The Global Governance of Trade from a Development Perspective

Home , Global governance

T h e G l o b a l Governance o f Tr a d e

A s I f

D e v e l o p m e n t

Really Mattered

Dani Rodrik

October 2001 UNITED NATIONS DEVELOPMENT PROGRAMME Contents

Executive Summary ...... 5

Introduction ...... 9

Growth versus Poverty Reduction: A Meaningless Debate ...... 12

Achieving Economic Growth: What Really Matters? ...... 14

Trade Liberalization, Growth and Poverty Reduction: What Do the Facts Really Show? ...... 21

The Integrationist Agenda and the Crowding Out of Development Priorities ...... 25

An International Trade Regime That Puts Development First: General Principles ...... 27

Conclusions: From a Market-Exchange Perspective to a Development Perspective ...... 34

Notes ...... 36

References...... 37

Copyright © 2001 United Nations Development Programme (UNDP) Social Development Group Bureau for Development Policy United Nat i o ns Devel o p ment Prog ra m m e One UN Plaza, 20th floor New York,NY 10017 About the Author http://www.undp.org/bdp

This is a background paper to the UNDP project on Trade and Dani Rodrik is the Rafiq Harir i Prof essor of Interna t i o nal Pol i t i c al Econo my at the Sustainable Human Development. Jo hn F. Kennedy School of Govern m e n t , H a rv a rd Univer s i t y. He has published wide- The responsibility for opinions in this paper rests solely with its author, and ly in the areas of interna t i o nal econ om i c s ,e c on omic devel o pm e n t , and political econo - publication does not constitute an my . What constitutes good econo mic policy and why some gover nments are better endorsement by the United Nations Development Programme or the insti- than others in adopting it are the central questions on which his res e a r ch focuses. tutions of the United Nations system. Rodrik is the research coordinator of the Group of 24 (G-24), and is also affili-

Ba ck g r ound papers are designed to gen- ated with the National Bureau of Economic Research and the Centre for Economic er ate discussion and fee d b a c k on issues Policy Research (London). He is the recipient of a Carnegie Scholar award for of trade and sustainable human devel - op ment for a series of cons u l t a t i o ns cul- 2001, and his work has been supported also by the Ford and Rockefeller minating in the UNDP Trade and Foundations. He is the faculty chair of the new MPA in International Human Devel o p ment Report in 2002. Development (MPAID) degree programme at Harvard. He holds a Ph.D. in eco- Design: VanGennep Design nomics and an MPA from Princeton University, and an A.B. from Harvard College. Preface

Events surrounding the WTO Ministerial meeting in Seattle in late 1999 became a kind of Rorschacht test for how different constituencies view globalization—how different people and groups look at the same pictures but draw different meanings from them.Many developing country governments noted the asymmetry in the multilateral trading regime, which they viewed as dominated by a narrow agenda of a few indus- trialized countries, thereby marginalizing the genuine development concerns of the vast majority of the people. Civil society organizations (CSOs) from both the South and North, for their part, were equally upset that their constituencies’ many concerns were once again excluded from the intergovernmental discussions and negotiations. The breakdown in Seattle opened up the opportunity for a much-needed breath- ing space to discuss and debate the significance of trade for achieving the Millennium Development Goals (MDGs).The controversy surrounding the global trading system is not about whether trade is necessary, but about how the multilateral trade regime can operate in ways that support and foster human development. As the dust settled on Sea t t l e , we were convinced that given UNDP’s vanguard rol e in advocating for human devel o p ment and its 1999 Human Devel o p ment Report on Gl o b a l i za t i on , our organizat i o n had a special res p on s i b i l i t y to cont r ibute to the trad e de b a t e . Our res p o nse was to conc e p t u a l i z e, design and implement a pro j e c t ,w h i ch cam e to be known as UNDP’s Trade and Sustainable Human Devel o p ment proj e c t . The project was approved in June 2000 and has four main pha s e s ; fi r s t , the com- mi s s i o ning of sever al respected scholars and experts to write consultant papers on differ ent aspects of trade and its global gover nance from a human devel o p ment perspec- ti ve ; se c on d , the convening of an advisory team of conc e r ned and interna t i on a l ly respected gover nment trade negotiators and diplom a t s ,a ca d e m i c s , civil society activists and senior UN colleagues to cri t i ca lly assess the consultant paper outlines and advise on the overa l l project strat e g y; th i rd , the use of the draf t papers as inputs into a serie s of cons u l t a t i o ns with both developing country gover nments and civil society organiza- ti on s , both to obtain their fee d b a c k on them and understand their conc e r ns more fully; and last but not least, dr awing upon all of these and other inputs, to prep a r e a UNDP rep o r t tentativel y entitled ‘Trade and Sustainable Human Devel o pm e n t . ’

The UNDP project has had three interrelated objectives: •To assist developing country governments and civil society organizations in ensuring that their countries can selectively and strategically seize the opportunities of global economic and trade integration for advancing national progress in human development and poverty eradication; •To strengthen the participation and substantive negotiating and advocacy posi-

3 P re f a c e

tions of developing countries in the debate and negotiations on the emerging global trading regime; •To present a UNDP position on the human development outcomes of the current global trading regime and the reforms needed to make it more inclusive and balanced,thereby enabling trade to become an instrument for enhancing human development and reducing poverty. While consultations continue and UNDP’s report is under preparation, the three consultant papers commissioned as part of the project are being made available.Indeed, an important part of the commitment of the project was to publish, in their independent right,each of the papers. We believe that they deserve to be widely read and used to inform the current debate on trade and development. This paper, by Professor Dani Rodrik of Harvard University, analyses the global governance of trade from a development perspective. Professor Rodrik looks at trade through a development lens, with particular emphasis on assessing the relationship between trade, growth and poverty. He provides an analysis of the strengths and weak- nesses of the existing trading system. The assumptions underlying trade liberalization and its relationship with growth and poverty are critically analysed. Based on the evidence, the paper makes proposals for how the multilateral trade regime and its agreements and practices need to change to better serve the goals of human development. The paper makes suggestions to developing countries on a range of pertinent issues, including on the crucial question of the degree of trade openness which is likely to be consistent with development objectives and outcomes under different country circumstances. We hope the reader will find the paper informative and useful as a contribution to the ongoing debate on trade and development.

Eimi Watanabe Assistant Administrator and Director Bureau for Development Policy UNDP October 2001

4 Executive Summary

It is widely accepted, not least in the agreement establishing the Wor ld Trade Organizat i o n (WTO) , that the purpose of the world trade regime is to raise living standards all around the wo rl d — r ather than to maximize trade per se. In c re a s i n g l y, how e ver , the WTO and mul t i l a t - er al lending agencies have come to view these two goa l s — p r omoting devel o p ment and maximizing trade—as syno nymo u s , to the point where the latter easily substitutes for the forme r . The net result is a confounding of ends and means. Trade has become the lens throu g h wh i c h devel o p ment is perce i ve d , rather than the other way arou n d . Imagine a trading regime in which trade rules are determined so as to maximize devel - op ment potential, pa rt i c u l a r ly that of the poorest nations in the world . Instead of asking, ‘Ho w do we maximize trade and market access?’ ne g otiators would ask, ‘Ho w do we enable co u n t r ies to grow out of pover ty? ’ Would such a regime look differ ent than the one that exists curren t ly ? The answer depends on how one interpr ets recent econo mic history and the role that tr ade openness plays in the course of econo mic devel o pm e n t . The prevailing view in G7 capitals and mul t i l a t e r al lending agencies is that econo mic growth is dependent upon integra t i o n into the global econom y . Successful integrat i o n in turn req u i r es both enhanced market access in the advanced industrial countries and a range of institutional ref o r ms at home (r anging from legal and administrat i v e ref o r m to safet y nets) to render econo mic openness viable and growth - p r omo t i n g . This can be cal led the ‘enlightened standard view’— e n l i g h t - ened because of its rec o g n i t i o n that there is more to integrat i o n than simply lowe r ing tarif f and non- t a r iff barriers to trad e , and standard because it rep r esents the conven t i o nal wisdom. In this conc e p t i on , the WTO’ s focus on expanding market access and deepening integrat i o n th r ough the harmon i za t i o n of a wide range of ‘t ra d e - re l a t e d ’ pr actices is prec i s e l y what devel - op ment req u i re s . This paper presents an alterna t i v e account of econo mic devel o pm e n t , one which ques- ti o ns the central i t y of trade and trade policy and empha s i z es instead the crit i c al role of do mestic institutional innova t i on s . It argues that econo mic growth is rare l y sparked by im p o r ted blueprints and opening up the econo my is hardly ever crit i c al at the outset. In i t i a l ref o r ms instead tend to combine unconven t i o nal institutional innova t i o ns with some elements from the orth o d o x rec i p e . Th e y are country- s p e c i f i c , based on local knowledge and ex p e ri m e n t a t i on .T h e y are targeted to domestic investors and tailored to domestic institu- ti o nal rea l i t i e s . A rei nv i go r ated focus on devel o p ment and pover ty red u c t i on ,a l o ng with an empirica l ly based understanding of the devel o p ment proc e s s , would have far-rea c hing implicat i o ns for the manner in which the interna t i o nal trading regime and the WTO function. This paper makes the case for such a reo ri e n t a t i on , arguing that developing countries are short- ch a n g i n g th e m s e l ves when they focus their complaints on specific asymm e t r ies in market access (tarif f

5 Dani Rodrik

peaks against developing country exports ,i n d u s t r ial country prot e c t i o n in agric u l t u r e and te x t i l e s ,e t c . ) . Th e y would be better served by pressing for changes that enshrine devel o p - ment at the top of the WTO agenda, and thereby provide them with a better mix of enhanced market access and roo m to pursue approp r iate devel o p ment strat e g i e s .

Growth vs. Poverty Reduction The paper first takes up the debate about whether growth or pover ty red u c t i o n strategies yie l d the greatest benefits, and argues that the distinction is not significan t , since policies targeted at the poor general ly tend to have growth payo f fs .E ven so, po ver ty red u c t i o n is a worth w h i l e po l i c y goal in itself, for three rea s on s : 1) growth is not a sufficient measure of social welfare, since it ignores both the level and distrib u t i o n of income , and competing growth strat e g i e s ma y have differ ent payof f s for the poor; 2) interven t i o ns to help the poor may be the best way to raise aver age income s , since they seek to close gaps between private and social costs; and 3) policies that target pover ty red u c t i o n seek to maximize people’s cap a b i l i t i e s ,i n c luding those of the poor, thus cont r ibuting to more sustainable devel o pm e n t . The problem with current trad e rules is not that they over - e m ph a s i z e trade and growth at the expense of pover ty red u c t i on , but that they over - e m ph a s i z e trade at the expense of pover ty red u c t i o n an d gr owth .

Alternative Development Strategies Tur ning to the determinants of econo mic growth , the paper discusses the enlightened stan- da r d view, wh i c h grew out of the failures of the Was h i n g t o n Consensus policies of the 1980s and 1990s. This view goes beyond liberal i za t i o n and priv a t i za t i o n to include the need for financial reg u l a t i o n and supervi s i on , legal and administrat i v e ref o rm , labour market flexibility and social safet y nets. Its ref o rm s ,h ow e ver , ar e biased towa r ds an Ang l o - Am e ri c an conc e p - ti o n of institutional soundness and are driv en largely by the req u i r ements of integrat i o n into the world econom y . Needed instead is an approa c h that empha s i z es domestic institutiona l in n ov a t i o ns (comp r ising a mix of orth o d o xy with ‘lo c al heresies’) and of investment strat e g i e s ta i l o r ed to each country. This argument is supported by an examination of three types of successful devel o pm e n t st ra t e g i e s : 1) import substitution strat e g i e s , based on (temporar y) import prot e c t i o n for home pro d u c e r s , as done successfully in scores of developing countries during the 1960s and early 19 7 0 s ; 2) outwa r d oriented industria l i za t i o n strat e g i e s , as pursued by the East Asian tigers in the 1980s, in which export led growth was made possible by gover nment support of priv a t e in ves t m e n t ,i n c luding credit subsidies, tax incentive s ,d u ty free access to inputs and cap i t a l goods as well as educat i o nal and infras t ru c t u r al devel o pm e n t ; and 3) two - t ra c k ref o r m st ra t e g i e s , as pursued for example by China and Mauritius in the late 1970s, that comb i n e ma r ket liberal i za t i o n and state reg u l a t i o n in differ ent ways.

Trade Liberalization, Growth and Poverty Reduction The third section examines the literat u r e on trade policy and econo mic perfo rm a n c e , wh i c h fo r ms the basis for the oft- h e a r d statements on the benefits of trade openness, and concl u d e s that there is no convincing evidence that trade liberal i za t i o n is pred i c t a b l y associated with subsequent econo mic growth . The claims for such links arise from the misattrib u t i o n of ma c ro e c on o mic phe n o mena (over valued curren c i e s ,m a c ro e c on o mic instability) ,i n s t i t u t i on a l

6 The Global Governance of Trade

fa i l u re s , or geograph i c al locat i o n to trade policies. The onl y systematic rel a t i o nship is that co u n t r ies dismantle trade res t ri c t i o ns as they get rich e r , wh i c h accounts for the fact that most of today’s ric h countries embarked on econo mic growth behind prot e c t i v e barrie r s ,w h i c h th e y subsequently lowe re d . This raises serious questions about the prio ri t y placed on integra- ti o nist policies in orth o d o x ref o r m prog ra m m e s . The problem is not trade liberal i za t i o n per se , but the diver s i o n of financial res o u r ces and political capital from more urgent devel o p - ment prio ri t i e s . To elaborate this point, the paper next presents some trad e - o f f s faced by devel o p i n g co u n t r ies in deciding to implement WTO agree m e n t s ,w h i c h reflect little conc e r n for needed de vel o p ment prio ri t i e s .B i l a t e r al and reg i o nal trade agreements are often worse in terms of the ob l i g a t i o ns req u i r ed in exchange for enhanced market access. Mo re o ver , their emphasis on eliminating the state from the formul a t i o n or reg u l a t i o n of econo mic policy undermines state cap a c i t y to undertake the institutional ref o r ms necessary to benefit from global integrat i on .

