global economic prospects - world bankpubdocs.worldbank.org/en/295201443469726534/global... · 1.4...

Global Economic Prospects

Global Economic Prospects

20052005Trade, Regionalism,and Development

Trade, Regionalism,and Development

34437P

ublic

Dis

clos

ure

Aut

horiz

edP

ublic

Dis

clos

ure

Aut

horiz

edP

ublic

Dis

clos

ure

Aut

horiz

edP

ublic

Dis

clos

ure

Aut

horiz

ed

GlobalEconomicProspects

Trade, Regionalism, and Development

2005

© 2005 The International Bank for Reconstruction and Development / The World Bank1818 H Street, NWWashington, DC 20433Telephone: 202-473-1000Internet: www.worldbank.orgE-mail: [email protected]

1 2 3 4 07 06 05 04

This volume is a product of the staff of the International Bank for Reconstruction andDevelopment / The World Bank. The findings, interpretations, and conclusions expressed hereindo not necessarily reflect the views of the Executive Directors of The World Bank or thegovernments they represent.

The boundaries, colors, denominations, and other information shown on any map in thisvolume do not imply on the part of the International Bank for Reconstruction andDevelopment / The World Bank any judgement on the legal status of any territory or theendorsement or acceptance of such boundaries.

Rights and Permissions

The material in this publication is copyrighted. Copying and/or transmitting portions or all ofthis work without permission may be a violation of applicable law. The International Bank forReconstruction and Development / The World Bank encourages dissemination of its work andwill normally grant permission to reproduce portions of the work promptly.

For permission to photocopy or reprint any part of this work, please send a request withcomplete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers,MA 01923, USA, telephone 978-750-8400, fax 978-750-4470, www.copyright.com.

All other queries on rights and licenses, including subsidiary rights, should be addressed tothe Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433,USA, fax 202-522-2422, e-mail [email protected].

0-8213-5747-6

Cover artist: John Yanson

iii

Foreword vii

Acknowledgments ix

Overview xi

Abbreviations xix

Frequently Cited Regional Trading Agreements and the Parties to Them xxiii

Chapter 1 Global Outlook and the Developing Countries 1The Global Economy: From Recovery to Expansion 3Commodity Markets 6World Trade 8International Finance 10Risks and Policy Ppriorities 12Long-Term Growth, Structural Change, and Poverty 14Structural Changes over Two Decades 16Structural Change in the Future 19Poverty Forecast 21Concluding Remarks 23Notes 23References 25

Chapter 2 Regional Trade and Preferential Trading Agreements: A Global Perspective 27

The Proliferation of Regional Preference Systems 28Trends in Trade and Growth by Region 42Changing Export Composition and the Rise of Global Production Networks 46Preferential Trade and Regional Outcomes 49Conclusion 53Notes 53References 54

Chapter 3 Regional Trade Agreements: Effects on Trade 57The Impact of RTAs on Merchandise Trade and Incomes 57Ingredients of Success 66

Contents

Conclusions: Preferential Trade Agreements and Economic Development 72Notes 74References 74

Chapter 4 Beyond Trade Policy Barriers: Lowering Trade Costs Together 77The Costs of Trade 78Regional Agreements to Facilitate Trade and Transport 79Standards, Conformity Assessments, and RTAs 85Trade-Related Regional Cooperation Agreements 91Conclusions 92Notes 93References 94

Chapter 5 Beyond Merchandise Trade: Services, Investment, Intellectual Property, and Labor Mobility 97

Services, Investment, and IPRs in Regional Agreements 98Economic Consequences of Services, Investment, and IPR Provisions in RTAs 104RTAs and Provisions for Movement of Labor 111Conclusions: Beyond Merchandise Trade 117Notes 119References 121

Chapter 6 Making Regionalism Complementary to Multilateralism 125Preferential Agreements within the Global Context 126Building Blocks versus Stumbling Blocks 133The Competitive Liberalization Game: The Case of Doha 136Multilateral Disciplines on Regional Arrangements 140Making Open Regionalism Work for Development 144Notes 148References 150

Figures1.1 A record year for developing countries in 2004 41.2 Strong growth across most regions 51.3 Tradable price developments 61.4 Terms of trade impacts from higher commodity prices, 2001–04 71.5 The oil-important burden for selected countries 71.6 Crude oil prices, 1960–2004 81.7 World trade rebounds 81.8 Export performance, percent change in market share since 2000 91.9 Trade balances in major regions 101.10 Developing countries’ debt and interest payments easing downward since 1999 101.11 Rising U.S. dollar reserves 111.12 Impact of a 200 basis point increase in interest rates 121.13 First year impacts of a $10 increase in oil prices 141.14 A rise in services 171.15 Rising openness to trade 18

G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 5

iv

1.16 Increased urbanization 191.17 Growth rate of labor supply declining 192.1 Number of RTAs exploded in the 1990s 292.2 Spaghetti and rigatoni: Multiple, overlapping RTAs, 2004 392.3 Trade within RTAs 412.4 Trade performance has differed across regions 432.5 Composition of trade has changed 472.6 Evolving trading blocs 503.1 Evolution of the share of intra-regional imports in total imports, 1960–2000 583.2 The ratio of external and intra-regional trade to GDP 593.3 Intra-regional trade grows faster when world trade growth is positive 603.4 Simulated welfare impact of various FTAs involving Chile 623.5 RTAs that divert imports tend to export less to global markets 643.6 Not all regions are open 663.7 Preferential tariffs in tandem with all tariffs in Latin America 683.8 Rules of origin in North–South agreements are more restrictive than in South–South

agreements 704.1 More efficient customs are associated with more trade 805.1 Share of EU nationals in total population, for selected EU countries and Norway,

1985–2000 1126.1 Global outcomes dominate alternatives 1316.2 Global reform dominates North–South agreements 1336.3 Global reform dominates South–South agreements 133

Tables1.1 The Global outlook in summary 21.2 Current account balances 111.3 Long-term prospects: Forecast growth of world GDP per capita 171.4 Labor market structure, 2005–15 201.5 Regional breakdown of poverty in developing countries 212.1 Most countries belong to more than one RTA 302.2 RTAs cover many topics besides merchandise trade 353.1 Implementation of tariff commitments by type of agreement, 1960–1999 695.1 Services, investment, and intellectual property: A comparison 995.2 Additional services liberalization in U.S. FTAs 1005.3 Summary of agreements by degree of labor mobility 1116.1 Comparison of bilateral agreements to global trade reform 1286.2 Comparison of bilateral agreements with global trade reform 129

Boxes1.1 The aggregation paradox 162.1 RTAs and types of trade liberalization 282.2 Reporting RTAs to the WTO 292.3 EPAs become the EU’s trade and development instrument: An experiment in

“North-South-South” integration 322.4 Labor in U.S. FTAs 33

C O N T E N T S

v

2.5 Trade agreements and the environment 362.6 Can RTAs prevent conflict? 382.7 Regional versus multilateral and unilateral liberalization:

What’s more important? 423.1 A primer on modeling of RTAs 613.2 Regional trade agreements in gravity models: A meta-analysis 633.3 Implementation matters 653.4 Restrictive rules of origin under NAFTA—the case of clothing 713.5 Monitoring implementation of preferential trade agreements:

“Single Market Scoreboard” in the European Union 734.1 Trading can be costly 794.2 Border delays tax trade 804.3 Standardization and simplification can increase trade volumes: The case of the

Trans-Kalahari Corridor 824.4 Logistics costs in Europe have fallen in the last two decades 834.5 The case of the Northern Corridor Stakeholders Consultative Forum 845.1 Not all investment is good investment 1065.2 Do more investor protections mean more investment? Lessons from bilateral

investment treaties 1075.3 Illegal migration: A growing global phenomenon 1135.4 U.S. temporary admission programs under NAFTA and unilateral policies 1166.1 Impacts of the new GTAP database 1276.2 Regional trade agreements, structural change, and congruence 1326.3 Choosing partners: Selection criteria for U.S. RTAs 1376.4 Sequencing of RTAs: Is there a good practice? 1396.5 RTAs and WTO disciplines 1416.6 Tunisia’s Association Agreement with the European Union 146


vi

THIS YEAR—2004—is shaping up to be the healthiest year for developing countries in the lastthree decades. East Asia has come out of the crisis of 1997–98 stronger and more vibrantthan ever; the countries of Europe and Central Asia are now almost completely out of the

long shadow of transition from socialism and are growing more rapidly; and South Asian coun-tries, on the strength of continuing reforms, are performing well. Moreover, countries in LatinAmerica, Sub-Saharan Africa, and the Middle East and North Africa had a much better year. Tobe sure, some countries within each region have not enjoyed the fruits of this recovery, and thesecountries remain a source of concern. But in aggregate, economic growth in 2004 was impressive.This performance reflects a fortuitous combination of (1) long-term secular trends built on a foun-dation of better macroeconomic management and (2) an improved domestic investment climateconverging with a cyclical recovery of the global economy.

This is no time for complacency. Lingering imbalances in the global economy associated withthe rising twin deficits in the United States, a delayed recovery in Europe, high and volatile oilprices, and questions about the path of China’s economy constitute risks to the pace of growthin developing countries over the medium term. Sustaining this pace is essential for the fate of mil-lions of the world’s poor.

World trade grew by 10.2 percent in 2004, and has played an important role in this year’s ex-emplary performance. On the policy side, the WTO discussions in August recouped the groundthat was lost in late 2003 (after the Cancun WTO ministerial), when governments concluded anagreement that provides a framework for pursuing the Doha Development Agenda. The frame-work, however, is only an outline—the actual agreement is still to come. Whether the final agree-ment contains provisions that will provide a meaningful impulse to development remains to beseen.

