and participatein economic activity (Kuznets 1966: 5). Or, there must be a sufficientlyhigh level of social capability (Ohkawa and Kohama 1989: 20716).

Catching Up: The Role of Technology and Social Capability

A prominent interpretation of catch-up and convergence empha-sizes the forces of technological congruence and social capabilitybetween the productivity leader and the followers. Abramovitz and David(1996) analyze how these forces relate to a countrys growth potentialand its actual ability to make the technological and organizational leapsthat the convergence hypothesis envisages. The constraints on the poten-tials of countries are divided into two categories. First are limitations oftechnological congruencelimitations that arise because the frontiersof technology do not advance evenly in all dimensions among nations,that is, with the same proportional impact on the productivities of la-bor, capital, and natural resource endowments, on the demands for thedifferent factors of production, and on the effectiveness of differentscales of output. The lagging countries have difficulty in adopting andadapting the current technological practices of the leader. The secondclass of constraints relates to social capability: levels of educationand technical competence; commercial, industrial, and financial insti-tutions; and political and sociocultural characteristics that influencerisk-taking, incentives, and rewards for economic activity.

Abramovitz and David summarize their general proposition as follows:

[C]ountries effective potentials for rapid productivity growth bycatch-up are not determined solely by the gaps in levels of tech-nology, capital intensity, and efficient allocation that separate them

chap2.p65 12/12/2000, 8:50 AM28

the old generation and the new 29

from the productivity leaders. They are restricted also by theiraccess to primary materials and mor