General Principles The final section of the paper develops some general principles for a world trade regime that puts development first. Such a regime would accept institutional diversity and the right of countries to ‘protect’ their institutional arrangements—so long as they do not seek to impose it on others. Once these principles are accepted and internal- ized in trade rules, priorities of poor nations and the industrial countries can be ren- dered compatible and mutually supportive. An ‘opt-out mechanism’ would essentially extend the existing safeguard agreement to permit countries to restrict trade or suspend WTO obligations for reasons that include social and distributional goals as well as development priorities. This would require replacing the serious injury test with the need to demonstrate broad domestic support for the proposed measure among all rele- vant parties–including exporters and importers as well as consumer and public interest groups–and could be complemented by WTO monitoring as well as an automatic sunset clause. The WTO is devoted largely to bargaining over market access. ‘Free trad e ’ is not the typi c al outcome of this proc e s s ; nor is consumer welfare what negotiators prio ri t i ze . In s t e a d , the negotiating agenda has been shaped in res p o nse to a tug-of-war between exporters and mul t i n a t i o nal corpo ra t i o ns in the advanced industrial countries on one side, and import- co mpeting interests (typi ca l ly,but not solely,labour) on the other.The differ ential trea t m e n t of manufactures and agric u l t u r e, or of clothing and other goods within manufacturin g , th e anti-dumping reg i m e , and the intellectual prop e r ty rights (IPR) reg i m e , for example, ar e all a result of this po l i t i c al pro c e s s .T h e r e is little in the struc t u r e of the negot i a t i o ns to ensure that their outcomes are consistent with devel o p ment goa l s , let alone that they seek to furth e r de vel o pm e n t . One result of a shift to a devel o p ment focus would be that developing nations arti c u l a t e their needs not in terms of market access, but in terms of the policy autono my needed to exe r cise institutional innova t i on s . Another is that the WTO should function to manage the in t e r face between differ ent national systems rather than to reduce national institutional differe n c e s . The most obvious advantage would be a more devel o pm e n t - f ri e n d ly interna t i on a l ec on o mic envi r onm e n t . Co u n t r ies would be able to use trade as a means for devel o pm e n t ,

7 Dani Rodrik

rather than being forced to view trade as an end in itself (and thereby sacrifice devel o pm e n t goa l s ) . It would save developing countries precious political capital by obviating the need to bargain for ‘special and differ ential trea t m e n t ’—a principle that in any case is more form than substance at this point. In addition ,v i ewing the WTO as managing institutional diver s i t y gets developing coun- tr ies out of a negotiating conu n d r um that arises from the incons i s t e n c y between their demands for flexibility to implement their devel o p ment policies, on the one hand, and their co mplaints about Nort h e r n prot e c t i o nism in agric u l t u r e, te x t i l e s , and labour and envi r onmental standards , on the other. As long as the issues are viewed in market-access terms , de veloping countries will remain unable to defend their need for flexibility. And the onl y wa y they can gain enhanced market access is by res t r icting their own policy autono my in ex cha n g e . Once the objective of the trade regime is viewed as letting differ ent national ec on o mic systems prosper side by side—the debate can centre on each nation’s institutiona l pri o r ities and how they may be ren d e r ed comp a t i b l e . Fin a l ly, the shift in focus provides a way to rec o ncile the perspectives of devel o p i n g co u n t r y govern m e n t s ,w h i c h complain about asymm e t r y in trade rul e s , and civil society or g a n i za t i on s ,p ri m a ri l y in the Nort h ,w h i c h charge that the system pays inadequate atten- ti o n to values such as tran s p a re n c y, ac c o u n t a b i l i t y, human rights and envi r onmental sustain- ab i l i t y. The often conflicting demands of these two grou p s — o ver issues such as labour and env i r onmental standards or the tran s p a re n c y of the dispute settlement proc e s s — h a v e pa ra l yzed the mul t i l a t e r al trade negot i a t i o n process and allo wed the advanced industria l co u n t r ies and the WTO leadership to seize the ‘mi d dl e ’ gro u n d . Ten s i o ns over these issues become manageable if the debate is couched in terms of de vel o p ment proc e s s e s — b ro a d ly defined—instead of the req u i r ements of market access. Vie wing the trade regime—and the gover nance cha l lenges it poses—from a devel o pm e n t pe r s p e c t i v e makes clear that developing country gover nments and NGO critics share the same goa l s :p o l i c y autonom y , po ver ty red u c t i on , and envi r onm e n t a l ly sustainable human de vel o pm e n t .

8 The Global Governance of Trade ...the WTO would no As If Development Really Mattered longer serve as an

instrument for the har-

In t ro d u c t i o n monization of economic What objectives does (or should) the Wor ld Trade Organizat i o n (WTO) serve? Th e policies and practices first substantive parag ra p h of the Agreement establishing the WTO lists the follow i n g as p i ra t i on s : across countries, but

raising standards of living, en s u r ing full employment and a large and steadily grow- become an organization ing volume of real income and effec t i v e demand, and expanding the prod u c t i o n of and trade in goods and servi c e s , while allo wing for the optimal use of the world ’s that manages the inter- res o u r ces in accordance with the objective of sustainable devel o pm e n t , se e k i n g face between different both to protect and pres e r ve the envi r onment and to enhance the means for doing so in a manner consistent with their res p e c t i v e needs and conc e r ns at differ ent lev- national practices and els of econo mic devel o pm e n t . (WTO 1995:9) institutions. A subsequent parag ra p h cites ‘mut u a l ly advantageous arrangements directed to the substantial red u c t i o n of tarif f s and other barriers to trade and to the elimination of dis- cri m i n a t o r y treatment in interna t i o nal trade rel a t i on s ’ as a means of ‘con t r ibuting to these objectives ’ (i b i d . ) . It is clear from this preamble that the WTO’ s framers placed pri o ri t y on raising standards of living and on sustainable devel o pm e n t . Ex p a n d i n g tr ade was viewed as a means towa r ds that end, rather than an end in itself. Re c e n t l y, pr omoting econo mic devel o p ment has acquired an even higher standing in the official 1 rhe t o r ic of the WTO, pa rt l y in res p o nse to its crit i c s . That the purpose of the world trade regime is to raise living standards all around the wo rl d — r ather than to maximize trade per se—has never been cont r over s i a l . In prac t i c e , how e ver , these two goa l s — p r omoting devel o p ment and maximizing trad e — h a v e increa s - in g l y come to be viewed as syno nymous by the WTO and mul t i l a t e r al lending agencies, to the point where the latter easily substitutes for the forme r . As the WTO’ s Mike Mo o r e (2000) puts it, ‘the surest way to do more to help the poor is to continue to open ma rk e t s . ’ This view has the apparent merit that it is backed by a voluminous empirica l li t e ra t u r e that identifies trade as a key determinant of econo mic growth . It also fits nicely with the trad i t i o nal modus operandi of the WTO, wh i c h is to focus pred om i n a n t l y on

This is a paper prepared for the UNDP. I thank Kamal Malhotra, Yilmaz Akyuz, Murray Gibbs, Gerry Helleiner, Gita Sen, UNDP staff and participants in a brainstorming meeting held in New York City 13-14 October 2000 for guidance and suggestions.The paper draws extensively on several of my previous writings, including, most notably, Rodrik 2001a, 2001b, 2000a, 2000b and 1999. 9 Dani Rodrik

rec i p ro c al market access (instead of devel o pm e n t - f ri e n d ly trade rul e s ) . How e ver , th e net result is a confounding of ends and means. Trade becomes the lens through which de vel o p ment is perce i ve d , rather than the other way arou n d . Imagine a trading regime that is true to the WTO prea m b l e , one in which trad e rules are determined so as to maximize devel o p ment potential, pa rt i c u l a r ly of the w o rl d’s poorest nations . Instead of asking, ‘Ho w do we maximize trade and marke t ac c e s s ? ’ ne g otiators would ask, ‘Ho w do we enable countries to grow out of pover ty? ’ Would such a regime look differ ent from the one that exists curren t ly ? The answer depends on how one interpr ets recent econo mic history and the rol e that trade openness plays in the course of econo mic devel o pm e n t . The preva i l i n g vi e w in G7 capitals and mul t i l a t e r al lending agencies is that integrat i o n into the global econo my is an essential determinant of econo mic growth . Successful integra- ti o n in turn req u i r es both enhanced market access in the advanced industrial coun- tr ies and a range of institutional ref o r ms at home (ranging from legal and adminis- tra t i v e ref o r m to safet y nets) to render econo mic openness viable and growth promo t i n g . This can be reg a r ded as the ‘enlightened standard view’— e n l i g h t e n e d be c ause of its rec o g n i t i o n that there is more to integrat i o n than simply lowe r ing tarif f and non- t a r iff barriers to trad e , and standard because it rep r esents the prevailing conven t i o nal wisdom (see Wor ld Bank/IMF 2000) In this conc e p t i on , to d a y’s WTO rep r esents what the doctor orde re d : the focus on expanding market access and deepening integrat i o n through the harmon i za t i o n of a wide range of ‘t ra d e - re l a t e d ’ pra c - tices is prec i s e l y what devel o p ment req u i re s . This paper presents an alterna t i v e account of econo mic devel o pm e n t , one that qu e s t i o ns the central i t y of trade and trade policy and empha s i z es instead the cri t i ca l role of domestic institutional innova t i o ns that often depart from prevailing orth o - dox y . In this view, tra n s i t i o ns to high econo mic growth are rare l y sparked by blue- pr ints imported from abroa d . Opening up the econo my is hardly ever a key factor at the outset. The initiating ref o r ms instead tend to be a comb i n a t i o n of unconven t i o n- al institutional innova t i o ns with some of the elements drawn from the orth o d o x rec i p e . These comb i n a t i o ns tend to be country- s p e c i f i c , req u i r ing local knowl e d g e and experim e n t a t i o n for successful implementation. Th e y are targeted on dome s t i c in vestors and tailored to domestic institutional rea l i t i e s . In this alterna t i v e view, a devel o pm e n t - f ri e n d ly interna t i o nal trading regime is one that does muc h more than enhance poor countrie s ’ access to markets in the advanced in d u s t r ial countrie s . It is one that enables poor countries to experiment with institu- ti o nal arrangements and leaves roo m for them to devise their own , po s s i b l y diver g e n t so l u t i o ns to the devel o p mental bottlenecks that they face. It is one that evaluates the demands of institutional ref o r m not from the perspective of integrat i o n (‘What do co u n t r ies need to do to integrate?’) but from the perspective of devel o p ment (‘What do co u n t r ies need to do achi e ve broa d - b a s e d , equitable econo mic growth ? ’ ) . In this vision, the WTO would no longer serve as an instrument for the harmon i za t i o n of econom i c policies and practices across countrie s , but become an organizat i o n that manages the in t e r face between differ ent national practices and institutions . This paper argues that a ren e wed focus on devel o p ment and pover ty red u c t i on ,

10 The Global Governance of Trade

al o ng with an empirica l ly-based understanding of the devel o p ment proc e s s , wo u l d ha v e far-rea c hing implicat i o ns for the way in which the interna t i o nal trading reg i m e and the WTO function. It focuses on broad prin c i p l e s , rather than specific rec o m- me n d a t i on s , be c ause it is onl y through a change in the overa l l perspective of trad e ne go t i a t i o ns that significant change can be accomp l i s h e d . One of the key prop o s i t i o ns is that developing countries are short- ch a n g i n g th e m s e l ves when they focus their complaints on specific asymm e t r ies in marke t access (tariff peaks against developing country exports , in d u s t r ial country prot e c t i o n in agric u l t u r e and textiles, et c . ) . This approa c h reflects acceptance of a marke t - a c c e s s pe r s p e c t i v e that does developing countries limited goo d . Th e y would be far better se r ved by pressing for changes that enshrine devel o p ment at the top of the WTO ag e n d a , and corres p on d i n g l y provide them with a better mix of enhanced marke t access and manoeuvring roo m to pursue approp r iate devel o p ment strat e g i e s . Since this paper is as muc h about the approa c h to devel o p ment that should inform vi e ws about the interna t i o nal trade regime as it is about the WTO itself, muc h of the di s c u s s i o n is devoted to the empiric al content of these ideas. The paper begins with an as s e rt i o n that the distinction between devel o p ment strategies that focus on growth ver - sus those that focus on pover ty red u c t i o n is a false one , since in prac t i c e , the two ends ar e inseparab l e . The main strike against existing trade rules is not that they over - em ph a s i z e trade and growth at the expense of pover ty red u c t i on , but that they over - em ph a s i z e trade at the expense of pover ty red u c t i o n and gr owth . It then argues that the enlightened standard devel o p ment model encompasses an impossibly broad and unfo- cused devel o p ment agenda, and one that is biased towa r ds a particular set of institu- ti o nal arran g e m e n t s . It empha s i z es instead the central i t y of domestic institutiona l in n ov a t i o ns (comp r ising a mix of orth o d o xy with ‘lo c al heresies’) and of inves t m e n t st r ategies that are tailored to the circumstances of each country. Mu c h of the paper focuses on the link between trade policy and econo mic perfo rm a n c e . The voluminous literat u r e in this area , wh i c h forms the basis for the oft- he a r d claims to the benefits of trade openness, is , up o n examination, less unequivo- cal . A close look reveals that there is no convincing evidence that trade liberal i za t i o n is pred i c t a b l y associated with subsequent econo mic growth . This raises serious ques- ti o ns about the prio ri t y that the integrat i o nist policy agenda typi ca l ly rec e i v es in ort h o d o x ref o r m prog ra m m e s . The problem is not trade liberal i za t i o n per se, but the di ve r s i o n of financial res o u r ces and political capital from more urgent and deservi n g de vel o p mental prio ri t i e s . Fin a l ly, the paper offers some general principles for a world trade regime that puts devel o p ment first. Fir s t , the trade regime must accept, rather than seek to elimi- na t e , in s t i t u t i o nal diver s i t y, al o ng with the right of countries to ‘pro t e c t ’ their institu- ti o nal arran g e m e n t s . How e ver , the right to protect one ’s own social arrangements is distinct from, and does not extend to, the right to impose it on others. Once these simple principles are accepted and interna l i z ed in trade rul e s , de vel o p mental prio r i- ties of poor nations and the needs of the industrial countries can be ren d e r ed compatible and mut u a l ly support i ve .