In the meantime, this year’s Global Economic Prospects examines the evidence on an impor-tant development that is reshaping the architecture of the world trading system: The dramatic pro-liferation of regional trade agreements (RTAs). These take various forms of preferential reciprocaltreaties—they can be bilateral or plurilateral free trade agreements or, less commonly, customsunions. In according preferential access to members, regional arrangements necessarily discrimi-nate against nonmembers. Even though arrangements in some instances can promote develop-ment, it is important to recognize that they also can lead to trade diversion in a way that hurts bothmember countries and excluded countries. Hence this year’s report identifies ways to design andimplement preferential trading agreements to maximize their benefits for participants and mini-mize their costs to nonmember developing countries. The key to making regional agreements com-plementary to a nondiscriminatory multilateral system is to strive for “open regionalism”—that

vii

Foreword

is, agreements with low external barriers to trade, nonrestrictive rules of origin, liberalized servicemarkets, and a strong focus on reducing transaction costs at borders.

The international community working together can leverage the August framework to achievean ambitious Doha deal. Multilateral liberalization has a greater positive impact on developmentthan do the myriad regional arrangements now being spawned seemingly in every corner of theglobe. Moreover, multilateral oversight of inherently discriminatory RTAs must be strengthened,and the first step is to increase transparency by empowering the WTO to collect and regularlymake public full details of all arrangements. If Doha can succeed in bringing down border pro-tection in agriculture and manufactures, it will reduce the discriminatory effects of regional agree-ments and lower the probability of costly trade diversion for participating countries. Said differ-ently, a strong Doha arrangement can contribute to open regionalism.

Along with other bilateral and multilateral institutions, the World Bank continues to promotethe integration of developing countries into the world economy so that the benefits of globaliza-tion can extend to the poor. The Bank now has 91 trade-related projects approved or planned in75 countries for the three-year fiscal period 2004–06. The actual and projected commitments fornew trade operations are at $2.9 billion—larger than the commitments of all ongoing operationsapproved over the preceding eight-year period of fiscal 1996–2003 ($2.4 billion). For trade fa-cilitation the growth is even larger. Projected commitments over fiscal 2004–06, at more than$1.2 billion, triple the commitments of ongoing trade facilitation operations approved betweenfiscal 1996–2003. To guide this lending and to provide policy advice, the World Bank researchprogram in trade works for client countries all over the world, and the program continues to beambitious.

There is still a lot to be done. Poverty remains high—with 2.7 billion people living on less than$2 dollars per day—and the global trading system is still riddled with obstacles that prevent theproducts of the world’s poor from reaching markets. Pursuing these challenges through multilat-eral, unilateral, and regional policies can contribute to poverty reduction around the world,which will pay high dividends for generations to come.

François BourguignonChief EconomistWorld Bank

November 2004


viii

THIS REPORT WAS produced by staff from the World Bank’s Prospects Group and from theWorld Bank Trade Department. Richard Newfarmer was the lead author and manager ofthe report, under the direction of Uri Dadush. The principal authors of chapter 1 were

Andrew Burns and Dominique van der Mensbrugghe. Chapter 2 was written by Paul Brenton,Carolina Diaz Bonilla, Denis Medvedev, and Philip Schuler; chapters 3 and 4 were written by PaulBrenton; chapter 5 was written by Richard Newfarmer; and chapter 6 was written by BernardHoekman, Jeffrey Lewis, and Dominique van der Mensbrugghe. Carolina Diaz Bonilla and DenisMedvedev provided research assistance that informed all chapters. The Global Outlook—the on-line twin publication of Global Economic Prospects—was led by Fernando Martel Garcia. Severalreviewers, including Michael Finger, Julio Nogues, Guillermo Perry, and Jeffrey Schott, as well asAlan Winters and Jaime De Melo, offered extensive and enormously useful comments. The reportwas produced under the general guidance of François Bourguignon.

Several people contributed substantively to the various chapters. In chapter 1, the GlobalTrends Team, under Hans Timmer’s leadership, was responsible for the projections, with contri-butions from John Baffes, Maurizio Bussolo, Betty Dow, Himmat Kalsi, Robert Keyfitz, AnnetteDe Kleine, Fernando Martel Garcia, Donald Mitchell, Mick Riordan, and Shane Streifel, and forthe poverty numbers, Shaohua Chen and Martin Ravallion. For chapter 2, Vlad Manole, WillMartin and Francis Ng, Sherman Robinson and Carolina Diaz Bonilla, Maurice Schiff andYanling Wang provided background papers and notes. For chapter 3, Souleymane Coulibaly pro-vided statistical analysis for the gravity model work, and Takako Ikezuki contributed backgroundmaterial. In chapter 4, Enrique Aldaz-Carroll, Gael Raballand, and Luc de Wulf wrote back-ground papers. In chapter 5, Ximena Clark and William Shaw, Carsten Fink and PatrickReichenmiller, Howard Mann and Aaron Cosbey, Aaditya Mattoo, and Beatriz Nofal providedbackground papers. Maurice Schiff and Alan Winters wrote papers for chapter 6. The GlobalOutlook benefited from the efforts of several specialists, among them Abhijeet Dwivedi, SabeeraKulkarni, Bulent Ozbilgin, and Carlos Rossel. Awatif H. Abuzeid was the senior staff assistant,who, together with Katherine Rollins, produced the several drafts of the report and managed thewriting stage. Dorota Nowak worked with the World Bank’s Office of the Publisher to managethe publication and dissemination process.

Many people provided written contributions and/or comments: Frederick Abbott, Hakim BenHamouda and staff at the Economic Commission for Africa, Carlos Braga, Harry Broadman,Aaron Cosbey, Robert Devlin and the staff of the Inter-American Development Bank, PeterDraper, Dorothy Dworskin and the staff at the U.S. Treasury and U.S. Trade Representative;

ix

Acknowledgments

Philip English, Antoni Estevadeordal, Alan Gelb, Caroline Freund, Mary Hallward-Driemeier,Lawrence Hinkle, Elena Ianchovichina, Roumeen Islam, Alejandro Jara, Sebastien Jean, StephenKaringi, Odin Knudsen, Hans Peter Lankes, Patrick Low, Nuno Limao, Muthukumara Mani,Howard Mann, Daniel Muller-Jentsch, Mustafa Nabli, Caglar Ozden, John Panzer, CarmenPont-Vieira, Jean-Francois Rischard, Kamal Saggi, Maurice Schiff, Raj Soopramanien, Peter VanDen Heuvel and staff of the EU Commission, Stephen Woolcock, and Gianni Zanini.


x

THE PROLIFERATION OF regional tradeagreements (RTAs) is fundamentally al-tering the world trade landscape. The

number of agreements in force now surpasses200, and it has risen sixfold in just twodecades. Today more than one-third of globaltrade takes place between countries that havesome form of reciprocal RTA.1 The EuropeanUnion (EU) and United States are playing aprominent role in this proliferation (figure 1).

This report addresses two questions:

• What are the characteristics of agree-ments that strongly promote—orhinder—development for member coun-tries?

• Does the proliferation of agreementspose risks to the multilateral tradingsystem, and how can those risks bemanaged?

Identifying What Works: OpenRegionalism

RTAs are often one component of a largerpolitical effort to deepen economic rela-

tions with neighboring countries.2 As such,they can create opportunities to expand tradethrough joint action to overcome institutionalas well as policy barriers to trade. At a basiclevel, it is often easier to motivate reciprocal re-ductions in border barriers when the partici-pants are fewer and the policymakers feel more

in control of outcomes. Moreover, RTAs havethe flexibility to pursue trade-expandingpolicies not addressed well in multilateral trad-ing rules. Trade agreements therefore usuallygo beyond slashing tariffs to include measuresto reduce trade impediments associated withstandards, customs and border crossings, andservices regulations—as well as broader rulesthat improve the overall investment climate.Finally, these agreements often form corner-stones of larger economic and political effortsto increase regional cooperation. RTAs canhelp motivate and reinforce broader reforms indomestic policy; they can be designed to con-tribute to a political environment that is moreconducive to stability, investment, and growth.

Not all agreements create new trade andinvestment. Those RTAs with high external bor-der protection are particularly susceptible to theadverse effects of trade diversion (figure 2). Infact, a statistical analysis based on findings fromseveral econometric studies suggests that manyagreements cost the economy more in lost traderevenues than they earn, because they discrimi-nate against efficient, low-cost suppliers in non-member countries. Of course, this finding doesnot take into account the potential dynamicgains, the positive effects associated with ser-vices liberalization, or any of the benefits fromadopting new regulations. But it does under-score the point that regional agreements carryrisks that merit close scrutiny by would-beparticipants.

xi

Overview

As agreements proliferate, a single countryoften becomes a member of several differentagreements. The average African countrybelongs to four different agreements, and the

average Latin America country belongs toseven agreements. This creates a “spaghettibowl” of overlapping arrangements (figure 3).Each agreement has different rules of origin,


xii

Figure 1 Regionalism spreads

a. The number of RTAs exploded in the 1990s

Annual numbers Percent of world trade coveredTotal numbers

b. EU and U.S. agreements were most important

a. EU-25 counted as single country.b. EU-15 counted as single country.

Source: World Trade Organization.

1958 1989198419761969 1994 200419990

5

10

15

20

25

30

40

Cumulative number of agreementsb

(right axis)

Agreements notnotified to the WTO

(left axis)

Agreementsnotifiedto the WTO(left axis)

300

350

250

200

150

100

50

0

Cumulative number of agreementsa

(right axis)

1990 1996 20020

15

10

5

25

20

30

35

U.S.

South-South

EU

Figure 2 External protection can differ among agreements

a. Only a few agreements have low external tariffs b. And so intraregional trade in some cases grows at theexpense of extraregional trade

0 5 10 15 20 25

SAPTA

COMESA

ECOWAS

EAC

MERCOSUR

AFTA

NAFTA

SADC

Note: In chart a, tariffs are import-weighted at the country level to arrive at RTA averages. In chart b, the bars show the magnitude ofthe dummy variables, capturing respectively the extent to which intraregional trade, overall imports, and overall exports differ from the“normal” levels predicted by the gravity model on the basis of economic size, proximity, and relevant institutional and historicalvariables, such as a common language.

Source: World Bank staff using UN TRAINS, accessed through WITS.

�4 �2 0 2 4 6 8

Estimated exponential impact on tradeAverage weighted tariffs

10

GCCAFTA

Overallexports

Overall importsIntraregional trade

CANCEMAC

MERCOSURSADC

SAPTACIS

EACWAEMU

COMESA

CACMECOWAS

EUNAFTASACU

different tariff schedules, and different periodsof implementation, and together they compli-cate customs administration. Customs agentsreport that it takes longer to process goodscovered by preferential arrangements, andlonger processing times drive up the cost oftrade. In general, the longer the delays in cus-toms, the smaller the role of trade in GDP.

So what characteristics lead to expandedtrade and development? A prerequisite forthe success of any trade policy is that it beintegrated into a sound domestic policy frame-work. It is virtually impossible for entrepre-neurs to take advantage of new opportunities—whether they originate in market accessthrough an RTA, through a multilateral agree-ment, or other sources—if the domestic invest-ment climate is not supportive. Macroeco-nomic stability, basic property rights, andadequate infrastructure regulation are all key.Indeed, trade agreements can reinforce positiveelements in the domestic reform program byanchoring policy to the agreement itself. Butan RTA cannot substitute for sound domesticpolicies.

With prerequisites in place, the RTAs mostlikely to increase national incomes over timeare those designed with:

• Low external MFN tariffs, • Few sectoral and product exemptions, • Nonrestrictive rules-of-origin tests that

build toward a framework common tomany agreements,

• Measures to facilitate trade,• Large ex-post markets,• Measures to promote new cross-border

competition, particularly in services, and• Rules governing investment and intellec-

tual property that are appropriate to thedevelopment context.

Low external tariffs and wide coverageminimize the risks of trade diversion, whilenonrestrictive rules of origin allow for in-creased trade. The practice of excluding manyagricultural products is common, and it canlimit development payoffs. Trade facilitationmeasures, though worthwhile in and of them-selves, receive more policymaker attention

O V E R V I E W

xiii

Figure 3 RTAs can complicate customs administration

a. African agreements are overlapping b. More efficient customs are associated with more trade

Ratio of trade to GDP (percent)

AlgeriaLibyaMoroccoMauritaniaTunisia

AMU

GhanaNigeria

Cape VerdeGambia

ECOWAS

BeninTogoCôte d’Ivoire

NigerBurkina Faso

Conseil deL’Entente

Guinea-Bissau MaliSenegal

WAEMU

LiberiaSierra Leone Guinea

Mano River Union

CILSS

CameroonCentral African Rep.GabonEquat. GuineaRep. Congo

Chad

Sao Tomé & Principe

ECCASCEMAC

Angola

BurundiRwanda

Egypt

DR Congo

DjiboutiEthiopiaEritreaSudan

KenyaUganda

Somalia

Tanzania

EAC

South AfricaBotswanaLesotho

NamibiaSwaziland

Mozambique

SACU

MalawiZambiaZimbabwe

MauritiusSyechelles

ComorosMadagascar

Reunion

IOCCBI

SADC

COMESA

Nile RiverBasin IGAD

Sources: Chart a, Schiff and Winters (2003). Chart b, Investment Climate Surveys data and Global Trends as cited in Subramanian andothers (2003).

00 5 10

Days through customs (imports)

R2 � �0.338

15 20 25

50

150

100

250

200

when they are embedded in an RTA, and theyoften have positive trade-creating effects forall trade partners.

Well designed agreements are of limitedvalue if they are not implemented, and manyRTAs have more life on paper than in reality.Weak implementation often afflicts South-South agreements. Monitoring mechanismsare often inadequate and do not receive thesustained high-level political attention neces-sary to drive institutional improvements in,for example, adherence to tariff reductionschedules, customs, and border crossings.

Against these benchmarks of success, it isdifficult to give universally high marks to anysingle category of agreement. In general,North-South agreements score better on im-plementation than South-South agreements.Because North-South agreements can inte-grate economies with distinct technologicalcapabilities and other different factor propor-tions, and because they usually result in largerpost-agreement markets, the potential gainsare usually greater. However, tighter rules oforigin, more restrictive exclusions for particu-lar sectors (such as agriculture), and a preoc-cupation with rules not calibrated to develop-ment priorities can undercut these benefits(figure 4). North-South agreements, particu-larly those with the United States, have beenmore effective in locking in new services liber-alization; they have pressed intellectual prop-erty rights beyond World Trade Organization(WTO) rules; and expanded the sphere of in-vestment protections; but they contain fewprovisions to liberalize the temporary move-ment of labor.

Some South-South agreements are betterat focusing on merchandise trade, minimiz-ing exclusions, adopting less restrictiverules of origin, and lowering the border costs.For example, the Caribbean Community(CARICOM) and the Common Market ofEastern and Southern Africa (COMESA) havehad some success in reducing border costs. Butin general, South-South agreements have notadhered to implementation schedules, andthey suffer from their small market size and

economic similarity. And like the North-Southagreements, South-South agreements rarelyprovide for the temporary movement of labor.

Consequences for the MultilateralSystem

The development consequences of RTAs arenot limited to their effects on members—

they also have cumulative effects on the multi-lateral system. In one sense, RTAs are a step to-ward greater openness in the whole system, bypromoting more trade and generating new do-mestic constituencies with an interest in open-ness. Moreover, some regional trade policiesare effectively nondiscriminatory, such as mea-sures to improve customs, speed transactions atports or border crossings, or in some cases openservices markets. These measures can comple-ment unilateral and multilateral policies.

However, this view overlooks the effectsthat RTAs can have on excluded countries.Preferences for some countries mean discrimi-nation against others. Indeed, the GeneralAgreement on Tariffs and Trade (GATT),borne out of the sad experience of discrimina-tion in the prewar years, was founded on the


xiv

Index of restrictiveness

Note: Higher values of the index equals to more restrictiverules of origin derived from Estevadeordal and Suominen(2004).

NAFTA

EU-Mex

ico

EU-Chil

e

SADC

Chile-

CACMAFTA

COMESA

ECOWAS

Figure 4 Rules of origin in North-Southagreements are more restrictive than inSouth-South agreements

0

2

1

3

5

4

6

principle of nondiscrimination. Today, theadverse consequences for the excluded coun-tries are much less severe than at GATT’sinception, because tariffs and other barriershave come down sharply, mitigating the ex-clusionary effects of regional arrangements.The exception—and it is not trivial—is agri-culture. Another mitigating factor is thatmany countries excluded by trade agreementsbetween the United States and the EU enjoysome degree of preferential access throughvoluntary preference schemes, such as theGeneralized System of Preferences (GSP),America’s Growth and Opportunity Act(AGOA), and the EU’s Everything But Arms(EBA) program. To be sure, these programslack the certainty of market access that MFNagreements and RTAs provide, because prefer-ences are voluntary and subject to politicalwhim, but they do mitigate the effects of ex-clusions for selected, very low-income coun-tries. Finally, some developing countries—thespokes in the hub-and-spoke analogy—aresigning bilateral agreements with each otherand with other hubs.

Inevitably some countries get left out oftrade agreements, either because they are notfavored politically, because they cannot affordthe costs of many separate negotiations, or be-cause their neighborhood is less open. Coun-tries as diverse as Bolivia, India, Mongolia,Pakistan, and Sri Lanka do not enjoy the samelevel of access to the United States or the EUas Chile, Jordan, or Mexico, and they see theirtrade diminished when bilateral agreementsare signed.

RTAs can also undercut the incentives ofgovernments to press for multilateral liberal-ization, which would improve global traderules. This study finds little evidence thatmajor players in the current WTO negotia-tions have changed their negotiating positionsor retreated from the multilateral process,even as they avail themselves of regional tradedeals. However, as the discussions becomepolitically difficult, the risk is ever present thateven they will abandon multilateralism infavor of “satisficing regionalism.” One

consequence of the spread of regional agree-ments is that many poorer developing coun-tries have diverted scarce negotiating re-sources to regional negotiations at the expenseof more active participation in the Doha dis-cussions. The average developing country be-longs to five separate RTAs and is negotiatingmore all the time. In the future, will countriesthat now enjoy preferences fight multilateralliberalization, or even oppose further regionalliberalization, to keep their privileged marketaccess? A few small developing countries areindeed likely to lose advantages in preferentialmarkets, and they may scuttle a deal if their le-gitimate concerns are not addressed.

The Importance of Doha to OpenRegionalism

The policy solution to these twin con-cerns—the need to design regional agree-

ments that create trade and regional agree-ments that have minimal exclusionaryeffects—comes together in the form of lowMFN tariffs and other border barriers. Anagreement that lowers border protectionaround the world promotes open regionalismby mitigating trade diversion. At the sametime, it would diminish the exclusionary ef-fects of discriminatory preferences built intoregional agreements. The first order of busi-ness for the international community is to ac-celerate progress on the Doha Agenda and tofill in the blanks of the August 2004 frame-work agreement with reductions in protec-tion, especially for products produced by theworld’s poor.

For Developing Countries,a Three-Part Strategy

Developing countries wishing to harnesstrade to their development strategy

should see regional integration as one elementin a three-pronged strategy that includes unilat-eral liberalization, multilateral liberalization,and regional liberalization.

O V E R V I E W

xv

Historically, unilateral liberalization, whichis usually linked to a broader program of do-mestic reform, has accounted for most of thereductions in border protection. Most com-prehensive trade reforms among large coun-tries (Argentina, Brazil, and China in the early1990s, and more recently, India) were primar-ily unilateral reforms that were undertaken toincrease the productivity of the domestic econ-omy. The same process took place in manysmall countries as well. In fact, of the 21 per-centage point cuts in average weighted tariffsof all developing countries between 1983 and2003, unilateral reforms account for roughlytwo-thirds of the reduction. Tariff reductionsassociated with the multilateral commitmentsin the Uruguay Round accounted for about25 percent, and the proliferation of regionalagreements amounted to about 10 percent ofthis reduction (see figure 5).

Autonomous liberalization promotes globalcompetitiveness by lowering costs of inputs, in-creasing competition from imports to drive pro-ductivity growth, and integrating the nationaleconomy into the global economy. Autonomoustrade reform is, ironically, more important thanever in the presence of RTAs; low border barri-ers minimize the risks of trade and investmentdiversion. Low external barriers promote tradein world markets, and this is highly correlated

with increases in intraregional trade, irrespec-tive of the presence of an RTA.