11 Dani Rodrik

Growth versus Poverty Reduction: A Meaningless Debate Should gover nments pursue econo mic growth first and forem o s t , or should they focus on po ver ty red u c t i o n? Recent debate on this question has become embroiled in broader polit- ic al cont r oversies on globalizat i o n and its impact on developing econom i e s . Cr itics of the WTO often take it to task for being over ly conc e r ned about the level of econo mic activity (and its growth) at the expense of pover ty red u c t i on . Sup p o r ters argue that expanded trad e and higher econo mic growth are the best ways to reduce pover ty. This largely steril e debate merel y diver ts attention from the real issues. In prac t i c e , ec on o mic growth and po ver ty red u c t i o n do tend to correlate ver y clo s e l y. How e ver , the real question is (or ought to be) whether open trade policies are a reliable mechanism for generating self-sustaining gr owth and po ver ty red u c t i on , the evidence for which is far less convi n c i n g . Re g a r ding the rel a t i o nship between growth and pover ty red u c t i on , le t ’s take some The real question is (or of the easier questions . Does growth benefit the poor? Yes , in general . The absolute number of people living in pover ty has dropped in all of the developing countries that ought to be) whether ha v e sustained rapid growth over the past few decad e s . In theory, a country could open trade policies are a en j o y a high aver age growth rate without any benefit to its poorest households, if in c o me disparities grew significan t l y—that is, if the ric h got ric her while the income s reliable mechanism for of the poor stagnated or decli n e d . This is unlikely, how e ver ; in c o me distrib u t i o n tends to be stable over time, and rare l y changes so muc h that the poor would experience an generating self- absolute decline in incomes while aver age incomes grow in a sustained fashion. sustaining growth and Mo re o ver , to the extent that income distrib u t i o n cha n g e s , its rel a t i o nship to econo mic growth varies from country to country. Gr owth has been accompanied by grea t e r poverty reduction.... eq u a l i t y of income in the Taiwan Province of China, Bangladesh and Egypt , for exam- pl e , but by greater inequality in Chile, China and Pol a n d . This suggests that the mag- nitude of the pover ty- re d u c t i o n payoff from growth depends, in part, on a country’s specific circumstances and policies. Is pover ty red u c t i o n good for growth? Again, yes , in general . It is hard to think of co u n t r ies where a large decrease in the absolute number of people living in pover ty has not been accompanied by faster growth . Just as we can imagine growth occurrin g without any red u c t i o n of pover ty, we can also imagine a strat e g y of pover ty red u c t i o n that relies exclu s i ve l y on red i s t r ibuting wealth from the ric h and the middle classes to the poor. In prin c i p l e , a country pursuing red i s t ri b u t i v e policies could reduce pover ty even if its total income did not grow. But we would be hard- p r essed to find rea l - w o rl d ex a m p l e s . Policies that increase the incomes of the poor, su c h as investments in prim a r y ed u ca t i on , rur al infras t ru c t u r e, health and nutrit i on , tend to enhance the prod u c t i v e cap a c i t y of the whole econom y , boosting the incomes of all grou p s . What does a high correl a t i o n between growth and the incomes of the poor tell us? Prac t i ca l ly nothing, for the rea s o ns outlined above. Al l it shows is that income distrib u - ti o n tends to be stable and fairly unres p on s i v e to policy cha n g e s . Mo re o ver , a strong co r re l a t i o n between econo mic growth and pover ty red u c t i o n is compatible with bo t h of the follo wing arguments: (1) onl y policies that target growth can reduce pover ty; an d (2) onl y policies that reduce pover ty can boost overa l l econo mic growt h .T h e re f o re, th e ob s e r ved correl a t i o n between growth and pover ty red u c t i o n is of little interest as far as po l i c y choices and prio r ities are conc e rn e d .

12 The Global Governance of Trade

A some what differ ent question is whether the well-being of the poor should enter as an independent determinant of policy cho i c e s , in addition to the usual focus on macroe c on o mic stability, mi c ro e c on o mic efficiency, and institutional quality. In other words , should econo mic ref o r m strategies have a pover ty focus? Yes , for at least three rea s on s . Fir s t , in cons i d e r ing social welfare, most people in ge n e ra l , and most democra t i ca lly elected gover nments in parti c u l a r , would give more weight to the well-being of the poor than to that of the rich . An econom y ’s growth rate is not a sufficient statistic for evaluating welfare because it ignores the distrib u - ti o n of the rewa r ds of growth . A policy that increases the income of the poor by one rupee can be worthwhile at the margin even if it costs the rest of society more than a rup e e . From this perspective, it may be entirel y rat i o nal and proper for a govern m e n t con s i d e r ing two competing growth strategies to choose the one that has grea t e r potential payoff for the poor even if its impact on overa l l growth is less assured . Sec on d , even if the welfare of the poor does not rec e i v e extra weight, in t e r ven t i on s aimed at helping the poor may still be the most effec t i v e way to raise aver age income s . Pover ty is natural ly associated with market imperfec t i o ns and incomp l e t e n e s s . The poor remain poor because they cannot borrow against future earnings to invest in educat i on , s k i ll s ,n ew crops and entrep re n e u r ial activities. Th e y are cut off from econo mic activity be c ause they are depriv ed of many colle c t i v e goods (e.g., pro p e r ty rig h t s , public safet y, in f ra s t ru c t u r e) and lack informa t i o n about market opportu n i t i e s . It is a standard tenet of econo mic theory that raising real aver age incomes req u i r es interven t i o ns targeted at closing gaps between private and social costs. Th e r e will be a prep on d e r ance of such op p o r tunities where there is a prep on d e r ance of pover ty. Th i rd , focusing on pover ty is also warranted from the perspective of an approa c h to devel o p ment that goes beyond an exclu s i v e focus on cons u m p t i o n or income level s to embrace human cap a b i l i t i e s . As Ama r tya Sen (1999) has empha s i ze d , the overa r - ching goal of devel o p ment is to maximize people’s capabilities—that is, their ability to lead the kind of life they value. The poor face the greatest hurdles in this area and ar e theref o r e the most deserving of urgent policy attention. Pol i c y-makers make choices and determine prio r ities all the time. The lens th r ough which they perce i v e devel o p ment prof o u n d ly affects their cho i c e s . Ke e p i n g po ver ty in sight ensures that their prio r ities are not distort e d .C onsider some illu s t r a- ti v e trad e o f fs . • Fis c al policy. Ho w should a gover nment res o l ve the trade-off between higher spending on pover ty- r elated projects (rur al infras t ru c t u r e, sa y) and the need for tight fiscal policies? Should it risk incurring the disapproval of financial markets as the price of better irrig a t i o n? How should it allo c ate its educat i o nal budget? Should more be spent on building prim a r y schools in rur al areas or on trai n i n g bank auditors and accountants? • Ma r ket liberal i za ti o n . Should the gover nment maintain price cont r ols on food cro p s , even if such cont r ols distort res o u r ce allo ca t i o n in the econo my? Should it rem o ve capital cont r ols on the balance of payme n t s , even if that means fiscal res o u r ces will be tied up in holding additional foreign res e r ves — re s o u r ces that could otherwise have been used to finance a social fund?

13 Dani Rodrik

• In s ti tu t ional refo rm . Ho w should the gover nment design its anti-corrup t i o n st ra t e g y? Should it target the large-scale corrup t i o n that foreign investors complain about or the petty corrup t i o n in the police and judicial systems that affec t s ord i n a r y citizens? Should legal ref o r m focus on trade and foreign investment or do mestic problems? Whose prop e r ty rights should rec e i v e pri o ri ty, peasants or fo r eign patent holders? Should the gover nment pursue land re f o rm , even if it th r eatens political ly powe r ful grou p s ? As these examples illu s t ra t e , in prac t i c e , even the standard, gr owth - o r iented desiderat a of macroe c on o mic stability, mi c ro e c on o mic efficiency and institutional ref o r m leave con- si d e r able roo m for manoeuvre .G over nments can use this roo m to better or worse effec t . A pover ty focus helps ensure that the rel e vant trad e - o f f s are cons i d e r ed explicitly. Since growth and pover ty red u c t i o n go largely hand in hand, the real questions are : What are the policies that yield these rewa r ds? How muc h do we know about po l i c y impacts? The honest answer is that we do not know nearly enough. We have evidence that land ref o rm s , ap p ro p ri a t e l y targeted price ref o r ms and certain types of health and educat i o n expenditures benefit the poor, but we are uncertain about many th i n g s . It is one thing to say that devel o p ment strategies should have a pover ty focus, another to identify the rel e vant policies. But this is not a strike against pover ty- o r iented prog ra m m e s , since we are equally un c e r tain about growth - o r iented prog ra m m e s . The uncomf o r table rea l i t y is that our kn o wledge about the kinds of policies that stimulate growth remains limited. We kn o w that large fiscal and macroe c on o mic imbalances are bad for growth . We know that ‘goo d ’ in s t i t u t i o ns are importa n t , even though we have ver y little idea about how co u n t r ies can acquire them. And , despite a voluminous literat u r e on the subject, we kn o w next to nothing about the kinds of trade policies that are most cond u c i v e to gr owth (see below) . For all of these rea s on s , it is not prod u c t i v e to make a sharp distinction betwe e n policies that promote growth and those at target the poor direc t l y. These policies are li k e l y to vary cons i d e ra b l y depending on institutional cont e x t , making it difficult to ge n e ra l i z e with any degree of prec i s i on . Our real focus should be on what works , ho w, and under what circu m s t a n c e s .

Achieving Economic Growth: What Really Matters? The enlightened standard view of devel o p ment policy grew out of dissatisfaction with the limited results yielded by the Was h i n g t o n Consensus policies of the 1980s and 19 9 0 s . The disappointing growth perfo r mance and increasing econo mic insecurit y in Latin Ame ri c a—the reg i o n that went furthest with policies of priv a t i za t i on , li b e ra l i z a- ti o n and openness—the failures in the former Soviet Union, and the Asian financial cr isis of 1997-98 all cont r ibuted to a ref a s h i on i n g , resulting in the ‘au g m e n t e d Was h i n g t o n Cons e n s u s ’ (s h o wn in Table 1). This goes beyond liberal i za t i o n and priv a - ti za t i o n to empha s i z e the need to create the institutional underpinnings of marke t ec on om i e s . Re f o r ms now include financial reg u l a t i o n and prudential supervi s i on , gover nance and anti-corrup t i on , legal and administrat i v e ref o rm , la b o u r - m a r ket ‘f l e x i b i l i t y’ and social safet y nets.

14 The Global Governance of Trade

T a b l e 1

The Original The Augmented Washington Consensus Washington Consensus

Fiscal discipline The original list plus: Reorientation of public Legal/political reform expenditures Regulatory institutions Tax reform Anti-corruption Financial liberalization Labour market flexibility Unified and competitive WTO agreements exchange rates Financial codes and standards Trade liberalization ‘Prudent’ capital-account opening Openness to DFI Non-intermediate exchange rate Privatization regimes Deregulation Social safety nets Secure property rights Poverty reduction

Op e ra t i on a l ly, these institutional ref o r ms are heavily influenced by an Ang l o - Ame ri c an conc e p t i o n of what constitutes desirable institutions (as in the prefe r ence for ar ms-length finance over ‘de vel o p ment banking’ and flexible labour markets over institu t i on a l i z ed labour marke t s ) . In addition, th e y are driv en largely by the req u i re m e n t s of integrat i o n into the world econom y : hence the emphasis on the interna t i o nal harmon i za t i o n of reg u l a t o r y prac t i c e s , as in the case of financial codes and standards and of the WTO agree m e n t s . Ma r ket econo mies rel y on a wide array of non- m a r ket institutions that perfo r m reg u l a t o r y, st a b i l i z i n g , and legitimizing functions (see Rodrik 2001a). Cro s s - n a t i on a l ec on om e t r ic work shows that the quality of a country’s public institutions is a crit i ca l , and perhaps the most importa n t , de t e r minant of a country’s long - t e r m devel o pm e n t (A cemoglu et al. 20 0 0 ) . While the recent emphasis on institutions is thus highly wel- com e , it needs to be borne in mind that the institutional basis for a market econo my is not uniquely determi n e d . Th e r e is no single mapping between a well- f u n c t i o ning market and the fo r m of non- m a r ket institutions req u i r ed to sustain it, as is clear from the wide varie t y of reg u l a t o r y, stabilizing and legitimizing institutions in today’s advanced in d u s t r ial societies. The Am e ri can style of capitalism is ver y differ ent from the Japanese style of cap i t a l i s m . Both differ from the European style . And even within Eu ro p e , th e r e are large differ ences between the institutional arrangements in, sa y, Sweden and Germa n y. Over the long term, ea c h of these have perfo r med equally well. 2 The point about institutional diver s i t y has in fact a more fundamental implicat i on . As Roberto Unger (1998) argues, the institutional arrangements in operat i o n today, va r ied as they are, th e m s e l ves constitute a subset of the full range of potential institu- ti o nal possibilities. Th e r e is no rea s o n to suppose that modern societies have exhausted

15 Dani Rodrik

al l useful institutional varia t i o ns that could underpin healthy and vibrant econom i e s . We must avoid thinking that a specific type of institution—mode of corpo r ate gover - na n c e , social securit y system or labour market legislation, for example—is the onl y one co mpatible with a well- f u n c t i o ning market econom y . Leaving aside the question of long - t e r m choice over institutional forms , th e enlightened standard view, insofar as it is presented as a recipe for stimulating econo m- ic growth , also suffers from a fatal flaw: it provides no sense of prio r ities among a long and highly demanding list of institutional prere q u i s i t e s . This kitchen-sink approa c h to de vel o p ment strat e g y flies in the face of prac t i c al rea l i t y and is at odds with the histori- cal experience of today’s advanced industrial econom i e s . What are today reg a r ded as ke y institutional ref o r ms in areas such as corpo r ate govern a n c e , financial supervi s i on , tr ade law and social safet y nets did not take place in Europe or No rt h e rn Ame ri c a until quite late in the econo mic devel o p ment process (Chang 2000). In d e e d , ma n y of the items on the augmented Was h i n g t o n Consensus agenda (Table 1) should be prop e r ly vi e wed as outcomes of successful econo mic devel o p ment rather than its prere q u i s i t e s . The rea l i t y of growth tran s f o rm a t i o ns is that they are instigated by an initially narrow set of policy and institutional initiatives , wh i c h might be cal led ‘in vestment strat e - gi e s ’ (R o d r ik 1999). Adequate human res o u rc e s , public infras t ru c t u r e, social peace and st a b i l i t y are all key enabling elements of an investment strat e g y. But often the crit i ca l factor is a set of targeted policy interven t i o ns that kindle the animal spirits of dome s t i c in ves t o r s . These investment strategies set off a period of econo mic growt h ,w h i ch in tu r n facilitates a cycle of institutional devel o p ment and further growth . The initiating ref o r ms are rare l y rep l i c as of each other, and they bear onl y partial resemblance to the req u i r ements highlighted by the enlightened standard view. Typi ca l ly, th e y entail a mix of orth o d o xy with unconven t i o nal domestic innova t i on s . An analysis of three sets of investment strategies will elucidate this central point and highlight the differ ent paths taken to greater pros p e ri t y: im p o rt - s u b s t i t u t i on , Ea s t - As i a n - s t yle outwa r d orie n t a t i o n and two - t ra c k ref o r m strat e g i e s . The list is not meant to be exhaustive, and in the future successful strategies are likely to differ from all three .