Multilateral liberalization leverages domes-tic reforms into increased market accessaround the world. Developing countries col-lectively stand to gain much more in the WTOarena than in any smaller regional market.Moreover, this multilateral forum is the onlyplace that developing countries, working to-gether, can press for more open markets inagriculture and can seek disciplines on trade-distorting agricultural subsidies and on con-tingent protection.

Some have argued that RTAs can be an al-ternative to multilateral liberalization. Theyare not. Gains for all developing countriesfrom these agreements, even under the mostgenerous of assumptions, are usually only afraction of those from full multilateral liberal-ization. Of course, if one of the partner coun-tries is a high-income, large-market economy,and if most other countries are excluded frompreferential access, the countries signing thefirst trade agreement may benefit individuallyand substantially—but those benefits witheras new countries sign additional agreements.In fact, the scenarios in this study show thatall developing countries would collectivelylose if they were all to sign preferential agree-ments with the Quad (Canada, the EU, Japan,and the United States) (figure 6). Therefore,developing countries have a powerful collec-tive interest in an effective Doha Agenda—even if they all are scrambling to gain prefer-ential market access to the Quad.

Forging policies on open regionalism is thethird component of trade policy strategy. De-sirable as multilateral liberalization is, theDoha Round is likely to realize only part of itsdevelopment potential. For some types of pol-icy, collective regional actions may be the first,best course, and may result in effective nondis-criminatory benefits.3 For example, RTAs canreduce regional political tensions, take advan-tage of scale economies in infrastructure pro-vision, and lead to joint programs to improveborder crossings or to motivate liberalizationin services. But countries should sign on with


xvi

Figure 5 Share of total tariff reduction, bytype of liberalization, 1983–2003

Autonomous Liberalization66%

MultilateralAgreements

25%

Regional Agreements

10%

Source: Martin and Ng 2004.

their eyes wide open. The lessons of this study(and others before it4) are that, much as withunilateral or multilateral policies, design andimplementation determine the ultimate effects.It is important to use trade policy to leveragedomestic reforms that promote growth. ForSouth-South agreements, it is essential that thefocus be on some combination of full tradeliberalization behind low external borderprotection, greater services deregulation andcompetition, and proactive trade facilitationmeasures that together positively affect bothintra- and extra-regional trade.

High-Income Countries andDevelopment

High-income countries, in order to realizetheir broad development objectives, must

intensify their efforts to realize the develop-ment promise of the Doha Agenda. This hasthe potential to open up trade, particularly inagriculture, in a way that would benefit low-income groups around the world. Becausethe high-income countries are the large

players in the system, they have a special inter-est in—and responsibility for—using effectivemultilateral reforms to discipline the discre-tionary aspects of the regional agreements.

Allowing developing countries to concen-trate scarce negotiating resources on the mul-tilateral agenda may require that high-incomecountries decelerate their efforts at expandingRTAs. Irrespective of the pace of new agree-ments, high-income countries could considerthe following rules of thumb when designingagreements to promote development. First, re-ducing the extensive exclusions for agriculturewould transfer the income gains to rural areasin participating developing countries. Second,adopting more common and nonrestrictiverules of origin across agreements wouldreduce the administrative barriers that oftenundermine agreements and that increase theburden on customs administration. Third,working with prospective partners to ensurethat new regulations regarding investment andintellectual property are appropriate to thelevel of development would reduce risks ofundue enforcement costs. Finally, providingtrade-related technical assistance, not only inthe implementation phase but also in the ne-gotiating phase, would promote greater liber-alization of services and lower MFN tariffs.

Acting Collectively to Mute theEffects of Discrimination

To minimize the discriminatory effects ofRTAs at the multilateral level, all countries

must assume greater responsibility for main-taining the multilateral system. The interna-tional community, working through the WTO,should revisit Article V of its charter. If thestated disciplines cannot be enforced in thenear term for collective political reasons, thenincreasing transparency and informationshould become a priority. At present, the WTOcollects little if any information updating spe-cific provisions, their implementation, and thetrade consequences. It even fails to take ad-vantage of extant public monitoring efforts in

O V E R V I E W

xvii

�1.5 �1.0 �0.5 0 0.5

Percent

1.0 1.5

Note: Global refers to the global merchandise trade reformscenario; JBIL corresponds to the simulation where alldeveloping countries sign bilateral agreements with theQuad-plus countries; and BILAT corresponds to thesimulation where the bilateral agreements are signedindividually. Results reflect unweighted regional averages.

Source: World Bank simulations with the Linkage model andGTAP release 6.04.

2.0

Figure 6 Multilateral liberalization is farmore beneficial than RTAs

Quad�

Low-income

Middle-income

Developing

Change in real income in 2015 compared to baseline

Global

BILATJBIL

specific regions, which could inform their datacollection effort. Collecting and publishingspecific information on RTAs would allowmembers that find themselves excluded tochallenge these agreements in the court of pub-lic opinion. Even the more modest goal oftransparency will require building a new con-sensus and providing the staff of the WTOwith more resources than they have currentlyavailable.

Nonetheless, WTO members should con-sider enhancing the existing rules to ensurethat regional agreements have positive devel-opment and systemic outcomes. This could in-clude (based on a modest tightening of currentpractice) setting quantitative indicators thatdefine “substantially all trade.” It could in-clude efforts to simplify and harmonize therules of origin that are applied to both devel-oped and developing countries. These itemsare on the Doha Agenda and may be ready foraction.

Organization of This Study

As is customary, chapter 1 of this study pre-sents the World Bank’s view of the global

economy. The short-term section analyzes themain forces shaping the global outlook and theimplications for developing countries; the long-term analysis focuses on structural changesin the global economy that will affect povertyrates and the prospects for attaining the Millen-nium Development Goals. A novel feature ofthis year’s report is the introduction of a com-panion online feature (see www.worldbank.org/prospects), where the reader can find addi-tional information on regional trends and com-modity prices, and tools to design scenarios tohis or her own specifications.

Chapter 2 introduces the issues associatedwith regional trade agreements and providesan overview of regional trading trends.Subsequent chapters focus on the content and

consequences of regional agreements for tradecreation (chapter 3), trade facilitation(chapter 4), and services, investment, intellec-tual property rights, and labor mobility(chapter 5). Chapter 6 returns to the issue ofmaking regional agreements more compatiblewith a nondiscriminatory multilateral system.

Notes1. Negotiated as bilateral or multicountry treaties,

regional trade agreements grant members assured pref-erential market access, usually at zero tariffs for eligi-ble products. Following WTO convention, the term“regional trade agreement” includes both reciprocalbilateral free trade or customs areas and multicountry(plurilateral) agreements. These are distinct from non-reciprocal voluntary agreements, such as the general-ized system of preferences (GSP). Also, for statisticalpurposes, unless otherwise noted, intra-EU trade is ex-cluded from quantitative trade analysis. The EU is de-fined as including the 15 countries that belonged to theunion before its enlargement in 2004.

2. See Devlin and Estevadeordal (2004) and Schiffand Winters (2003), among others.

3. See Robert Lawrence (1997), who develops theidea of subsidiarity as applied to regional agreements.

4. See Schiff and Winters (2003).

ReferencesDevlin, Robert and Antoni Estevadeordal. 2004. Trade

and Cooperation: A Regional Public Goods Ap-proach. In Regional Public Goods: From Theory toPractice, eds. A. Estevadeordal, Brian Frantz,and Tam Robert Nguyen. Washington, DC: Inter-American Development Bank.

Lawrence, Robert. 1997. Preferential Trading Arrange-ments: The Traditional and the New. In RegionalPartners in Global Markets: Limits and Possibili-ties of the Euro-Med Agreements, eds. A. Galaland B. Hoekman. London: Center for EconomicPolicy Research/Egyptian Center for EconomicStudies.

Martin, W., and F. Ng. 2004. Sources of Tariff Reduc-tions. Background paper.

Schiff, Maurice, and L. Alan Winters. 2003. RegionalIntegration and Development. Washington, DC:World Bank.


xviii

ACP African Caribbean and Pacific states

ACPEU African Caribbean and Pacific states European Union

AFTA ASEAN Free Trade Area

AGOA African Growth and Opportunity Act

ANZCERTA Australia-New Zealand Closer Economic Relations Trade Agreement

APEC Asia Pacific Economic Cooperation

ASEAN Association of Southeast Asian Nations

BITS Bilateral investment treaties

CAFTA Central America Free Trade Agreement

CARICOM Caribbean Community

CEC Commission for Environmental Cooperation

CEMAC Economic and Monetary Community of Central Africa

CEPR Center for Economic and Policy Research

CGE Computable general equilibrium

CIS Commonwealth of Independent States

COMESA Common Market for Eastern and Southern Africa

CRTA Committee on Regional Trade Agreement

EAC East African Community

EBA Everything but arms

EC European Community

ECO Economic Cooperation Organization

ECOWAS Economic Community of West African States

EEC European Economic Community

EFTA European Free Trade Association

EPAs Economic Partnership Agreements

EU European Union

FDI Foreign direct investment

xix

Abbreviations

FTAA Free Trade Area of the Americas

GAO General Accounting Office

GATS General Agreement on Trade in Services

GATT General Agreement on Tariffs and Trade

GCC Gulf Cooperation Council

GDP Gross domestic product

GSP Generalized System of Preferences

GTAP Global Trade Analysis Project

HS Harmonized system

IDB Inter-American Development Bank

IFC International Finance Corporation

IISD International Institute for Sustainable Development

IMF International Monetary Fund

INS Immigration and Naturalization Services

IOM International Organization for Migration

IPR Intellectual Property Rights

IRCA Immigration and Regularization Control Act

IRPA Immigration and Refugee Protection Act

LAFTA Latin America Free Trade Area

LDCs Least developed countries

MDGs Millennium Development Goals

MERCOSUR Southern Lone Common Market

MFN Most favored nation

MRA Mutual recognition agreement

NAFTA North America Free Trade Agreement

NBER National Bureau of Economic Research

OECD Organisation for Economic Co-operation and Development

PRSP Poverty Reduction Strategy Paper

PTAs Preferential trade agreements

RTAs Regional trade agreements

SAARC South Asia Association for Regional Cooperation

SACU South African Customs Union

SAD Single administrative document

SADC Southern African Development Community

SAPP Southern African Power Pool

SAFTA South Asian Free Trade Area

SAPTA South Asia Preferential Trade Agreement

SPS Sanitary and phyto-sanitary standards


xx

TBT Technical barriers to trade

TFP Total factor productivity

TRIM Trade-related investment measures

TRIPS Trade-related aspects of intellectual property rights

TTF Transport and trade facilitation

UEMOA/WAEMU West African Economic and Monetary Union

UNCTAD United Nations Conference for Trade and Development

UNESCAP United Nations Economic and Social Commission for Asia and the Pacific

USAID United States Agency for International Development

USTR United States Trade Representative

WCO World Customs Organization

WIPO World Intellectual Property Organization

WTO World Trade Organization

A B B R E V I A T I O N S

xxi

Agreement Full name Members

AFTA ASEAN Free Brunei, Darussalam, Cambodia, Indonesia, Lao Trade Area People’s Democratic Republic, Malaysia, Myanmar,

Philippines, Singapore, Thailand, Vietnam

APEC Asia Pacific Economic Australia, Brunei, Canada, Chile, China, Cooperation Hong Kong (China), Indonesia, Japan, Korea,

Malaysia, Mexico, New Zealand, Papua NewGuinea, Peru, Philippines, Russia, Singapore, Taiwan(China), Thailand, United States, Vietnam

CACM Central American Costa Rica, El Salvador, Guatemala, Honduras, Common Market Nicaragua

CAFTA Central America Free United States, Costa Rica, El Salvador, Guatemala, Trade Area Honduras, Nicaragua, Dominican Republic

CAN Andean Community Bolivia, Colombia, Ecuador, Peru, RepúblicaBolivariana de Venezuela

CARICOM Caribbean Community Antigua and Barbuda, Bahamas, Barbados, Belize,and Common Market Dominica, Grenada, Guyana, Haiti, Jamaica,

Monserrat, Trinidad and Tobago, St. Kitts andNevis, St. Lucia, St. Vincent and the Grenadines,Suriname

CEFTA Central European Bulgaria, Czech Republic, Hungary, Poland, Free Trade Agreement Romania, Slovak Republic, Slovenia

CEMAC Economic and Monetary Cameroon, Central African Republic, Chad, Community of Republic of Congo, Equatorial Guinea, GabonCentral Africa

CER Closer Economic Australia, New Zealand(ANZCERTA) Relations Trade

Agreement

xxiii

Frequently Cited Regional Trading Agreements and theParties to Them

Agreement Full name Members

CIS Commonwealth of Azerbaijan, Armenia, Belarus, Georgia, Moldova, Independent States Kazakhstan, Russian Federation, Ukraine,

Uzbekistan, Tajikistan, Kyrgyz Republic

COMESA Common Market for Angola, Burundi, Comoros, Democratic Republic of Eastern and Congo, Djibouti, Arab Republic of Egypt, Eritrea, Southern Africa Ethiopia, Kenya, Madagascar, Malawi, Mauritius,

Namibia, Rwanda, Seychelles, Sudan, Swaziland,Uganda, Zambia, Zimbabwe

EAC East African Community Kenya, Tanzania, Uganda

ECOWAS Economic Community Benin, Burkina Faso, Cape Verde, Gambia, Ghana, of West African States Guinea, Guinea-Bissau, Côte d’Ivoire, Liberia, Mali,

Niger, Nigeria, Senegal, Sierra Leone, Togo

EEA European EU, Iceland, Liechtenstein, NorwayEconomic Area

EFTA European Free Iceland, Liechtenstein, Norway, SwitzerlandTrade Association

EMFTA Euro-Mediterranean EU, Algeria, Cyprus, Egypt, Israel, Jordan, Lebanon, Free Trade Area Malta, Morocco, Syrian Arab Republic, Tunisia,

Turkey, Palestinian Authority

FTAA Free Trade Area Antigua and Barbuda, Argentina, Bahamas, of the Americas Barbados, Belice, Bolivia, Brazil, Canada, Chile,

Colombia, Costa Rica, Dominica, DominicanRepublic, Ecuador, El Salvador, Grenada,Guatemala, Guyana, Haiti, Honduras, Jamaica,Mexico, Nicaragua, Panama, Paraguay, Peru, St. Kitts and Nevis, St. Lucia, St. Vincent and theGrenadines, Suriname, Trinidad and Tobago, United States, Uruguay, Venezuela

GAFTA Greater Arab Free Bahrain, Egypt, Iraq, Jordan, Kuwait, Lebanon, Trade Area Libya, Morocco, Oman, Palestine, Qatar, Saudi

Arabia, Somalia, Sudan, Syria, Tunisia, United Arab Emirates, Yemen

GCC Gulf Cooperation Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, Council United Arab Emirates

MERCOSUR Southern Common Argentina, Brazil, Paraguay, UruguayMarket

NAFTA North American Canada, Mexico, United StatesFree Trade Agreement

SACU Southern African South Africa, Botswana, Lesotho, Swaziland, Customs Union Namibia


xxiv

SADC Southern African Angola, Botswana, Democratic Republic of Congo, Development Lesotho, Malawi, Mauritius, Mozambique, Community Namibia, South Africa, Swaziland, Seychelles,

Tanzania, Zambia, ZimbabweSAFTA South Asian Free Trade Bangladesh, Bhutan, India, Maldives, Nepal,

Area Pakistan, Sri Lanka

SAPTA South Asian Preferential Bangladesh, Bhutan, India, Maldives, Nepal, Trade Arrangement Pakistan, Sri Lanka

WAEMU West African Economic Benin, Burkina Faso, Côte d’Ivoire, Guinea Bissau, and Monetary Union Mali, Niger, Senegal, Togo

R E G I O N A L T R A D I N G A G R E E M E N T S A N D T H E P A R T I E S T O T H E M

xxv

World growth accelerated sharply in 2004,with GDP advancing an estimated 4 percent(table 1.1). All developing regions are nowgrowing faster than their average growth ratesof the 1980s and 1990s. The ongoing eco-nomic boom in China was a major factor, aswere the surges in activity registered in Japanand the United States. The economic recoverywas slower to take hold among Europeanhigh-income countries, which contributed tothe less marked increase in growth rates there.Meanwhile, very strong import demand—because of the torrid expansion in China andthe continued tendency for domestic demandin the United States to substantially exceedproduction—contributed to an exceptional10.2 percent increase in world trade volumes.

Economic growth is expected to slow in2005 and 2006, expanding by 3.2 percent ineach year. Several factors are likely to con-tribute to this more moderate pace of activity.First, the investment cycle in the United Stateshas likely peaked, implying a slowdown ingrowth there.1 Second, world demand hasoutstripped supply, resulting in substantial in-creases in oil and other commodity prices thathave cut into incomes, moderating demand inmany countries. Third, higher interest rateswill slow investment growth as central bankscontinue shifting monetary policy from aloose to a more neutral stance. Fourth, thelarge fiscal impulse that has helped propel theU.S. economy in recent years will weaken in2004—although the deficit will remain high;

and in Europe, budgetary policy is expected totighten as countries seek to regain controlover deficits, which in many cases exceedMaastricht limits. Finally, efforts in China tobring growth down to a more sustainable paceshould also contribute to weaker, but stillstrong, demand over the medium term.

Given this external environment and espe-cially the less rapid expansion of trade,growth in most low- and middle-income coun-tries is also expected to moderate but remainstrong. The extent of the slowdown should bemitigated because of the far-reaching struc-tural reforms carried out in many countries,which have contributed to recent gains in mar-ket share and economic growth. Recent effortsto reduce general government and current ac-count deficits and to pay down debt shouldenable most developing countries to withstandthe higher interest rates expected over the nextfew years without excessive adjustmentcosts. However, there is little room for com-placency—especially for the more highlyindebted countries.

These favorable prospects for the next twoyears represent a solid starting point forlonger-term growth through 2015 and increasethe likelihood that developing countries meetthe Millennium Development Goals (MDGs).Improvements in macroeconomic fundamen-tals, enhanced structural flexibility, a strongerinvestment climate, and further progress to-ward reducing trade barriers should, ifsustained, support the ability of developing

1

Global Outlook and theDeveloping Countries

1

Table 1.1 The global outlook in summaryPercentage change from previous year, except interest rates and oil prices

Forecast

2002 2003 2004e 2005 2006

Global ConditionsWorld Trade Volume 3.7 5.5 10.2 8.4 7.8Consumer Prices

G-7 Countriesa,b 1.0 1.6 1.7 1.4 1.2United States 1.6 2.3 2.7 2.2 1.7

Commodity Prices (USD terms)Non-oil commodities 5.3 10.2 17.0 �3.1 �4.2

Oil Price (World Bank average)c 24.9 28.9 39.0 36.0 32.0Oil price (percent change) 2.4 15.9 35.0 �7.7 �11.1

Manufactures unit export valued �1.3 7.4 5.2 �0.8 �0.3Interest Rates

$, 6-month (percent) 1.8 1.2 1.6 3.5 4.7€, 6-month (percent) 3.3 2.3 2.1 2.4 3.6

Real GDP growthe

World 1.7 2.7 4.0 3.2 3.2Memo item: World (PPP weights)f 2.9 3.9 4.9 4.2 4.1

High income 1.3 2.1 3.5 2.7 2.7OECD Countriesg 1.3 2.0 3.5 2.6 2.6Euro Area 0.9 0.5 1.8 2.1 2.3Japan �0.3 2.5 4.3 1.8 1.6United States 1.9 3.0 4.3 3.2 3.3Non-OECD countries 2.2 3.1 5.9 4.6 4.4