Im p o r t-Substituting Industrialization (ISI) Im p o r t-substituting industria l i za t i o n is based on the idea that domestic investment and te ch n o l o g i c al capabilities can be spurred by providing home producers with (temporar y) prot e c t i o n against imports . Although this approa c h has fallen into disgrace since the 1980s, it actually did quite well for a substantial period of time in scores of devel - oping nations . Until the first oil shock hit in 1973, no fewer than 42 developing coun- tr ies grew at rates exceeding 2.5 per cent per capita per annum (see Rodrik 1999: ch. 4 ) . At this rat e , in c o mes would double ever y 28 years or less. Most of these countries follo wed ISI policies. The list includes 12 countries in South Ame ri ca , six in the Middle East and Nor th Africa , and 15 in Sub - Sa h a r an Afri ca . In fact, th e r e were no less than six Sub - Sa h a r an Afric an countries among the 20 fastest-growing developing countrie s in the world prior to 1973: Swa z i l a n d , Bo t s w a n a , Côte d’I v o i r e, Les o t h o , Ga b o n and Tog o, with Kenya ranking 21st. Th e r e can be little doubt that econo mic growth led to substantial improvements in the living cond i t i o ns of the vast majorit y of the house-

16 The Global Governance of Trade

holds in these countrie s . Be t ween 1967 and 1977, li f e expectancy at birth increased by four years in Brazil (from 58 to 62), by five years in Cote d’I v o i r e (from 43 to 48), by fi v e years in Mexico (from 60 to 65), and by five years in Pakistan (from 48 to 53). In Ke ny a , infant morta l i t y fel l from 112 (per 1,000 live births) in 1965 to 72 in 1980. ISI policies spurred growth by creating protected and theref o r e profitable home ma r kets for domestic entrep r eneurs to invest in. Con t ra r y to rec e i v ed wisdom, IS I - d ri v - en growth did not produce techn o l o g i c al lags and inefficiency on an econom y - w i d e sca l e . In fact, the prod u c t i v i t y perfo r mance of many Latin Ame ri c an and Middle Ea s t e r n countries was, in comp a ra t i v e perspective, exe m p l a r y. Acc o r ding to estimates pr oduced by Collins and Bosworth (1996), not onl y was aver age total factor prod u c t i v i t y (TFP) growth during the period preceding the first oil shock quite high in the Middle East and Latin Ame ri c a (at 2.3 and 1.8%, res p e c t i ve ly ) , it was actually significan t l y higher than in East Asia (1.3%)! Countries such as Braz i l , the Domi n i c an Republic and Hence, as an industrial- Ecuador in Latin Ame ri ca ; Iran , Mo r occo and Tunisia in the Middle East; and Côte ization strategy intended d’I v o i r e and Kenya in Afric a all experienced more rapid TFP growth than any of the East Asian countries in this early period (with the possible exce p t i o n of Hong Kong , fo r to raise domestic invest- wh i c h comp a r able data are not available). Me x i c o , Bo l i v i a , Pan a m a , Eg ypt , Al g e ri a , Tan z ania and Zai r e experienced higher TFP growth than all but Tai w a n . Of course, no t ment and enhance al l countries follo wing ISI policies did well: Argentina is a striking counter-example, productivity, import with an aver age TFP growth of onl y 0.2 per cent from 1960 to 1973. The dismal rep u t a t i o n of ISI is due partl y to the subsequent econo mic colla p s e substitution apparently ex p e r ienced by many of the countries pursuing it in the 1980s, and partl y to the ex t re m e l y influential studies of Little, Scott and Scitovsky (1970) and Bela Balassa worked pretty well in a (1 9 7 1 ) . What these two studies did was to document in detail some of the st atic ec o - very broad range of no mic inefficiencies generated by high and extrem e l y dispersed effec t i v e rates of pro- te c t i o n (ERP) in the manufacturing sectors of the countries under study. The discover y countries until at least of cases of negative value-added at world prices—that is, cases where countries would ha v e been better off by throwing away the inputs than by processing them as they did the mid-1970s.As an in highly protected plants—was parti c u l a r ly shocki n g . How e ver , neither study cla i m e d industrialization strategy to show that countries which had follo wed ‘ou tw a r d orie n t e d ’ st r ategies had been sys- te m a t i ca l ly immune from the same kind of inefficiencies. In fact, their evidence can be intended to raise domes- read as suggesting that there was no such clear dividing line.3 In addition, the evi d e n c e on TFP growth revi e wed above shows that the idea that ISI produced more dyna m i c tic investment and in e f f i c i e n c y than did ‘ou tw a r d orie n t a t i o n’ is simply incorrec t . enhance He n c e , as an industria l i za t i o n strat e g y intended to raise domestic investment and enhance prod u c t i v i t y, im p o r t substitution apparen t l y worked pret t y well in a ver y broa d productivity, import range of countries until at least the mid-1970s. How e ver , st a r ting in the second half of the 1970s, a disaster befel l the vast majorit y of the econo mies that had been doing well. Of the 42 countries with growth rates above 2.5 per cent prior to 1973, less than a th i r d (12) managed the same rec o r d over the next decad e . The Middle East and Lat i n Ame ri ca , wh i c h had led the developing world in TFP growth prior to 1973, not onl y fel l behind, but actually began to experience ne gat i ve TFP growth on avera g e . On l y East Asia held its own , while South Asia actually improved its perfo r mance (see Co l lins and Bosworth 1996).

17 Dani Rodrik

Was this a result of the ‘ex h a u s t i o n’ of import- s u b s t i t u t i o n policies? As I have argued elsewh e r e (Rodrik 1999), the comm o n timing implicates the turbulence experienced in the world econo my follo wing 1973—the abandonment of the Bret t o n Woo d s sy stem of fixed exchange rat e s , two major oil shocks , va r ious other comm o d i t y boom- and-bust cycle s , plus the U.S .Fed e r al Reserve interes t - r ate shock of the early 1980s. The fact that some of the most ardent follo wers of ISI policies in South Asia—espe- ci a l ly India and Pakistan — managed to either hold on to their growth rates after 1973 (P akistan) or increase them (India) also suggests that more than just ISI was invo l ved . 4 The actual story implicates macroe c on o mic policies rather than the trade reg i m e . The proximate rea s o n for the econo mic collapse was the inability to adjust macroe c o - no mic policies approp ri a t e l y in the wake of these external shocks . Ma c ro e c on o mic mal- adjustment gave rise to a range of synd r omes associated with macroe c on o mic instabili- The real determinants of ty—high or rep r essed inflation, sca rc i t y of foreign exchange and large black- m a r ket pre m i u m s , ex t e r nal payments imbalances and debt cris e s — w h i c h grea t l y magnified the growth performance real costs of the shocks . Co u n t r ies that suffer ed the most were those with the largest after the 1970s ar e in c r eases in inflation and black- m a r ket premiums for foreign curren c y. The culprit s we r e poor mone t a r y and fiscal policies and inadequate adjustments in excha n g e - ra t e rooted in the ability of po l i c y, so metimes aggravated by shortsighted policies of creditors and the Bret t o n Woods institutions . The bottom line is that in those countries that experienced a debt domestic institutions to cri s i s , the crisis was the product of mone t a r y and fiscal policies that were incomp a t i b l e manage the distributional with sustainable external balances: th e r e was too little expenditure reducing and expen- di t u r e switchi n g . Trade and industrial policies had ver y little to do with bringing on the conflicts triggered by cri s i s . Why were some countries quicker to adjust their macroe c on o mic policies than the external shocks of others? The real determinants of growth perfo r mance after the 1970s are rooted in the the period. ab i l i t y of domestic institutions to manage the distrib u t i o nal conflicts trig g e r ed by the ex t e r nal shocks of the perio d . Social conflicts and their management—whether successful or not—played a key role in transmitting the effects of external shocks on to ec on o mic perfo rm a n c e . Societies with deep social cleavages and poor institutions of co nflict management proved worse at handling shocks (see Rodrik 1999).

‘O u t w a r d-Oriented’ Industrialization The experience of the East Asian tigers is often presented as one of export-led growth , in which opening up to the world econo my unleashed powe r ful forces of industria l di ve r s i f i ca t i o n and techn o l o g i c al cat ch - u p . How e ver , the conven t i o nal account over - lo o ks the active role taken by the gover nments of Taiwan Province of China and the Republic of Korea (and Japan before them) in shaping the allo ca t i o n of res o u rc e s . In neither of these countries was there significant import liberal i za t i o n early in the proc e s s of growth . Most of their trade liberal i za t i o n took place in the 1980s, when high gr owth was already firml y established. The key to these and other East Asian countrie s ’ success was a cohe r ent strat e g y of raising the ret u r n to private inves t m e n t , th r ough a range of policies that included cred i t subsidies and tax incentives , ed u ca t i o nal policies, establishment of public enterpri s e s , ex p o r t inducements, du ty - f r ee access to inputs and capital goods and actual gover n-

18 The Global Governance of Trade

ment coordi n a t i o n of investment plans. In the Republic of Korea , the chief form of in vestment subsidy was the extension of credit to large business groups at negative rea l in t e r est rat e s . Ko r ean banks were nationa l i z ed after the military coup of 1961, an d con s e q u e n t l y the gover nment obtained exclu s i v e cont r ol over the allo ca t i o n of in vestible funds in the econom y . Another important manner in which investment was su b s i d i z ed in Korea was through the socializat i o n of investment risk in selected sectors. This emerged because the gover nment—most notably President Park — p r ovided an implicit guarantee that the state would bail out entrep r eneurs investing in ‘de s i ra b l e ’ activities if circumstances later threatened the prof i t a b i l i t y of those inves t m e n t s . In Tai w a n , in vestment subsidies took the form of tax incentives . In both the Republic of Korea and Tai w a n , public enterpr ises played a ver y important role in enhancing the prof i t a b i l i t y of private investment by ensuring that key inputs were available local ly for private producers down s t re a m . Not onl y did public In both the Republic of en t e rp r ises account for a large share of manufacturing output and investment in each Korea and Taiwan, public co u n t r y, their importance actually increased during the crit i c al take-off years of the 19 6 0 s . Sin g a p o r e also heavily subsidized inves t m e n t , but this country differs from the en t e r p r i s e s played a very Republic of Korea and Taiwan in that its investment incentives centred heavily on foreign inves t o r s . important role in enhanc- While trade policies that spurred exports were part of this complex arsenal of ing the profitability of in c e n t i ve s , in vestment and its promo t i o n was the key goal in all countrie s . To that end, gover nments in the Republic of Korea and Taiwan free l y res o r ted to unorth o d o x strat e - private investment by gi e s : th e y protected the home markets to raise prof i t s , implemented generous export su b s i d i e s , en c o u r aged their firms to reverse-engineer foreign patented prod u c t s , an d ensuring that key inputs imposed perfo r mance req u i r ements such as export- i m p o r t balance req u i r ements and were available locally for do mestic content req u i r ements on foreign investors (when foreign companies were all o wed in). Al l of these strategies are now severe l y res t r icted under the WTO agree - private producers me n t s . downstream. The Two - T rack Strategy A rel a t i ve l y minimal set of ref o r ms in China in the late 1970s set the stage for the phe - no menal econo mic perfo r mance that has been the envy of any poor country since. In i t i a l ref o r ms were rel a t i ve l y simple: th e y loosened the comm unal farming system and allow e d fa r mers to sell their crops in free markets once they had fulfilled their quota obligations to the state. Subsequent ref o r ms allo wed the crea t i o n of township and village enterpri s e s and the extension of the ‘ma r ket track ’ into the urban and industrial sectors. Special econo mic zones were created to attract foreign inves t m e n t . What stands out about these ref o r ms is that they are based on dual trac ks (state and marke t ) , on gradualism and on ex p e ri m e n t a t i on . One can interpr et Chinese-style gradualism in two ways. One perspective, rep r e- sented force f u l ly in work by Sac hs and Woo (2000) underpl a ys the rel e vance of Chinese parti c u l a r ism by arguing that the successes of the econo my are not due to any special aspects of the Chinese tran s i t i o n to a market econom y , but instead are largely due to a co n v erge n c e of Chinese institutions with those in non-socialist econom i e s . In this view, the faster the conver g e n c e , the better the outcome s : ‘f a v o r able outcomes have

19 Dani Rodrik

emerged not because of grad u a l i s m , but de s p i t e gra d u a l i s m ’ (i b i d : 3 ) . The policy message that follo ws is that countries that look to China for lessons should focus not on institutional experim e n t a t i o n but on harmo nizing their institutions with those ab ro a d . The alterna t i v e perspective, pe r haps best developed in work by Qian and Ro l a n d , is that the peculiarities of the Chinese model rep r esent solutions to parti c u - lar political or informa t i o nal problems for which no blueprin t - s t yle solution exists. Hence Lau , Qian and Roland (1997) interpr et the dual-trac k approa c h to liberal i z a- ti o n as a way of implementing Par eto-efficient ref o rm s : an alterat i o n in the planned ec on o my that improves incentives at the margin, enhances efficiency in res o u r ce allo - ca t i on , and yet leaves none of the plan beneficiaries worse off. Qi a n , Roland and Xu (1999) interpr et Chinese-style decentral i za t i o n as allo wing the devel o p ment of superior institutions of coordi n a t i on : when econo mic activity req u i r es products with ma t c hed attrib u t e s , lo c al experim e n t a t i o n is a more effec t i v e way of processing and using local knowl e d g e . These analysts find muc h to praise in the Chinese model be c ause they think the system generates the right incentives for developing the tacit kn o wledge req u i r ed to build and sustain a market econom y , and theref o r e they are not over ly bothered by some of the econo mic inefficiencies that may be generat e d al o ng the way. A less well- k n o wn instance of a successful two - t ra c k strat e g y is that of Ma u ri t i u s , wh e r e superior econo mic perfo r mance has been built on a peculiar comb i - na t i o n of orth o d o x and heterod o x strat e g i e s . An export processing zone (EPZ ) , op e r ating under free - t r ade prin c i p l e s , enabled an export boom in garments to Eu r opean markets and an accomp a n ying investment boom at home . Yet the island’s ec on o my has combined the EPZ with a domestic sector that was highly prot e c t e d until the mid-1980s: the IMF gave the Mauritian econo my the highest (i.e., wo r s t ) sc o r e on its ‘po l i c y res t ri c t i ve n e s s ’ index for the early 1990s, reck o ning it was one of the world most protected econo mies even by the late 1990s (see Sub r amanian 2001). Ma u r itius is essentially an example of an econo my that has follo wed a two - t ra c k st ra t e g y not too dissimilar to that follo wed by China, but which was underpinned by social and political arrangements that encouraged parti c i p a t i on , rep re s e n t a t i o n and co a l i t i on - b u i l d i n g . The circumstances under which the Mauritian EPZ was set up in 1970 are in s t ru c t i v e, and highlight the manner in which parti c i p a t o r y political systems help design crea t i v e strategies for building local ly adapted institutions . Gi v en the small si z e of the home marke t , it was evident that Mauritius would benefit from an outwa rd - o r iented strat e g y. But as in other developing countrie s , po l i c y-makers had to co ntend with the import-substituting industrialists who had been propped up by the res t ri c t i v e comm e r cial policies of the early 1960s prior to independence, and who we r e natural ly opposed to relaxing the trade reg i m e . A Was h i n g t o n econo mist would have advocated acros s - t h e - b o a r d liberal i za t i on , without reg a r d to what that might do the pre ca rious ethnic and political balance of the island. The EPZ scheme provided a neat way around the political difficulties. The crea t i o n of the EPZ generated new opportunities of trade and of employme n t ,

20 The Global Governance of Trade

without taking prot e c t i o n away from the import-substituting groups and from the male wo r kers who dominated the established industrie s . The segmentation of labour markets early on between male and female workers—with the latter pred om i n a n t l y em p l o yed in the EPZ—was parti c u l a r ly cruc i a l , as it prevented the expansion of the EPZ from driving wages up in the rest of the econom y , th e r eby disadvantaging import- substituting industrie s . New employment and profit opportunities were created at the ma r g i n , while leaving old opportunities undisturbe d . This in turn paved the way for the mo r e substantial liberal i za t i o ns that took place in the mid-1980s and in the 1990s. By the 1990s, the female-male earning ratio was higher in the EPZ than in the rest of the ec on o my (ILO 2001, table 28). Ma u r itius found its own way to econo mic devel o p - ment because it was able to devise a strat e g y that was unorth o d ox , yet effec t i ve .