Developing countries 3.4 5.2 6.1 5.4 5.1East Asia and Pacific 6.7 7.9 7.8 7.1 6.6Europe and Central Asia 4.6 5.9 7.0 5.6 5.0Latin America and the Caribbean �0.6 1.6 4.7 3.7 3.7Middle East and North Africa 3.2 5.7 4.7 4.7 4.5South Asia 4.6 7.5 6.0 6.3 6.0Sub-Saharan Africa 3.1 3.0 3.2 3.6 3.7

Memorandum itemsDeveloping countries

excluding transition countries 3.2 5.1 5.9 5.4 5.1excluding China and India 2.1 3.8 5.4 4.6 4.3

Note: PPP � purchasing power parity; e � estimate.a. Canada, France, Germany, Italy, Japan, the United Kingdom, and the United States.b. In local currency, aggregated using 1995 GDP weights.c. The World Bank average is the unweighted mean of one barrel of West Texas Intermediate, Brent, and Dubai oil.d. Unit value index of manufactured exports from major economies, expressed in U.S. dollars.e. GDP in constant dollars at 1995 prices and market exchange rates.f. GDP measured at 1995 PPP weights.g. Now excludes the Republic of Korea, which has been reclassified as high-income OECD. Source: World Bank.

countries to achieve rapid and sustained percapita growth at a level of 3.5 percent perannum between 2006 and 2015—double thegrowth rate of the 1990s. Such growth wouldenable many developing countries to halve theincidence of extreme poverty by 2015, which isa key development goal. However, even if thehigher growth of recent periods were sus-tained, some regions, notably Sub-Saharan

Africa, will fail to reduce poverty to thisdegree. In Sub-Saharan Africa, per capitagrowth has been slow, and progress to reducepoverty has been minimal. It would take im-plausibly high growth rates during the next10 years to achieve the poverty target alongwith substantial enhancements to pro-poorpolicies and significantly more assistance.Finally, even if many regions are expected to


2

achieve the MDG to reduce poverty, many areoff track for reaching other important MDGs,such as reducing child and maternal mortality.In many cases economic growth is not enough.A more targeted approach and a realignmentof spending priorities are also necessary.

Despite the relatively positive picture forboth medium- and long-term prospects,downside risks are ever present and couldhave negative impacts in the near future and inthe long term. An additional rise in oil prices,or a failure of them to moderate, could furtherrestrain global demand and reduce incomes inmost less developed countries. While oil pricesare expected to decline from present highs,especially given substantial efforts to increasesupply by oil exporting countries, existingdemand conditions are such that a significantincrease cannot be ruled out. Such a risewould have important negative effects on alloil-importing economies, particularly thoseof low- and middle-income countries thatface current account constraints. For thesecountries, difficulties accessing internationalfinance mean that they cannot absorb the in-creased costs associated with higher oil pricesby increasing their current account deficit.Instead, the additional costs must be accom-modated by lower imports, consumption, andinvestment volumes—implying a significantreal-side adjustment. For the most vulnerableof such countries, an additional $10 a barrelincrease in oil prices could reduce domestic in-comes by as much as 4 percent. On average,incomes of oil-importing low-income countrieswould fall by about 1 percent of GDP.

Financing requirements of the U.S. currentaccount and government deficits, and reneweddownward pressure on the dollar, may causelong-term interest rates to rise more than fore-casted. If interest rates rise, short- to medium-term impacts might include a slowing in worldeconomic growth, sharply increased financingcosts, and economic hardship for heavily in-debted countries. Increased financial-marketturbulence might also ensue—especially forthose developing countries most exposed to theU.S. dollar. Over the medium- to long-term,

failure to rein in the U.S. budget deficit, whichwould also tend to reduce its current accountdeficit, could result in an ever increasing stockof dollar-denominated debt and rising futurefinancing burdens. Moreover, higher interestrates would depress investment levels, provok-ing a prolonged slowing in the rate of increaseof potential output. All of these factorsheighten the risk of a resurgence in protection-ist sentiment, which would thwart the pace atwhich developing countries are able to achievetheir poverty reduction objectives.

Finally, if current efforts to slow the unsus-tainable pace of growth in China fail, majordisruptions could result. Currently, investmentlevels may be unsustainably high, and thereare some signs that rapidly rising food-priceincreases are feeding into production costs,which could ultimately choke off competitive-ness, (although for the moment there are noclear indications that this is happening). Eitherproblem could provoke a much more abruptslowdown than described in the baseline.Given China’s growing importance as a driverof world trade growth, such a sharp slow-down could have a significant damping effecton global economic activity, particularlyamong China’s major trading partners.

The Global Economy: FromRecovery to Expansion

The world economy accelerated sharply in2004, expanding by an estimated 4 per-

cent (figure 1.1). The United States and Japan,whose economies grew by more than 4 percent,continued to lead Europe in the recovery. Evenstronger growth was experienced by a numberof large developing countries, notably China(8.8 percent), Russia (8.0 percent), and India(6.0 percent). Their performance helped powerdeveloping countries as a whole to an antici-pated 6.1 percent growth rate in 2004—anexpansion without precedent over the past30 years. Moreover, it marks a second year ofvery strong growth, and it may be the first timethat recovery in developing countries preceded,rather than followed, recovery in high-income

G L O B A L O U T L O O K A N D T H E D E V E L O P I N G C O U N T R I E S

3

countries. In contrast to the United States,where the surge was initially led by investmentand household consumption, exports were themain source of growth in Europe and Japan—and much of the increase in external demandcame from developing countries.

Across the developing world virtuallyevery region enjoyed solid growth, andrapidly rising trade volumes played an impor-tant role. Even excluding China, India, andRussia, economic activity in developing coun-tries is expected to have risen 5 percent in2004. While easy credit contributed toChina’s remarkable performance, the benefitsof WTO accession were also a major factor,and the increase of over 30 percent in Chineseimport demand helped underpin growthamong neighboring East Asian countries.Russia and the oil-producing countries in theMiddle East and North Africa Region bene-fited from very strong oil revenues, whichwere reflected in strong import demand andthe solid export performance of their tradingpartners. Increasing market shares, followingsubstantial inward investment flows associ-ated with the accession of many of the Europeand Central Asian Region’s members to theEU, also contributed to these positive out-comes. Elsewhere, a strong cyclical recovery

is under way in Latin America, and there aresigns of a more modest recovery in Sub-Saharan Africa.

Growth should moderate in 2005 and2006, led by a slowing of the expansion amongdeveloped countries. In the United States, asthe output gap closes, productivity growth isprojected to slow and unit labor costs to rise;these factors, in addition to external inflation-ary pressures from commodity prices, likely re-flect the Fed’s decision to tighten monetaryconditions. This, plus the maturation of the in-vestment cycle, a tailing off of fiscal stimulus,and the impact of higher oil costs, will con-tribute to slowing growth. Similar factors ex-plain the anticipated slowdown in Japan,where output is expected to increase at abouttrend rates. In contrast, because of its laterstart and the fact that investment is onlynow beginning to recover, Europe’s growth isexpected to continue gaining momentumthrough 2005 and into 2006, notwithstandingfiscal tightening and a slowdown in the rate ofgrowth of world demand. Overall estimatessuggest that the hike in oil prices already ob-served can be expected to dampen output in2005 by about 0.5 percent of GDP.

Moderating growth in the OECD economiesand a soft landing in China should translate intoslower but still buoyant growth in developingcountries (figure 1.2).

• In East Asia, efforts to stem the flow ofcredits into selected sectors of the Chi-nese economy are already having observ-able effects (figure 1.2a). The growth ofimports of raw materials such as steel,copper, and various ores have moderatedsignificantly in recent months. Steel im-ports have collapsed, although iron oreimport volumes were growing by morethan 25 percent (year/year) in September.However, there are indications thatconsumption demand continues to growrapidly, and the Chinese authoritiesreport that GDP increased 9.1 percent inthe third quarter. The baseline forecastpredicts that a soft landing (growthslowing to 7.1 percent by 2006) will be


4

1980

1983

1986

1989

1992

1995

1998

2001

2004

WorldDeveloping countries

GDP growth, 1995 US$ (percent)

Figure 1.1 A record year for developingcountries in 2004

0

3

2

1

4

5

6

Source: World Bank.


5

1980–2002 2003 2004* 2005 2006

a. East Asia and Pacific

Growth of real GDP (percent)

b. South Asia


c. Europe and Central Asia


d. Middle East and North Africa

Forecast Forecast

Forecast Forecast

Forecast Forecast


e. Latin America and the Caribbean


f. Sub-Saharan Africa


Figure 1.2 Strong growth across most regions

1

0

2

3

4

5

6

7

9

8

1

2

3

4

5

6

7

8

1

2

3

4

5

6

7

8

1980–2002 2003 2004* 2005 2006

1

0

2

3

4

5

6

7

9

8

1980–2002 2003 2004* 2005 20060

1980–2002 2003 2004* 2005 20060

1980–2002 2003 2004* 2005 2006

1.0

0.5

1.5

0

2.0

2.5

3.0

3.5

4.0

4.5

5.0

1980–2002 2003 2004* 2005 2006

1.0

0.5

1.5

0

2.0

2.5

3.0

3.5

4.0

4.5

5.0

Note: * � estimate.

Source: World Bank estimates.

achieved and will contribute to slowingthroughout the region.

• In South Asia, despite the moderation ofthe Chinese and OECD economies,growth is expected to accelerate in 2005,reflecting the enduring impacts of struc-tural reforms, market opening, andstronger domestic demand as the damp-ening impact of last year’s poor cropfades. As agricultural production and re-lated incomes return to trend growthrates in 2006, GDP growth is projectedto moderate somewhat.

• Output in Europe and Central Asia is fore-cast to remain strong, with still-high oilprices supporting demand in Russia andthe exports of its trading partners. Centraland Eastern European countries will con-tinue to benefit from rapid investmentgrowth following the EU accession ofsome of their members. However, policy-makers need to prepare for the next down-turn by pursuing fiscal consolidation to re-duce worryingly high government and, insome cases, current account deficits.

• Growth in the Middle East and NorthAfrica region is expected to remain robust,but well below the highs observed in 2003,which were boosted by sharp increases inoil production. All countries, but espe-cially those of the Maghreb, should benefitfrom the strengthening export demandemanating from Western Europe; but con-sumption demand, reflecting still high oilincomes, will continue to be the mainsource of growth for the region as a whole.

• The return to growth in Latin America andthe Caribbean is projected to continue,with only Argentina experiencing a signif-icant slowdown as the competitive advan-tage from its depreciation in 2002 wearsoff. Elsewhere, growth should remainstrong, with Brazil expanding steadily atbetween 3.7 and 3.9 percent. BecauseLatin America and the Caribbean is aheavily indebted region, outturns will ulti-mately depend on the success with whichpolicymakers deal with rising interest rates

and higher payments on debt (see the riskssection in this chapter). Here, country-specific conditions and the degree towhich fiscal consolidation programs aremaintained will play an important role.

• Sub-Saharan Africa will also benefitfrom the revival in Europe, its main trad-ing partner, but many oil-importingcountries in Africa remain vulnerabledue to high oil prices. Notwithstandingsubstantially improved performance,growth in the region will continue to lagthe rest of the world by a significant mar-gin, implying a further widening of in-come gaps. Moreover, the terms of tradeappear to be turning against this regionas non-oil commodity prices are ex-pected to ease. Although additional de-velopment aid and debt relief wouldhelp, continued efforts to improve fun-damentals and the efficiency of publicexpenditure are also required to speedthe pace at which these countries achievetheir poverty-reduction objectives.

Commodity Markets

Strong world demand and supply shortageswere responsible for commodity prices re-

bounding sharply during the global recovery(figure 1.3). In dollar terms, metals and minerals


6

Petroleum

Cumulative percent change, 2001–2004

Agriculturalproducts

Exchange-rateadjusteda

Metals andminerals

Manufactures

Figure 1.3 Tradable price developments

�10

20

10

0

40

30

50

60

Nominal

EUN

a. The adjusted data show the price change expressed as a trade-weighted average of domestic currency prices.

Source: World Bank.

prices have increased the most since 2001 (up al-most 60 percent), but the 40 percent hike inpetroleum prices has had the largest economiceffect. In domestic currency terms, the impact ofthese price hikes was less important for manycountries because of the 15 percent deprecia-tion2 of the dollar over the same period.

Higher commodity prices since 2001 haveboosted incomes of low- and middle-incomecountries as a whole by an estimated 1.1 per-cent of GDP. However, virtually all of the gainaccrued to low- and middle-income oil ex-porters. Most developing country oil im-porters suffered net terms of trade losses (fig-ure 1.4). The major beneficiaries were theMiddle East and North Africa, Europe andCentral Asia, and Latin America and theCaribbean Regions—all of which includemajor oil exporters. In contrast, the net gainsfrom non-oil commodity prices for low-income countries were modest or even nega-tive. This is partly because most of the non-oilcommodity price gains were concentratedin metals and minerals prices, which restrictedthe benefits to a few resource-rich countries.Moreover, many industrializing low-incomecountries, notably India and Pakistan, arenow net commodity importers. The terms-of-trade impact on incomes of oil exporting

developing countries was 5.6 percent of GDP,whereas for oil importers the impact was aloss of 0.3 percent.

For the poorest oil-importing countries,high oil prices have dramatically exacerbatedalready serious poverty. Many of thesecountries remain particularly vulnerable tohigh oil prices. Even before the oil price hikes,a number of these countries were spendingmore than 5 percent of GDP to cover oil im-ports. The unweighted average of West-TexasIntermediate, Brent, and Dubai crude oils isestimated to have been $39 in 2004.3 At thislevel, it is estimated that as many as sevencountries will have oil-import bills in excess of10 percent of GDP; these countries would beforced to make substantial cuts in spendingelsewhere in their economies to compensatefor the additional burden (figure 1.5). Indeed,for the poorest countries the net additionalburden in 2004 is expected to consume 75percent of the World Bank funding they re-ceive for all development programs, and


7

Percent of GDP

Oil exporters

Oil importers

Source: World Bank.

Low a

nd

midd

le inc

ome

Low in

com

e

Highly

indeb

ted

Low a

nd

midd

le inc

ome

Low in

com

e

Low in

com

e

(exc

luding

India

)

Highly

indeb

ted

Figure 1.4 Terms of trade impacts fromhigher commodity prices, 2001–04

�2

2

0

4

8

6

10

Source: World Bank.

Figure 1.5 The oil-import burden for selected countries

0 5 10 15 20 25

Mali

Nicaragua

Moldova

TajikistanBosnia and

HerzegovinaSwaziland

Mongolia

Lesotho

Mauritania

Lao PDR

Malawi

Pakistan

Burkina Faso

Kyrgyzstan

Ethiopia

Pre-hike burden

Additional burden

Percent of GDP

12 percent of all the bilateral aid they receive.To keep development projects on track, high-income countries will need to increase com-mitments substantially—at least as long ashigh oil prices continue.

Although a substantial rise in oil prices isnot the most likely scenario, given newsources of supply and reduced oil intensities inthe world economy, there remains consider-able scope for higher oil prices, particularlygiven the current sensitivity of oil markets to

localized disruptions in production (fig-ure 1.6). Indeed, OPEC excess capacity is esti-mated to have fallen from 4.6 million barrelsper day in 2001 to only 1.4 million barrels perday in 2004. Moreover, oil prices remain wellbelow past peaks. Corrected for inflation andexpressed in 2003 dollars, oil prices averagedmore than $72 in 1980, and actually reachedmore than $100 in November of the previousyear. Viewed from this perspective, furtherhikes would not be unprecedented.

World Trade

World trade growth averaged 10.2 per-cent in 2004, reflecting rapid increases

in industrial production and investment activ-ity (figure 1.7). The expansion in trade vol-umes in 2004 is reminiscent of the increase ob-served in 2000 and mirrors the rapid recoveryin industrial production that began to takeshape in the second half of 2003 and contin-ued into 2004. More than 20 percent of theincrease in world merchandise trade volumeswas represented by China, whose imports in-creased by 32 percent—reflecting both thepositive impact of its accession to the WTOand unsustainable rates of investment andconsumption demand.


8

Figure 1.6 Crude oil prices, 1960–2004

0

30

20

10

50

40

60

70

80

Price per barrel ($)

Nominal dollars

1960

1964

1968

1972

1976

1980

1984

1988

1992

1996

2000

2004

a. Nominal petroleum prices deflated by the U.S. GDPdeflator, 2003 � 1.

Source: World Bank.

Constant dollarsa

Source: World Bank.

Figure 1.7 World trade rebounds

15.0

12.5

10.0

7.5

5.0

2.5

0

�2.5

�5.0

�7.5Q1

1991Q1

1993Q1

1995Q1

1997Q1

1999Q1

2001

Industrial production World trade

Q12003

�5.0

�2.5

0

2.5

5.0

7.5

World industrial production (percent change, year/year) Merchandise export volumes (percent change, year/year)

Trade in raw materials and investment goodswas particularly strong. As discussed above, ro-bust demand for raw materials was an impor-tant factor underlying the trade expansion in anumber of developing countries. In particular,oil, steel, and minerals trade was strongly influ-enced by the rapid increase in Chinese manu-facturing and construction sectors. Similarly,fast-growing global investment expenditureswere particularly important in spurring exportdemand in countries such as Germany andJapan that specialize in the fabrication of ma-chinery and other physical capital.

As a whole, developing countries havegrown their share in world markets by about19 percent (figure 1.8), up from 19 to 23 per-cent since 2000. Much of this rise is attributedto China, which has seen its share in worldexports double from 2.9 to 5.8 percent be-tween 2000 and 2004. Excluding China, theimprovement in the export share of low- andmiddle-income countries has been more mod-est (from 16 to 17 percent), although develop-ing countries in the South Asia and Europeand Central Asia regions have increased theirmarket shares considerably. Other regionseither maintained their market share (the restof the Eastern Asia and Pacific and the Middle

East and North Africa) or lost market share(Sub-Saharan Africa and Latin America andthe Caribbean).

Within regions the performance of specificcountries continues to be dictated, in part, bydomestic factors. So notwithstanding verystrong Chinese import demand, exports in therest of East Asia failed to increase as quickly,partly because political instability held backindustrial and investment activity in thePhilippines and Indonesia. In Latin Americanand the Caribbean, export volumes in Braziland Argentina grew briskly under the contin-ued influence of currency devaluations 2 yearsago, while strong world demand for metalsand minerals gave special impetus to Chileanexports.

Slower activity throughout the global econ-omy should translate into less rapid trade ex-pansion in 2005 and 2006. Trade in goods andnonfactor services is forecast to expand byabout 8.5 percent in 2005, down from an esti-mated 10 percent in 2004. Much of the decel-eration is conditional on the success of effortsto dampen the pace of activity in China, whichshould be reflected in slower import growth inChina and slower exports among its tradingpartners. Looking to other regions, the easingof activity in the United States, coupled withbroadly stable growth in Europe, is expected toresult in a somewhat more pronounced decel-eration of trade volumes in Latin America ascompared with Africa, the Middle East, andEastern European areas.

Major imbalances in the world trading en-vironment persisted during 2004 and willlikely continue to play a large role in 2005–06(figure 1.9). Notwithstanding the sharp accel-eration in world import volumes, the U.S. cur-rent account deficit reached 5.7 percent ofGDP in the second quarter of 2004, as Amer-ican consumption and investment volumes ex-ceeded domestic production by a wide margin(higher oil prices represented 0.6 percentagepoints of the 1.4 percentage point deteri-oration in the current account since the firstquarter of 2002). The expansion in the tradedeficit since the mid-1990s has been the main


9

Source: World Bank.

Figure 1.8 Export performance, percentchange in market share since 2000

�80

1989

199019

9119

9219

9319

9419

9519

9619

9719

98

China

East Asia(excluding China)

Developingcountries

(excluding China)

Developedcountries

1999

2000

200120

0220

0320

04�40

�30

�20

�10

0

10

20

30

40

50

60

Percent China, percent

120

60

80

100

40

20

0

�20

�40

�60

factor behind the rise in the U.S. current ac-count deficit—itself a major factor behind the15 percent real effective depreciation of thecurrency since February 2002. Barring a sub-stantial increase in domestic savings by, for ex-ample, a tightening of fiscal policy, downwardpressure on the U.S. dollar is likely to resumeas U.S. foreign borrowing requirements re-main high, and the already large amounts ofexternal debt continue to accumulate (see, forexample, Bergsten and Williamson 2004).