The Bottom Line Most of the institutional These examples suggest that while market incentives , ma c ro e c on o mic stability and development occurs sound institutions are crit i c al to econo mic devel o pm e n t , th e y can be generated in a number of differ ent ways—by making the best use of existing capabilities in light of alongside economic res o u r ce and other cons t ra i n t s . Th e r e is no single model of a successful tran s i t i o n to a hi g h - g r owth path. Ea c h country has to figure out its own investment strat e g y. On c e development, not as a the approp r iate strat e g y is identified (or stumbled upon) , the institutional ref o rm s prerequisite to it. needed may not be extensive. Most of the institutional devel o p ment occurs along s i d e ec on o mic devel o pm e n t , not as a prer equisite to it.

Trade Liberalization, Growth and Poverty Reduction: What Do the Facts Really Show? Co nsider two countrie s , A and B. Co u n t r y A engages in state trad i n g , maintains import mon o p o l i e s , retains quantitative res t ri c t i o ns and high tarif f s (in the range of 30-50 per- cent) on imports of agric u l t u r al and industrial products and is not a member of the WTO. Co u n t r y B, a WTO member, has slashed import tarif f s to a maximum of 15 per cent and rem o ved all quantitative res t ri c t i on s , ea r ning a rar e comm e n d a t i o n from the U.S . State Department that ‘t h e r e are few significant barriers to U.S . ex p o rt s ’ (US Sta t e De p a r tment 1999). One of the two econo mies has experienced GDP growth rates in ex cess of 8 per cent per annum, has sharpl y reduced pover ty, has expanded trade at double-digit rat e s , and has attracted large amounts of foreign inves t m e n t . The other ec on o my has stagnated and suffer ed deterio r ating social indicat o r s , and has made little pro g r ess in integrating with the world econo my as judged by trade and foreign inves t - ment flows . Co u n t r y A is Viet Nam , wh i c h since the mid-1980s has follo wed Chinese-style gr adualism and a two - t ra c k ref o r m prog ra m m e . Co u n t y B is Haiti. Viet Nam has been phe n om e n a l ly successful, ach i e ving not onl y high growth and pover ty red u c t i on , but also a rapid pace of integrat i o n into the world econo my despite high barriers to tra d e . Ha i t i ’s econo my has gone nowh e r e, even though the country underto o k a com- pre h e n s i v e trade liberal i za t i o n in 1994-95. The cont r asting experiences of these two countries highlight two important points. Fir s t , a leadership committed to devel o p ment and standing behind a cohe r ent growth

21 Dani Rodrik

st ra t e g y counts for a lot more than trade liberal i za t i on , even when the strat e g y departs s h a rp ly from the enlightened standard view on ref o rm . Sec on d , in t e g ra t i o n with the wo r ld econo my is an outcome , not a prere q u i s i t e , of a successful growth strat e g y. Protected Viet Nam is integrating with the world econo my significan t l y more rap i d ly than is open Haiti, be c ause Viet Nam is growing and Haiti is not. This comp a ri s o n illu s t r ates a comm o n misdiagnosis. A typi c al Wor ld Bank exerc i s e co nsists of cla s s i f y ing developing countries into ‘gl o b a l i ze r s ’ and ‘non - g l o b a l i ze r s ’ based on their rates of growth of trade volumes. The analyst asks whether globalizers (i.e., those with the highest rates of trade growth) have experienced faster income growt h ,g reater pover ty red u c t i o n and worsened income distrib u t i o n (see Dollar and Kraa y 2000).The answers tends to be yes , yes , and no. As the Viet Nam and Haiti cases show, how e ver , this is a highly mis- leading exerc i s e . Trade volumes are the outcome of many differ ent things, in c luding most Integration with the im p o rt a n t l y an econom y ’s overa l l perfo rm a n c e . Th e y are not something that govern m e n t s con t r ol direc t l y. What gover nments cont r ol are trade po l i c i e s : the level of tariff and no-tarif f world economy is an ba r ri e r s , membership in the WTO, co mpliance with its agreements and so on. The rel ev a n t outcome, not a prerequi- qu e s t i o n is: Do open trade po l i c i e s rel i a b l y produce higher econo mic growth and grea t e r po ver ty red u c t i o n? site, of a successful Cro s s - n a t i o nal comp a ri s o n of the literat u r e reveals no systematic rel a t i on s h i p be t ween a country’s aver age level of tariff and non- t a r iff res t ri c t i o ns and its subsequent growth strategy. ec on o mic growth rat e . If anyth i n g , the evidence for the 1990s indicates a po s i t i ve (b u t st a t i s t i ca l ly insignificant) rel a t i o nship between tarif f s and econo mic growth (see Fig u r e 1) . The onl y systematic rel a t i o nship is that countries dismantle trade res t ri c t i o ns as they get rich e r . That accounts for the fact that today’s ric h countrie s , with few exce p t i on s , em b a r ked on modern econo mic growth behind prot e c t i v e barrie r s , but now have low tr ade barrie r s . The absence of a robust positive rel a t i o nship between open trade policies and econo mic growth may come as a surpr ise in view of the ubiquitous claim that trade liberal - iza t i o n promotes higher growth . In d e e d , the literat u r e is replete with cros s - n a t i on a l studies conc luding that growth and econo mic dynamism are strong l y linked to more li b e r al trade policies. For example, an influential study by Sac hs and War ner (1995) found that econo mies that are open, by their definition, gr ew 2.4 percentage points faster annually than did those that are not—an enormous differe n c e . Without such st u d i e s , or g a n i za t i o ns such as the Wor ld Bank, IMF and the WTO could not have been so vocifer ous in their promo t i o n of trad e - c e n t r ic devel o p ment strat e g i e s . Up o n closer look ,h ow ever , these studies turn out to be flawed. The cla s s i f i ca t i o n of co u n t r ies as ‘op e n ’ or ‘clo s e d ’ in the Sach s - War ner study, for example, is not based on actual tr ade policies but largely on indicators related to exchange rate policy and locat i o n in Sub - Sah a r an Africa . Their cla s s i f i ca t i o n of countries in effect conflates macroe c on om i c s ,g e o g - rap hy and institutions with trade policy. It is so correlated with plausible groupings of al t e rn a t i v e explanatory varia b l e s — m a c ro e c on o mic instability, poor institutions , lo ca t i o n in Af ri c a—that one cannot draw from the subsequent empiric al analysis any strong infer - ences about the effects of openness on growth (see Rodrig u e z and Rodrik 2001). The problem is a general one . In a revi e w of the best-known literat u r e (Dolla r 19 9 2 ; Ben-David 1993; Ed w a r ds 1998; Frankel and Romer 1999; Sac hs and Warn e r

22 The Global Governance of Trade

F i g u r e 1 Low import tariffs are good for growth? Think again

Source: All data are averages for the 1990s,and come from the Dollar and Kraay (2000) data set. Specifications are based on Dollar and Kraay (2000), replacing trade/GDP with tariff levels and control- ling separately for initial income, government consumption/GDP and inflation rate.

19 9 5 ) , Francisco Rodrig u e z and I found a major gap between the policy concl u s i on s that are typi ca l ly drawn and what the res e a r ch has actually shown . A comm o n prob - lem has been the misattrib u t i o n of macroe c on o mic phe n o mena (e.g., over valued cur- rencies or macroe c on o mic instability) or geograp hic locat i o n (e.g., in the trop i ca l zone) to trade policies. Once these problems are correc t e d , an y meaningful rel a t i o n- ship across countries between the level of trade barriers and econo mic growth eva p o - rates (see also Helleiner 1994). In prac t i c e , the rel a t i o nship between trade openness and growth is likely to be a co ntingent one , dependent on a host of internal and external cha ra c t e ri s t i c s . The fact that prac t i ca l ly all of today’s advanced countries embarked on their growth behind ta r iff barrie r s , and reduced prot e c t i o n onl y subsequently, su re l y offers a clue of sorts . Mo re o ver , the modern theory of endogenous growth yields an ambiguous answer to the question of whether trade liberal i za t i o n promotes growth , one that depends on whether the forces of comp a ra t i v e advantage push the econom y ’s res o u r ces towa rd s activities that generate long - r un growth (res e a r ch and devel o pm e n t , expanding prod - uct varie t y, up g r ading product quality, etc.) or diver t them from such activities. No country has developed successfully by turning its back on interna t i o nal trad e and long - t e r m capital flows . Ver y few countries have grown over long periods of time without experiencing an increase in the share of foreign trade in their nationa l pro d u c t . In prac t i c e , the most comp e l ling mechanism that links trade with growth in de veloping countries is that imported capital goods are likely to be significan t l y cheaper than those manufactured at home . Policies that res t r ict imports of cap i t a l

23 Dani Rodrik

eq u i pm e n t , raise the price of capital goods at home and thereby reduce real inves t - ment levels have to be viewed as undesirable on the face of it – although this does not rule out the possibility of selective infant industry policies in certain segments of cap i t a l - g oods industrie s . Ex p o rt s , in turn, ar e important since they permit the pur- chase of imported capital equipme n t . But it is equally true that no country has developed simply by opening itself up to foreign trade and inves t m e n t . The tric k has been to combine the opportu n i t i e s of fe r ed by world markets with a domestic investment and institution-building st ra t e g y to stimulate the animal spirits of domestic entrep re n e u r s . Almost all of the outstanding cases—East Asia, Ch i n a , India since the early 1980s—invo l ve parti a l and gradual opening up to imports and foreign inves t m e n t . The experiences of China and India are parti c u l a r ly notewo rt h y, as they are two huge countries that have done extrem e l y well rec e n t l y, and are often cited as examples of what openness can achi e ve (see Ste r n 2000:3). The rea l i t y, once again, is more com- pl i ca t e d . In both India and China, the main trade ref o r ms took place about a decad e af t er the onset of higher growth . Mo re o ver , these countrie s ’ tr ade res t ri c t i o ns rem a i n am o ng the highest in the world . As noted brie f l y above, the increase in China’s gr owth started in the late 1970s with the introd u c t i o n of the household res p on s i b i l i t y sy stem in agric u l t u r e and of two-tier pric i n g . Trade liberal i za t i o n did not start in ea r nest until muc h later, du r ing the second half of the 1980s and especially during the 19 9 0 s , once the trend growth rate had already increased substantially. The case of India is shown in Fig u r e 2. As the figure makes cle a r , In d i a ’s tren d gr owth rate increased substantially in the early 1980s (a fact that stands out parti c u l a r ly cle a r ly when it is comp a r ed against other developing countrie s ) , while serious trad e ref o r m did not start until 1991-93. The tariff aver ages displayed in the cha r t show that ta ri f f s were actually higher in the rising growth period of the 1980s than in the low- gr owth 1970s. To be sure, ta ri f f s hardly constitute the most serious trade res t ri c t i o ns in In d i a , but they nonetheless display the trends in Indian trade policy fairly accurat e l y. Of course, both India and China did ‘pa r ticipate in interna t i o nal trad e, ’ and by that measure they are both globalizer s . But the rel e vant question for policy-makers is not whether trade per se is good or bad—countries that do well also increase their tr ade/GDP ratios as a by-product—but what the correct sequence of policies is and ho w muc h prio ri t y deep trade liberal i za t i o n should rec e i v e early in the ref o r m pro c e s s . With reg a r d to the latter questions , the experiences of India and China are su g g e s t i v e of the benefits of a grad u a l , sequenced approa ch . To rep e a t , the approp r iate concl u s i o n is not that trade prot e c t i o n is inheren t l y pre fe r able to trade liberal i za t i on ; ce rt a i n l y, th e r e is scant evidence from the last 50 years that inwa rd - l o o king econo mies experience syst e m a t i ca l ly faster econom i c gr owth than open one s . But the benefits of trade openness are now grea t l y over s o l d . Deep trade liberal i za t i o n cannot be relied on to deliver high rates of econom i c gr owth and theref o r e does not deserve the high prio ri t y it typi ca l ly rec e i v es in the de vel o p ment strategies pushed by leading mul t i l a t e r al organizat i on s . 5 As Helleiner (2000: 3) puts it, th e r e are ‘f ew reputable developing country analysts or gover nments who question the positive potential roles of interna t i o nal trade or

24 The Global Governance of Trade

F i g u r e 2 Tariffs and Growth in India

—l — Tariffs —u — Import Duty — — Gr owth Relative to All Developing Countries l

l u u The rules for admission u u into the world economy

not only reflect little

awareness of develop- Source: Author’s calculations from data in Dollar and Kraay (2000) and World Bank,World Development Indicators 2000, CD-Rom. ment priorities, they are capital inflow in econo mic growth and overa l l devel o pm e n t . Ho w co u l d th e y question often completely unrel a t e d the inevitable need for parti c i p a t i o n in, indeed a cons i d e r able degree of integrat i o n to sensible economic wi t h , the global econom y ? ’ The real debate is not over whether integrat i o n is good or ba d , but over matters of policy and prio ri t i e s : ‘It isn’t at all obvi o u s ei t h e r (1) that fur- principles. ther external liberal i za t i o n (‘openness’) is now in ever y country’s interest and in all di m e n s i o ns or (2) that in the overa r ching sweep of global econo mic history what the wo r ld now most req u i r es is a set of global rules that promote or ease the path to gr eater free d o m for global market actors, and are universal in applicat i o n’ (i b i d : 4) .

The Integrationist Agenda and the Crowding Out of Development Priorities Prio r ities are important because in the enlightened standard view, in s e rt i o n into the wo r ld econo my is no longer a matter of simply rem o ving trade and investment barri- er s . Co u n t r ies have to satisfy a long list of institutional req u i r ements in order to ma x i m i z e the gains and minimize the risks of parti c i p a t i o n in the world econom y . Global integrat i o n remains the key prer equisite for econo mic devel o pm e n t , but there is now a lot more to it than just throwing the borders open. Reaping the gains from openness req u i r es a full complement of institutional ref o rm s . So trade liberal i za t i o n entails not onl y the lowe r ing of tariff and non- t a r iff barri- er s , but also compliance with WTO req u i r ements on subsidies, in t e l lectual prop e r ty, cu s t o ms proc e d u re s , sa n i t a r y standards and policies vis-à-vis foreign inves t o r s . Mo re o ver , these legal req u i r ements have to be complemented with additional re f o rm s to ensure favourable econo mic outcome s : tax ref o r m to make up for lost tariff reven u e s ; social safet y nets to compensate displaced worke r s ; cre d i b i l i t y enhancing

25 Dani Rodrik

in s t i t u t i o nal innova t i o ns to quell doubts about the permanence of the ref o rm s ; la b o u r - m a r ket ref o r m to enhance labour mobility across industrie s ; te ch n o l o g i ca l assistance to upgrade firms adver s e l y affected by import comp e t i t i on ; tr aining pro- gr ammes to ensure that export- o r iented firms and investors have access to skille d wo rk e r s ; and so on. Reading Wor ld Bank rep o r ts on trade policy, one can be excu s e d for thinking that the list of comp l e m e n t a r y ref o r ms is virtu a l ly endle s s . Not withstanding the over ly Ang l o - Am e ri c an conc e p t i o n of institutional possibilities reflected in the Was h i n g t o n agenda for integrat i o nist ref o rm , ma n y of the pr oposed institutional ref o r ms are perfec t l y sensible one s , and in a world without fi n a n c i a l , ad m i n i s t ra t i v e or political cons t ra i n t s , th e r e would be little argument about the need to adopt them. But in the real world , fi s c al res o u rc e s , ad m i n i s t ra t i v e cap a - bilities and political capital are all scarc e , and choices need to be made about how to de p l o y them. In such a world , vi e wing institutional prio r ities from the vantage point of inserti o n in the global econo my has real opportu n i t y costs. Some trad e - o f f s are illu s t ra t i ve . It has been estimated that it costs a typi ca l de veloping country $150 milli o n to implement req u i r ements under just three of the WTO agree m e n t s : cu s t o ms valuation, sa n i t a r y and phyto s a n i t a r y measures (SPS) and intellectual prop e r ty rights (TRI P S ) . As the Wor ld Bank’s Micha el Fin g e r points out, this is a sum equal to a yea r ’s devel o p ment budget for many of the least- de veloped countries (Finger and Schuler 1999). In the area of legal ref o rm , should the gover nment focus its energies on ‘im p o r t- in g ’ legal codes and standards , or on improving existing domestic legal institutions ? In Turk e y, a weak coalition gover nment spent sever al months gathering political sup- po r t for a bill that would provide foreign investors the prot e c t i o n of interna t i on a l arb i t ra t i on . Wou l d n ’t it have been a better strat e g y for the long run to ref o r m the existing legal regime for the benefit of foreign an d do mestic investors alike? In public health, should the gover nment pursue tough policies on comp u l s o r y licensing and/or paral lel importa t i o n of basic medicines, even if that means run n i n g afoul of existing WTO rules? The United States has charged that Braz i l ’s highly successful treatment prog r amme for HIV/AIDS violates WTO rules because it allow s the gover nment to seek comp u l s o r y licensing when a foreign patent holder does not ‘wo rk ’ the patent local ly. In industrial strat e g y, should the gover nment simply open up and let the chi p s dr op wherever they might, or should it emulate East Asian experience of industria l policies through export subsidies, di r ected credit and selective prot e c t i on ? Ho w should the gover nment focus its anti-corrup t i o n strat e g y? Should it target the ‘g ra n d’ co r ru p t i o n that foreign investors complain about, or the petty corrup t i o n that affects the poor the most? Perh a p s , as prop o nents of permanent normal trad e rel a t i o ns with China argued in the recent U.S . Con g re s s i o nal debate, a govern m e n t that is forced to protect the rights of foreign investors becomes more inclined to protect the human rights of its own citizens too. But isn’t this at best a trick l e - d ow n st ra t e g y of institutional ref o r m? Sho u l d n ’t institutional ref o r m be targeted on the de s i r ed ends direc t l y—whether those ends are the rule of law, im p r oved observa n c e of human rights or reduced corrup t i o n?

26 The Global Governance of Trade

The rules for admission into the world econo my not onl y reflect little awareness of de vel o p ment prio ri t i e s , th e y are often comp l e t e l y unrelated to sensible econo mic prin - ci p l e s . WTO rules on anti-dumping, subsidies and countervailing measures , ag ri c u l - tu r e, te x t i l e s , tr ade related investment measures (TRI M S) and trade related intelle c t u a l pro p e r ty rights (TRI P S ) are utterly devoid of any econo mic rat i o nale beyond the mer- cantilist interests of a narrow set of powe r ful groups in the advanced industrial coun- t ri e s . The devel o p mental pay-off of most of these req u i r ements is hard to see. Bi l a t e r al and reg i o nal trade agreements are often far worse, as they impose even tighter prer equisites on developing countries in ret u r n for crumbs of enhanced ‘ma rk e t ac c e s s ’ in the larger partn e r s . The Afric a Growth and Opportu n i t y Act passed by the U.S . Con g r ess in 2000, for example, co ntains a long list of eligibility cri t e ri a ,i n cluding the req u i r ement that Afric an gover nments minimize interfer ence in the econom y . It provi d e s fr ee access to U.S . ma r kets onl y under strict rules of orig i n , th e r eby ensuring that few eco- In each of these areas, a no mic linkages are generated in the Afric an countries themselves . The U.S. - J o r dan Free strategy focused on inte- Trade Agreement imposes more res t ri c t i v e intellectual prop e r ty rules on Jordan than exist under the WTO. gration crowds out more In each of these area s , a strat e g y focused on integrat i o n crowds out more devel o p - ment- frie n d ly alterna t i ve s . Ma n y of the institutional ref o r ms needed for inserti o n in development- friendly the world econo my can be independently desirab l e , or produce broader spillo ver s . Bu t alternatives. these prio r ities do not necessaril y coincide with the prio r ities of a broader devel o pm e n t ag e n d a . A strat e g y that focuses on getting the state out of the way of the market over - lo o ks the important functions that the state must play during the process of econom i c tra n s f o rm a t i on . What belongs on the agenda of institutional ref o r m is building up state cap a c i t y—not diminishing it (Evans 2000). Wor ld markets are a source of techn o l o g y and cap i t a l ; it would be silly for the de veloping world not to exploit these opportu n i t i e s . Bu t , as I have argued above, su c - cessful devel o p ment strategies have alwa ys req u i r ed a judicious blend of imported prac - tices with domestic institutional innova t i on s . Pol i c y-makers need to forge a do m e s t i c gr owth strat e g y, rel ying on domestic investors and domestic institutions . The most co s t l y downside of the integrat i o nist agenda is that it is crowding out serious thinking and efforts along such lines.

An International Trade Regime That Puts Development First: General Principles Access to the markets of the industrial countries matters for devel o pm e n t . But so does the autono my to experiment with institutional innova t i o ns that diverge from orth o - dox y . The exchange of reduced policy autono my in the South for improved marke t access in the Nor th is a bad bargain where devel o p ment is conc e rn e d . Co nsider the old GATT syst e m , under which the interna t i o nal trade regime did not rea c h muc h beyond tariff and non- t a r iff barriers to trad e . The developing coun- tr ies were effec t i ve l y exempt from prevailing disciplines. The ‘most favoured nation’ pr inciple ensured that they benefited from the tariff cuts negotiated among the indus- tr ial countrie s , while they themselves ‘ga v e up’ little in ret u rn . The resulting pattern of li b e ra l i za t i o n may have been asymm e t r ic (with many products of interest to devel o p i n g

27 Dani Rodrik

co u n t r ies either excluded or receiving less beneficial trea t m e n t ) , but the net effect for the developing world was still highly salutary. It is in such an envi r onment that the most successful ‘gl o b a l i ze r s ’ of an earli e r er a—the East Asian tigers—managed to pros p e r . These countries were free to do their own thing, and did so, co mbining trade reliance with unorth o d o x policies—export su b s i d i e s , dom e s t i c - c o ntent req u i re m e n t s , im p o rt - e x p o r t linkages, patent and copyrig h t in f ri n g e m e n t s , res t ri c t i o ns on capital flows (including direct foreign inves t m e n t ) ,d i re c t e d cr edit and so on—that are largely prec luded by today’s rul e s . 6 In fact, su c h policies we r e part of the arsenal of today’s advanced industrial countries until quite rec e n t l y (see Sch e r er and Watal 2001). The envi r onment for today’s globalizers is significan t l y more res t ri c t i v e (see Amsden 2000). For the world ’s poorest econom i e s , the so-cal led least developed countrie s Is it possible to preserve (LL DC s ) , so mething along the old GATT lines is still achi ev a b l e , and would cons t i t u t e a more devel o pm e n t - f ri e n d ly regime than the one that exists curren t l y. LLD Cs are developing countries’ ec on o mies that are individually and colle c t i ve l y small enough that ‘ad j u s t m e n t ’ issues in autonomy while also the advanced countries are not a serious obstacle to the provi s i o n of one-sided free - ma r ket access in the Nor th to the vast majorit y of products of interest to them. In s t e a d respecting the legitimate of encumbering these countries with all kinds of institutional req u i r ements that come at t a c hed to a ‘single underta k i n g, ’ it would be far better to leave them the roo m to follo w objectives of advanced their own institutional prio ri t i e s , while providing them with access into north e r n industrial countries to ma r kets that is both duty free and free of quantitative res t ri c t i on s . In prac t i c e , this can be done either by extending existing ‘pha s e - i n ’ pe r iods until certain income thres h o l d s maintain high labour, ar e rea ch e d , or incorpo r ating a general LLD C exce p t i on . In the case of middle - i n c o me and other developing nations , it is unrealistic to social and environmental expect that advanced industrial countries would be willing to accept a similar arran g e - standards at home? me n t . The amount of political opposition that imports from developing countries gen- er ate in the advanced industrial countries is already disprop o rt i o nate to the volume of tr ade in question. Some of these objectives have a legitimate core, and it is importa n t that developing nations understand and accept this (see Mayda and Rodrik 2001). Under a sensible set of global trade rul e s , in d u s t ri a l i z ed countries would have as muc h right to protect their own social arrangements—in areas such as labour and envi r onmental standards , we l f a r e-state arran g e m e n t s , rur al commu n i t i e s , or industrial or g a n i za t i o n—as developing nations have to adopt divergent institutional prac t i c e s . Co u n t r ies such as India, Bra z i l , or China, whose exports can have a sizable impact on, sa y, la b o u r - m a r ket institutions and employment rel a t i o ns within the advanced countrie s , cannot ask importing countries to ove rl o ok these effects while demanding at the same time that the cons t r aints on their own devel o p mental agenda be lifte d . Mi d dl e - i n c om e de veloping countries have to accept a more balanced set of rights and obligations Is it possible to pres e r ve developing countrie s ’ au t on o my while also respecting the legitimate objectives of advanced industrial countries to maintain high labour, so c i a l and envi r onmental standards at home? Would such a regime of world trade avoid col- lapsing into prot e c t i on i s m , bi l a t e r alism or reg i o nal trade blocs? Would it in fact be de vel o pm e n t - f ri e n d ly? The answer to all these questions is yes , pr ovided we accept five simple prin c i p l e s .

28 The Global Governance of Trade

Trade is a means to an end, not an end in itself. Step number one is to move away from attaching normative significance to trade itself. The scope of market access generated by the international trade regime and the volume of trade thereby stimulated are poor measures of how well the system functions. As the WTO’s preamble emphasizes, trade is useful only insofar as it serves broader developmental and social goals. Developing countries should not be obsessed with market access abroad, at the cost of overlooking more fundamental developmental challenges at home. Industrial countries should balance the interests of their exporters and multinational companies with those of their workers and consumers. Adv o c ates of globalizat i o n lecture the rest of the world incessantly about the adjustments countries have to undertake in their policies and institutions in order to expand their interna t i o nal trade and become more attrac t i v e to foreign inves t o r s . This is another instance of confusing means for ends. Trade serves at best as an instrument for achi e v- Trade rules should seek ing the goals that societies seek: pro s p e ri t y, st a b i l i t y, fre e d o m and quality of life. peaceful co-existence Nothing enrages WTO bashers more than the suspicion that, when push comes to sh o ve, the WTO allo ws trade to trump the envi r onment or human rig h t s . And devel - among national practices, oping countries are right to resist a system that evaluates their needs from the perspective of expanding world trade instead of pover ty red u c t i on . not harmonization. Re versing our prio r ities would have a simple but powe r ful implicat i on . Instead of asking what kind of mul t i l a t e r al trading system maximizes foreign trade and inves t m e n t op p o rt u n i t i e s , we would ask what kind of mul t i l a t e r al system best enables nations ar ound the world to pursue their own values and devel o p mental objectives .

Trade rules have to allow for diversity in national institutions and standards. As I have emphasized above, there is no single recipe for economic advancement. This does not mean that anything and everything works: market-based incentives, clear property- control rights, competition and macroeconomic stability are essential everywhere. But even these universal requirements can be and have been embodied in diverse institutional forms. Investment strategies, needed to jump-start economies, can also take different forms. Mo re o ver , ci t i z ens of differ ent countries have varying prefe r ences over the role of gover nment reg u l a t i o ns or provi s i o n of social welfare, how e ver imperfec t l y these prefe r - ences are articulated or determ i n e d .T h ey differ over the nature and extent of reg u l a t i on s to gover n new technologies (such things as genetical ly modified organisms) or protect the env i r onm e n t , of policies to extend social safet y nets and, mo r e broa d ly, about the entire rel a t i o nship between efficiency and equity. Ri c h and poor nations have ver y differe n t needs in the areas of labour standards or patent prot e c t i on . Poor countries need the space to follo w devel o p mental policies that ric her countries no longer req u i re . When countrie s use the trade system to impose their institutional prefe r ences on others, the result is ero- si o n of the syst e m ’s legitimacy and efficac y. Trade rules should seek peaceful co-existence am o ng national prac t i c e s , not harmon i za t i on .

29 Dani Rodrik

Non-democratic countries cannot count on the same trade privileges as democratic ones. National standards that deviate from those in trade partners and thereby provide ‘trade advantages’ are legitimate only to the extent that they are grounded in free choices made by citizens. Think of labour and environmental standards, for example. Poor countries argue that they cannot afford to have the same stringent standards in these areas as the advanced countries. Indeed, tough emission standards or regulations against the use of child labour can easily backfire if they lead to fewer jobs and greater poverty. Democratic countries such as India and Brazil can legitimately argue that their practices are consistent with the wishes of their own citizens, and that therefore it is inappropriate for labour groups or NGOs in advanced countries to tell them what standard they should have. Of course, democracy never works perfectly (in either developing countries or in advanced countries), and one would not want to argue that there are no human rights abuses in the countries just mentioned. The point is simply that the presence of civil liberties and political freedoms provides a presumptive cover against the charge that labour, environmental and other standards in the developing nations are inappropriately low. But in non- d e m o c r atic countrie s , su c h as China, the asserti o n that labour rig h t s and the envi r onment are trampled for the benefit of comm e r cial advantage cannot be as easily dismissed. Con s e q u e n t l y, ex p o r ts of non- d e m o c r atic countries deserve grea t e r sc ru t i n y when they entail costly dislocat i o ns or adverse distrib u t i o nal consequences in im p o r ting countrie s . In the absence of the pres u m p t i v e cover provided by democrat i c ri g h t s ,s u ch countries need to make a ‘de vel o pm e n t a l ’ case for policies that generat e adjustment difficulties in the importing countrie s . For example, mi n i m um wages that ar e significan t l y lower than in ric h countries or health and other benefits that are less ge n e r ous can be justified by pointing to lower labour prod u c t i v i t y and living standards in poor nations . Lax child labour reg u l a t i o ns can sometimes be justified by the argument that under cond i t i o ns of widespread pover ty it is not feasible or desirable to with- dr aw young workers from the labour force . In other cas e s , the ‘af f o rd a b i l i t y’ ar g u m e n t car r ies less weight: non - d i s c ri m i n a t i on , fre e d o m of association, co ll e c t i v e bargaining, pr ohi b i t i o n of forced labour do not ‘co s t ’ an yth i n g ; co mpliance with these ‘co r e labour rig h t s ’ does not harm, and indeed possibly benefits, ec on o mic devel o pm e n t . The latter ar e examples that do not pass the ‘de vel o p ment test.’

Countries have the right to protect their own institutions and development priorities. Opponents of today’s trade regime argue that trade sets off a ‘race to the bottom,’ with nations converging towards the lowest levels of environmental, labour and consumer protections. Advocates counter that there is little evidence that trade leads to the ero- sion of national standards. Developing nations complain that current trade laws are too intrusive, and leave little room for development-friendly policies. Advocates of the WTO reply that these rules provide useful discipline to rein in harmful policies that would otherwise end up wasting resources and hampering development. One way to cut through this impasse is to accept that countries can uphold nationa l st a n d a r ds and policies in these area s , by withholding market access or suspending WTO ob l i g a t i o ns if necessary, when trade demons t ra b l y undermines domestic practices that

30 The Global Governance of Trade

en j o y broad popular support. For example, poor nations might be allo wed to subsidize in d u s t r ial activities (and indirec t l y, their exports) when this is part of a broa d ly supported devel o p ment strat e g y aimed at stimulating techn o l o g i c al cap a b i l i t i e s . Adv a n c e d co u n t r ies might seek temporar y prot e c t i o n against imports originating from countrie s with weak enforcement of labour rights when such imports serve to worsen worki n g con d i t i o ns at home . The WTO already has a ‘sa fe g u a r d’ sy stem in place to prot e c t fi r ms from import surges. An extension of this principle to protect devel o p mental pri- or ities or envi r onm e n t a l , labour and cons u m e r - s a fe t y standards at home–with approp r i- ate proc e d u r al res t r aints against abuse–might make the world trading system more de vel o pm e n t - f ri e n d ly, mo r e resilient and less resistant to ad-hoc prot e c t i on i s m . Cu r re n t l y, the Agreement on Safe g u a r ds allo ws (temporar y) increases in trad e res t ri c t i o ns under a ver y narrow set of cond i t i o ns (see Rodrik 1997). It req u i r es a determi n a t i o n that in c r eased im p o r ts ‘cause or threaten to cause serious injury to the dome s - tic industry,’ that cau s a l i t y be firml y established and that if there are multiple cau s e s , in j u r y not be attributed to imports . Safe g u a r ds cannot be applied to devel o p i n g - co u n t r y exporters unless their share of imports of the product conc e r ned is above a th re s h o l d . A country applying safeg u a r d measures has to compensate the affec t e d ex p o r ters by providing ‘equivalent conc e s s i on s , ’ la c king which the exporter is free to ret a l i a t e . A broader interpre t a t i o n of safeg u a r ds would ackn o wledge that countries may le g i t i m a t e l y seek to res t r ict trade or suspend existing WTO obligations—to exerc i s e what I cal l ‘opt-outs’—for rea s o ns going beyond comp e t i t i v e threats to their indus- tri e s . Amo ng such rea s o ns are, as I have discussed, de vel o p mental prio r ities as well as distrib u t i o nal conc e r ns or conflicts with domestic norms or social arrangements in the industrial countrie s . We could imagine rec asting the current agreement into an Ag r eement on Devel o p mental and Soc i a l Safe g u a rd s , wh i c h would permit the applicat i o n of opt-outs under a broader range of circu m s t a n c e s . This would req u i r e re- casting the ‘se r ious injury’ test and replacing it with the need to demons t r ate broa d do mestic support, among all concerned parti e s , for the proposed measure.

To see how that might work in prac t i c e , co nsider what the current agreement says:

A Member may apply a safeg u a r d measure onl y follo wing an inves t i g a t i o n by the competent authorities of that Member pursuant to proc e d u r es previ o u s l y established and made public in cons o nance with Arti c le X of the GATT 19 9 4 . This inves t i g a t i o n shall include rea s o nable public notice to all interes t - ed parties and public hearings or other approp r iate means in which im p o r ters , exp o r ters and other interes t ed parties could present evidence and their views , in c luding the opportu n i t y to res p o nd to the pres e n t a t i o ns of other parties and to submit their views , inter alia, as to wh e t h e r or not the app l i c ation of a safe- gu a r d measure would be in the public interes t . The competent authorities shall publish a rep o r t setting forth their findings and rea s o ned concl u s i o ns rea ch e d on all pertinent issues of fact and law. (WTO 1995:9; em p hasis added)

The main shortc o ming of this clause is that while it allo ws all rel e vant grou p s ,

31 Dani Rodrik

and exporters and importers in parti c u l a r , to make their views known , it does not ac t u a l ly compel them to do so. Con s e q u e n t l y, it results in a strong bias in the do mestic inves t i g a t i v e process towa r ds the interests of import- c o mpeting grou p s , who are the petitioners for import relief and its obvious beneficiarie s . In d e e d , this is a key problem with hearings in anti-dumping proc e e d i n g s , wh e r e testimony from other groups besides the import- c o mpeting industry is typi ca l ly not allow e d . The most significant and reliable guarantee against the abuse of opt-outs is in f o r med deliberat i o n at the national level . A crit i c al ref o rm , th e n , would be to req u i r e the inves t i g a t i v e process in each country to: (1) gather testimony and views fr om all rel e vant parti e s , in c luding consumer and public-interest grou p s , im p o rt e r s and exporte r s , civil society organizat i on s , and (2) determine whether there exists su f - ficiently broa d support am o ng these groups for the exer cise of the opt-out or safeg u a r d An automatic sunset in question. The req u i r ements that groups whose incomes might be adver s e l y affec t - ed by the opt-out—importers and exporters—be comp e l led to testify, and that the clause could ensure that in ves t i g a t i v e body trade off the competing interests in a tran s p a r ent manner would trade restrictions and help ensure that prot e c t i o nist measures that benefit a small segment of industry at a large cost to society would not have muc h chance of success. When the opt-out in opt-outs do not become qu e s t i o n is part of a broader devel o p ment strat e g y that has already been adopted af ter broad debate and parti c i p a t i on , an additional inves t i g a t i v e process need not be entrenched long after la u n ch e d . This last point deserves to be highlighted in view of the emphasis placed their perceived need has on ‘lo c al own e r s h i p ’ and ‘pa rt i c i p a t o r y mecha n i s m s ’ in strategies of pover ty red u c t i o n and growth promoted by the interna t i o nal financial institutions . disappeared. The main advantage of this proc e d u r e is that it would force a public debate on the legitimacy of trade rules and when to suspend them, en s u r ing that all sides would be heard. This is something that rare l y happens even in the industrial countrie s , le t al o ne in developing nations . This proc e d u r e could be complemented with a st r engthened moni t o r ing and survei l lance role for the WTO, to ensure that dome s t i c opt-out proc e d u r es are in compliance with the expanded safeg u a r d cla u s e . An automatic sunset clause could ensure that trade res t ri c t i o ns and opt-outs do not become en t re n c hed long after their perce i v ed need has disappeared . All o wing opt-outs in this manner would not be without its ris k s . The possibility that the new proc e d u r es would be abused for prot e c t i o nist ends and open the door to un i l a t e r al action on a broad front , despite the high threshold envisaged here, has to be taken into account. But as I have already argued, the current arrangements also ha v e ris k s . The ‘mo r e of the same’ ap p ro a c h embodied in the industria l i z ed coun- tri e s ’ ef f o r ts to launch a comp re h e n s i v e new round of trade negot i a t i o ns is unlikely to pr oduce benefits for developing nations . Absent crea t i v e thinking and novel institu- ti o nal designs, the narrowing of the roo m for institutional divergence harms devel o p - ment pros p e c t s . It may also lead to the emergence of a new set of ‘gr ey area ’ me a s - ur es entirel y outside mul t i l a t e r al discipline. These are consequences that are far worse than the expanded safeg u a r d regime I have just describ e d .

But countries do not have the right to impose their institutional preferences on others. The exercise of opt-outs to uphold a country’s own priorities has to be sharply dis-

32 The Global Governance of Trade

tinguished from using them to impose these priorities on other countries. Trade rules should not force Americans to consume shrimp that are caught in ways that most Americans find unacceptable; but neither should they allow the United States to use trade sanctions to alter the way that foreign nations go about their fishing business. Citizens of rich countries who are genuinely concerned about the state of the environment or of workers in the developing world can be more effective through channels other than trade—via diplomacy or foreign aid, for example. Trade sanctions to promote a country’s own preferences are rarely effective, and have no moral legitimacy (except for when they are used against repressive political regimes). This and the previous principle help us draw a useful distinction between two st yles of ‘un i l a t e ra l i s m’ — o ne that is aimed at protecting differe n c e s , and the other aimed at reducing them. When the European Union drags its feet on agric u l t u ra l tr ade liberal i za t i on , it is out of a desire to ‘pro t e c t ’ a set of domestic social arran g e - ments that Europ e a n s , th r ough their democratic proc e d u re s , ha v e decided are worth ma i n t a i n i n g .Whe n , on the other hand, the United States threatens trade sanctions against Japan because its retailing practices are perce i v ed to harm Ame ri c an exporte r s or against South Afric a because its patent laws are perce i v ed as too lax, it does so out of a desire to bring these countrie s ’ pr actices into line with its own . A well- d e s i g n e d wo r ld trade regime would leave roo m for the forme r , but prohibit the latter.

Other development-friendly measures. In addition to providing unrestricted access to least developed countries’ exports and enabling developing countries to exercise greater autonomy in the use of subsidies, ‘trade-related’ investment, patent regulations and other measures, a development-friendly trade regime would do the following (see UNCTAD 2000; Raghavan 1996):

• gre a t l y res t r ict the use of anti-dumping (AD) measures in advanced industria l co u n t r ies when exports originate from developing countrie s . A small, bu t im p o r tant step would be to req u i r e that the rel e vant investigating bodies take fu l ly into account the consumer costs of anti-dumping action. • all o w greater mobility of workers across interna t i o nal boundarie s , by liberal i z i n g for example the movement of natural persons connected to trade in labour- in t e n s i v e services (such as cons t ru c t i on ) . • req u i r e that all existing and future WTO agreements be fully costed out (in te r ms of implementation and other costs). It would cond i t i o n the phasing in of these agreements in the developing countries on the provi s i o n of comm e n s u ra t e financial assistance. • req u i r e additional comp e n s a t i o n when a dispute settlement panel rules in favour of a developing country comp l a i n a n t , or (when comp e n s a t i o n is not forth c o m- ing) req u i r e that other countries join in the ret a l i a t i on . • pr ovide expanded legal and fact-finding assistance to developing country mem- bers of the WTO in pros p e c t i v e dispute settlement cas e s .

33 Dani Rodrik

Conclusions: From a Market-Exchange Perspective to a Development Perspective Ec on o mists think of the WTO as an institution designed to expand free trade and th e r eby enhance consumer welfare, in the South no less than in the Nort h . In rea l i t y, it is an institution that enables countries to bargain about market access. ‘Free trad e ’ is not the typi c al outcome of this proc e s s ; nor is consumer welfare (muc h less devel o pm e n t ) what the negotiators have chi e f l y in mind. Trad i t i on a l ly, the agenda of mul t i l a t e r al trad e ne go t i a t i o ns has been shaped in res p o nse to a tug-of-war between exporters and mul t i - na t i o nal corpo ra t i o ns in the advanced industrial countries (which have had the upper ha n d ) , on the one hand, and import- c o mpeting interests (typi ca l ly, but not solely, la b o u r ) on the other. The chief textbo o k beneficiaries of free trad e — c o nsumers—do not sit at the table. The WTO can best be understood in this cont e x t , as the product of intense lo b b y ing by specific exporter groups in the United States or Europe or of specific com- pr omises between such groups and other domestic grou p s . The differ ential treatment of ma n u f a c t u r es and agric u l t u r e, or of clothing and other goods within manufacturin g , th e anti-dumping reg i m e , and the intellectual prop e r ty rights (IPR) reg i m e , to pick some of the major anom a l i e s ,a re all results of this political proc e s s . Understanding this is essen- ti a l , as it underscores the fact that there is ver y little in the struc t u r e of mul t i l a t e r al trad e ne go t i a t i o ns to ensure that their outcomes are consistent with devel o p ment go a l s ,l e t al o ne that they be designed to further devel o pm e n t . Hence there are at least three sources of slippage between what devel o pm e n t req u i r es and what the WTO does. Fir s t , even if free trade were optimal for devel o p - ment in its broad sense, the WTO does not fundamentally pursue free trad e . Se c on d , even if it did, th e r e is no guarantee that free trade is the best trade policy for co u n t r ies at low levels of devel o pm e n t . Th i rd , co mpliance with WTO rul e s , even when these rules are not harmful in themselves , cr owds outs a more fully devel o p - mental agenda—at both the interna t i o nal and national level . My main argument has been that the world trading regime has to shift from a ‘ma r ket access’ pe r s p e c t i v e to a ‘de vel o pm e n t ’ pe r s p e c t i v e (see Helleiner 2000:19). Es s e n t i a l ly, the shift means that we should stop evaluating the trade regime from the pe r s p e c t i v e of whether it maximizes the flow of trade in goods and servi c e s , and ask in s t e a d , ‘Do the trading arran g e m e n t s — c u r r ent and prop o s e d — m a x i m i z e the possibilities of devel o p ment at the national level ? ’ I have discussed why these two per- sp e c t i v es are not the same, even though they sometimes overl a p , and have outlined so me of the operat i o nal implicat i o ns of such a shift. One is that developing nations ha v e to articulate their needs not in terms of market access, but in terms of the policy au t on o my that will allo w them to exer cise institutional innova t i o ns that depart from pr evailing orth o d ox i e s . A second is that the WTO should be conc e i v ed of not as an in s t i t u t i o n devoted to harmon i za t i o n and the red u c t i o n of national institutional differe n c e s , but as one that manages the interface between differ ent national syst e m s . This shift to a devel o p ment perspective would have sever al important advan- ta g e s . The first and more obvious is that it would provide for a more devel o pm e n t - fri e n d ly interna t i o nal econo mic envi r onm e n t . Co u n t r ies would be able to use trad e as a means for devel o pm e n t , rather than being forced to view trade as an end in itself

34 The Global Governance of Trade

(and being forced to sacrifice devel o p ment goals in the bargain). It would save de veloping countries precious political capital by obviating the need to bargain for ‘special and differ ential trea t m e n t ’—a principle that in any case is more form than substance at this point. Se c on d , vi e wing the WTO as an institution that manages institutional diver s i t y (r ather than imposing uniformi t y) provides developing countries a way out of a con u n d r um inherent in their current negotiating stance. The problem arises from the incons i s t e n c y between their demands for space to implement their devel o pm e n t policies on the one hand, and their complaints about north e r n prot e c t i o nism in agri- cu l t u r e, textiles and labour and envi r onmental standards , on the other. As long as the issues are viewed in market-access terms , de veloping countries will be unable to make a sound and principled defense of their legitimate need for space. And the onl y way th e y can gain enhanced market access is by res t r icting their own policy autono my in The WTO should be ex cha n g e . Once the objective of the trading regime is seen as letting differe n t conceived of not as an na t i o nal econo mic systems prosper side by side, the debate can become one about ea c h nation’s institutional prio r ities and how they may be ren d e r ed compatible in a institution devoted to de vel o pm e n t - f ri e n d ly way. The third advantage of this shift in perspective is that it provides a way out of harmonization and the the impasse that the trading system finds itself post-Sea t t l e . At pres e n t , two grou p s reduction of national feel parti c u l a r ly excluded from the decision-making machi n e r y of the global trad e reg i m e : de veloping country gover nments and north e r n NGOs. The former comp l a i n institutional differences, about the asymm e t r y in trade rul e s , while the latter charge that the system pays inadequate attention to values such as tran s p a re n c y, ac c o u n t a b i l i t y, human rights and but as one that manages env i r onmental sustainability. The demands of these two disenfran ch i z ed groups are the interface between of ten perce i v ed to be conf l i c t i n g — o ver questions such as labour and envi r onm e n t a l st a n d a r ds or the tran s p a re n c y of the dispute settlement proc e d u re s — a ll o wing the different national advanced industrial countries and the WTO leadership to seize the ‘m i d dl e’ gro u n d . It is the demands of these two grou p s , and the apparent tension between them, th a t systems. has paral yzed the process of mul t i l a t e r al trade negot i a t i o ns in recent yea r s . But once the trade regime—and the gover nance cha l lenges it poses—is seen fr om a devel o p ment perspective, it becomes clear that developing country gover n- ments and many of the north e r n NGOs share the same goa l s : po l i c y autono my to pursue independent values and prio ri t i e s , po ver ty red u c t i on , and human devel o pm e n t in an envi r onm e n t a l ly sustainable manner. The tensions over issues such as labour st a n d a r ds become manageable if the debate is couched in terms of devel o pm e n t pro c e s s e s — b ro a d ly defined—instead of the req u i r ements of market access. On all co u n t s , th e n , the shift in perspective provides a better foundation for the mul t i l a t e ra l tr ading reg i m e .

35 Dani Rodrik

No t e s

1. See , for example, Mike Moore (2000) or his speech at the Lond o n Ministeria l rou n d t a b l e , 19 March 2001 (www.wt o. o r g / e n g l i s h / n ew s ) .

2. The supposition that one set of institutional arrangements must dominate in terms of overa l l perfo r mance has produced the fads of the decad e : Eu ro p e , with its low unem- pl o yme n t , high growth and thriving culture, was the continent to emulate throu g h o u t muc h of the 1970s; du r ing the trad e - c o nscious 1980s, Japan became the exemplar of cho i c e ; and the 1990s have been the decade of U.S. - s t yle free wheeling cap i t a l i s m .

3. For example, although Taiwan and Mexico are comm on l y reg a r ded as follo wing dia- me t ri ca l ly opposed devel o p ment paths, fi g u r es provided by Little et al. (1 9 7 0 : 17 4 - 90) show that long after introducing trade ref o rm s , Taiwan had a higher avera g e ERP in manufacturing and greater varia t i o n in ERP s than did Mexico.

4. Although India did grad u a l ly liberal i z e its trade regime after 1991, its rel a t i v e perfo r mance began to improve a full decade before these ref o r ms went into effect (in the ea r ly 1980s).

5. The same is true of the promo t i o n and subsidizat i o n of inwa r d flows of direct forei g n in vestment (see Hanson 2001).

6. A recent illu s t ra t i o n is the dispute between Brazil and Canada over Braz i l ’s subsi- di za t i o n of its aircra f t manufacturer , Em b r aer . Br azil lost this case in the WTO, an d wi l l either rem o ve the subsidies or have to put up with ret a l i a t i o n from Canada. Th e Republic of Korea , Tai w a n , pr ovince of China and Mauritius subsidized their export in d u s t r ies for years without incurring similar sanctions .

36 The Global Governance of Trade

Re f e re n c e s

Ac e m o g l u ,D a ron, Sim o n John s o n and James A. R o b i n s on .2 0 0 0 . ‘The Colonial Orig i n s of Comp a ra t i v e Devel o pm e n t : An Empiric al Inves t i g a t i on , ’ unpublished paper, Ma s s a c husetts Institute of Tech n o l o g y, Ca m b ri d g e , MA .

Ams d e n ,A l i c e .2 0 0 0 . ‘In d u s t ri a l i za t i o n Under New WTO Law, ’ paper prep a r ed for the Hi g h - L evel Round Table on Trade and Devel o pm e n t ,U N C T AD X, Ba n g k ok , Feb ru a r y 12.

Ba l a s s a , Be l a . 1 9 7 1 . The Struc tu r e of Prot ection in Developing Countrie s . Ba l t i m o re : Joh n s Hopkins Univer s i t y Pres s .

Be n - D a v i d , Da n . 19 9 3 . ‘Equalizing Excha n g e : Trade Liberal i za t i o n and Income Co nver g e n c e, ’ Qua r terly Journal of Ec o n o m i c s 10 8 , no . 3.

Ch a n g , Ha - J o on .2 0 0 0 . ‘In s t i t u t i o nal Devel o p ment in Developing Countries in a Historica l Per s p e c t i v e,’ unpublished paper, Fac u l t y of Econo mics and Pol i t i c s ,C a m b r idge Univer s i t y.

Co ll i n s , Susan and Barry Bosworth . 19 9 6 . ‘Ec on o mic Growth in East Asia: Acc u mu l a t i o n versus Assimilation, ’ Bro ok i n g s Pape rs on Economic Activi t y 2: 13 5 - 9 1 .

Do ll a r , Da v i d . 1 9 9 2 .‘ Ou tw a rd - O riented Developing Econo mies Really Do Grow More Rap i d ly: Evidence from 95 LDCs , 19 7 6 - 8 5 , ’ Economic Development and Cultur al Change, pp . 52 3 - 4 4 .

Do ll a r , David and Aart Kraa y. 20 0 0 . ‘Trad e , Gr owth , and Pover ty,’ Wor ld Bank, Was h i n g t on , DC, Oc t o b e r .

Ed w a rd s , Seb a s t i a n . 19 9 8 . ‘O p e n n e s s , Prod u c t i v i t y and Growth : What Do We Really Kn ow ? ’ Economic Journal 108 (March) : 38 3 - 9 8 .

Ev a n s , Pet e r . 20 0 0 . ‘Ec on o mic Gover nance Institutions in a Global Pol i t i c al Econom y : Im p l i ca t i o ns for Developing Countrie s , ’ paper prep a r ed for the High-Level Round Tab l e on Trade and Deve l o pm e n t , ’U N C TAD X, Ba n g k ok , Feb ru a r y 12.

Fin g e r , Mi ch a el J. and Philip Schu l e r . 19 9 9 . ‘Im p l e m e n t a t i o n of Urug u a y Round Com m i t m e n t s : the Devel o p ment Challe n g e, ’ Wor ld Bank policy res e a r ch working paper no . 22 1 5 , Sep t e m b e r .

Fra n k e l ,J e f f rey and David Rome r . 19 9 9 . ‘Does Trade Cause Growth ? ’ Ame ric an Economic Rev i e w 89 , no .3 (Ju n e ) : 37 9 - 9 9 .

Ha n s on , Go rd on . 2 0 0 1 .‘ Should Countries Promote For eign Direct Inve s t m e n t , ’G ro u p of 24 Discussion Paper no. 9, Feb ru a r y.

He ll e i n e r , Ge r ald K. 20 0 0 . ‘Ma rk e t s , Pol i t i c s , and the Global Econom y : Can the Global Ec on o my Be Civilized ? ’ Preb i s c h Lec t u r e, UN C TA D , Ge n ev a .

—— . 1 9 9 4 . Trade Pol i c y and Industria l i z ation in Turbulent Tim e s . New York : Ro u t l e d g e .

In t e rn a t i o nal Labour Office. 20 0 1 . Mau ri t i u s . St udies on the Social Dimensions of Gl o ba l iza t i o n. G e n ev a :I LO.

37 Dani Rodrik

Lau , Law rence J., Yin g yi Qian and Gerar d Roland. 19 9 7 . ‘Re f o r m Without Los e r s : An In t e rp re t a t i o n of China’s Dua l - Trac k App ro a c h to Tran s i t i on , ’ unpublished paper, Novem b e r .

Li t t l e , Ia n , Tibor Scitovsky and Maurice Scott. 19 7 0 . In d u s t ry and Trade in Some De veloping Countrie s . Lond o n and New York : Oxf o r d Univer s i t y Press for OECD.

Ma yda , Anna Maria and Dani Rodrik . 2001 ‘Why Are Some People (or Countrie s ) Mo r e Prot e c t i o nist than Others? Harva r d Univer s i t y, Ca m b ri d g e , MA .h t tp : / / w w w / h a r - va rd . e d u / ro d ri k .

Mo o r e, M i k e .2 0 0 0 . ‘The WTO Is a Friend of the Poo r, ’ Financial Tim e s , June 19, 20 0 0 .

Qi a n , Yin g yi ,G e ra rd Roland and Chenggang Xu. 1 9 9 9 .‘ C o o rdinating Changes in M- For m and U-For m Organizat i on s , ’ paper prep a r ed for the Nobel Symp o s i u m , Aug u s t .

Rag h a v a n ,C h a k ra v a rt h i . 19 9 6 . ‘The New Issues and Developing Countrie s , ’ TWN Trade & Devel o p ment Se ri e s ,n o. 4 ,T h i rd Wor ld Netw o rk , Pen a n g , Ma l a ysi a .

Ro d r í g u ez , Francisco and Dani Rodrik . 20 0 1 . ‘Trade Pol i c y and Econo mic Growth : A Ske p t i c ’s Guide to the Cros s - Na t i o nal Literat u re . ’Fort h c o ming in Ben Bernanke and Kenneth S. Ro go f f , ed s . , NB E R Macr o Annual 2000. Ca m b ri d g e , MA : NB E R. (h t tp : / / w w w. k s g. h a rv a rd . e d u / ro d ri k / s k e p t i 1 2 9 9 . p d f) .

R o d ri k ,D a n i . 20 0 1 a . ‘In s t i t u t i o ns for High-Qua l i t y Growth : What Th e y Are and How to Acq u i r e Th e m , ’ St udies in Comparat i ve International Develo p m e n t , fo rt h c om i n g .

______. 20 0 1 b . ‘Trading in Illu s i on s , ’ For eign Pol i c y, Ma r ch- Ap ri l .

______.2 0 0 0 a . ‘Fiv e Simple Principles for Wor ld Trad e, ’ The Ame ric an Pros p e c t , Ja n u a r y 17.

______. 20 0 0 b . ‘G r owth versus Pover ty Reduction: A Hollo w Debate,’ Finance & De velo p m e n t 37 , no .4 (December).

______. 19 9 9 . The New Global Economy and the Developing Countrie s : Making Openness Work , Overseas Devel o p ment Council, Was h i n g t on , DC, 19 9 9 .

______. 19 9 7 . Has Global i z ation Gone Too Far? Institute for Interna t i o nal Econom i c s , Was h i n g t on , DC 1997.

Sach s , Je f f r ey D., and Wing Th ye Woo . 20 0 0 . ‘Understanding China’s Econom i c Perf o rm a n c e, ’ Jo u r nal of Pol i c y Refo r m 4, no . 1.

Sach s , Je f f r ey, and And r ew Warn e r , 19 9 5 . ‘Ec on o mic Reform and the Process of Global In t e g ra t i on , ’ Bro ok i n g s Pape rs on Economic Activi t y, no. 1 : 1 - 1 1 8 .

Sch e re r ,F.M . and Jaya s h r ee Wat a l . 20 0 1 . ‘Pos t -T R IPS Options for Access to Pat e n t e d Medicines in Developing Countrie s , ’ Jo hn F. Kennedy School of Govern m e n t , Ha rv a r d Un i ve r s i t y.

Sen , Ama r ty a .1 9 9 9 . De velopment As Free d o m . New York : Al f r ed Knopf.

______. 19 9 3 .‘ C a p a b i l i t y and Well - b e i n g .’ In Martha Nussbaum and Ama r tya Sen , ed s . , The Qua l i t y of Li f e .Oxf o rd : Cl a re n d o n Pres s .

38 The Global Governance of Trade

Ste rn , Ni ch o l a s .2 0 0 0 .‘ G l o b a l i za t i on and Pover ty,’ pr esented at the Institute of Econom i c and Social Research, Fac u l t y of Econom i c s , Un i ve r s i t y of Indone s i a , December 15.

Sub ra m a n i a n , Arv i n d . 20 0 1 . ‘Ma u ri t i u s ’ Trade and Devel o p ment Stra t e g y: What Les s on s Does It Offer ? ’ paper presented at the IMF High-Level Seminar on Globalizat i o n and A f ri ca ,M a rch.

UN C TA D . 20 0 0 . Pos i t i ve Agenda and Futu r e Trade Negot i a t i o n s . Ge n ev a : UN C TA D .

Un g e r , Ro b e r to Mangabeira. 19 9 8 . De m o c r acy Realize d : The Prog res s i ve Alt erna t i ve. Lond o n and New York : Verso Books .

United Sta t e s , De p a r tment of Sta t e . 19 9 9 . ‘1999 Country Report on Econo mic Pol i c y and Trade Prac t i c e s : Ha i t i , ’ ht tp : / / w w w. s t a t e . g ov/ w w w / i s s u e s / e c on om i c / t ra d e _ re p o rt s / 1 9 9 9 / h a i t i . p d f .

Wor ld Bank. 20 0 0 . Wor ld Development Indicators 2000.Was h i n g t on , DC: Wor ld Bank.

Wor ld Bank and IMF. 20 0 0 . ‘Trad e , De vel o p ment and Pover ty Reduction, ’ joint paper pre p a r ed for the cons i d e ra t i o n of the Devel o p ment Comm i t t e e , Ma r ch 31.

Wor ld Trade Organizat i o n (WTO) . 19 9 5 . ‘Ag r eement Establishing the Wor ld Trad e Or g a n i za t i on . ’ WTO Informa t i o n and Media Relations Divisions , Ge n ev a . ht tp : / / wt o. o r g / e n g l i s h / d o c s _ e / l e g a l _ e / 0 4 - wt o. p d f .

39 Trade and Sustainable Human Development Project Bureau for Development Policy United Nations Development Programme 1 United Nations Plaza New York, NY 10017 www.undp.org/bdp BAC KG ROUND PA PER

T h e G l o b a l Governance o f Tr a d e

A s I f

D e v e l o p m e n t

Really Mattered

Dani Rodrik

United Nations Development Programme