The U.S. trade deficit is largely a home-grown problem. While bilateral trade deficitswith specific countries are large, notably withrespect to China, the fact that these countrieshave only small overall surpluses supports theview that the deficit with the United States ismore a reflection of U.S. trade patterns than anindication of unfair trading practices. For ex-ample, China’s large bilateral surplus with theUnited States (but very small global surplus)reflects its specialization in the production offinal consumption goods (sold to the UnitedStates) based on intermediate and primary

imports from other developing countries withwhom China has a cumulatively large tradedeficit (Lau 2003).4

Failure to address the twin U.S. deficits couldhave significant impacts on developing coun-tries, especially if that failure leads to an increasein protectionist behavior. This is especially rele-vant because the substantial improvements inliving standards, wages, and incomes in manyupper-lower and middle-income countries havebeen the result of expanding their world marketshare in manufactures. An increase in protec-tionism could halt these countries’ progress anddeny other poor countries the same avenue todevelopment. Moreover, a retreat from recentefforts to reduce trade barriers or a failure tomake further progress—especially concerningagricultural subsidies—could have substantialnegative consequences on many of the world’spoorest countries.

International Finance

Over the past several years, favorableglobal conditions, strong growth, rapidly

expanding trade, and domestic reforms (in-cluding lower fiscal deficits and inflation) haveallowed developing countries to substantiallyimprove their financial positions (figure 1.10).


10

Source: World Bank.

Figure 1.10 Developing countries’ debtand interest payments easing downwardsince 1999

2.5

3.0

2.0

1.5

1.0

0.5

00

10

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

20

30

40

50

60

Percent of GNI

Interest payments(right axis)

Total debt, outstanding anddisbursed (left axis)

Percent of GNI

Billions of U.S. dollars, 2004

a. European Union refers to the 15 member countries, priorto the latest expansion.

Source: National agencies.

Figure 1.9 Trade balances in majorregions

�800

�700

�600

�667

52

260

134

960

170

�500

�400

�300

�200

�100

300

200

100

0

United

Sta

tes

Other

NAFTA

Europ

ean

Uniona

Japa

n

China

Other

deve

loping

Latin

Am

erica

and

the

Caribb

ean

On average, their debt to GNI ratio has fallenfrom 44 to 37 percent since its peak in 1999.This progress, plus low interest rates andstrong growth, has substantially lowered thedebt-servicing burden for most countries.While the situation of the most heavily in-debted countries remains serious, they havemade the greatest gains—debt to GDP ratiosfor these countries are down from 161 to 86percent since 1994—partly because of debt-relief programs instituted over this period.

These favorable conditions have also al-lowed many countries to strengthen their ex-ternal position. Most countries have succeededin improving their structural positions so that,even in the face of higher oil prices or a moremoderate pace for growth, their current ac-count positions should not deteriorate to thepoint where financing becomes problematic.As a whole, the current account position of themajor groups of developing countries is closeto balance or in surplus (table 1.2).

Developing countries have become majorsources of international capital. Since 2000, thecentral banks of some of the largest developingcountries have increased their foreign reserves bymore than 80 percent. Taken as a group, the re-serves of Brazil, China, India, Mexico, Thailand,and Turkey now represent over 45 percent of de-veloping country reserves. Indeed, following

private investors’ retreat from equity and bondinvestments in U.S. dollar-denominated assets,5

the central banks of these countries have becomeone of the most important sources of financingfor the large U.S. current account deficit, absorb-ing 51 percent of the overall increase in foreignofficially-held U.S. treasury bills between March2000 and January 2003. While this has allowedthese countries to increase their reserves by a sub-stantial margin, it has been achieved at the ex-pense of increasing their exposure to the U.S. dol-lar (figure 1.11). Among these countries, theshare of U.S. treasury bills in their official re-serves has increased by as much as 20 percentagepoints and equals almost 70 percent in the case ofMexico, and 58 percent in China. Should thesecountries decide to rebalance their reserve port-folio by slowing the pace at which they accumu-late dollar-denominated reserves, either


11

2000 2003

Othercurrencies

28%

U.S. dollar72%

Billions of dollars

Developing country reserves Share in increase

Figure 1.11 Rising U.S. dollar reserves

50

0

100

150

200

250

300

400

450

350

500

Source: World Bank.

China

Other developing countries

Table 1.2 Current account balances

Percent of GDP in 2004

East Asia and Pacific 1.5South Asia �0.5Middle East and North Africa 14.4Sub-Saharan Africa 1.3Europe and Central Asia 0.1Latin America and the Caribbean 0.7High-income countries �0.8

Source: World Bank estimates.

downward pressures on the dollar will accentu-ate, or interest rates will have to rise in order toattract sufficient private capital inflow.

Notwithstanding robust aggregate perfor-mance, many countries have been lesssuccessful in reaping the benefits of the lastfew years of strong economic conditions, andtheir high current account deficits could im-peril their stability—especially in the contextof slower growth in trade and world economicactivity. More than 50 developing countrieshave current account deficits that exceed5 percent of GDP. As a result, even the moder-ate hikes in interest rates, deterioration interms of trade, and the slower export demandprojected in the baseline will likely requirethese countries to undergo significant cuts toimports and domestic consumption in order tomaintain external stability. If trade growthwere to slow more than currently predicted, orif terms of trade were to deteriorate more be-cause of an additional hike in oil prices, the re-quired adjustment could be severe.

Risks and Policy Priorities

Forceful steps are required to reduce thetwin deficits in the United States. As

the preceding discussion has indicated, overthe past few years, private sector equity and di-rect investment financing of the very large U.S.current account deficit has dried up,6 havingbeen replaced to a large extent by increasedpurchases of U.S. bonds by foreign centralbanks, notably those of developing countries.While these countries’ build up of reserves hashelped improve their external financial posi-tion, the stock of U.S. dollars that they nowhold is very high and represents a dispropor-tionate share of their assets. It is not clear thatthey can or should increase these stocks fur-ther by continuing to absorb the lion’s share ofnet new U.S. treasury bills (6 developing coun-tries absorbed more than half of net new issuessince 2000).7 Assuming their appetite for trea-suries wanes, downward pressure on the U.S.dollar is likely to re-emerge, and yields will

probably have to rise in order to motivateprivate investors to re-enter the market.8

Simulations suggest that a 200 basis point in-crease in long-term interest rates could reduceworld GDP over the short- to medium-term byabout 0.5 percent per annum;9 the impactwould be somewhat stronger for developingcountries, because higher rates will raise debtservicing burdens, which require additional cutsto spending and demand(figure 1.12). Over thelonger term, if the twin deficits in the UnitedStates are not addressed (a tightening of fiscalpolicy would reduce both deficits by increasingU.S. savings10), the problem is likely to intensify.Permanently higher long-term interest rateswould render a wide range of investment pro-jects uneconomic and slow the pace of potentialoutput for a considerable time11— leading, per-haps, to a period of stagflation similar to thatobserved during the 1970–80s.

While higher U.S. interest rates mightmaintain investor interest in the dollar, theywould have serious disruptive impacts oncountries with large U.S. dollar debts. Forcountries such as Brazil, Indonesia, the Philip-pines, Poland, and Turkey, a 200 basis pointincrease in dollar interest rates would signifi-cantly increase debt-servicing charges. In-creased outflows could provoke large depreci-ations in their currencies (as much as9 percent), which would only increase the


12

Deviation from baseline (percent of GDP)

Source: World Bank.

Figure 1.12 Impact of a 200 basis pointincrease in interest rates

0

0.1

2005 2006 2007 2008

0.7

0.6

0.5

0.4

0.3

0.2

Developedcountries

Developingcountries

domestic burden of their external debt andgenerate further downward pressure on theircurrencies. Maintaining stability would, in alllikelihood, require a substantial reduction inimports, consumption, and investment, whichwould result in slower growth and impede in-creases in poverty reduction.

The risk of such outcomes makes redress-ing imbalances all the more pressing. Amongdeveloped countries, steps need to be taken toreduce the U.S. government deficit, whichwould lower overall borrowing requirementsand investor’s concerns over the long-termfinancing of the debt. In Europe and otherOECD countries, more resolute steps to re-dress government deficits and to create thefiscal room necessary to deal with the fiscalconsequences of aging will be necessary iflong-term interest rates are to remain low. Fordeveloping countries, a gradual appreciationof some currencies relative to the dollar couldhelp by permitting a further depreciation ofthe dollar. However, in the absence of fiscaltightening in the United States, such measuresare unlikely to have a significant impact. Fis-cal consolidation is also required in many de-veloping countries. Recent steps to lower ex-isting government deficits move in the rightdirection and need to be pursued—as do ef-forts to reduce trade barriers so that exportopportunities can increase. While these ac-tions may well imply hardship and impose realpolitical costs, the human and political conse-quences of entering into a period of higher in-terest rates without external and internal fi-nances on a firm footing would be even moredramatic. Finally, funding for initiatives to re-lieve the debt burden of the poorest countriesneeds to be increased.

Should oil prices rise even further, theeconomies of low-income countries are likelyto be among the hardest hit. Oil prices are as-sumed to moderate in the base case, fallingfrom $39 per barrel (for the average of West-Texas Intermediate, Brent, and Dubai oils)12

in 2004 to $32 in 2006. However, given sup-ply and geopolitical conditions, there is a realrisk that prices will either remain at current

levels ($46.8 in October 2004 for thisaverage—$49.5 for Brent) or rise evenfurther. Simulations suggest that were eventsto temporarily disrupt supply by about 1 mil-lion barrels per day, oil prices could be ex-pected to increase by about $10 a barrel. Inmacroeconomic terms, such an increasewould slow economic growth by about 0.5percentage points in the following year.13

However, the resulting terms of trade shockwould be larger in many poorer countries(�2.4 percent of GDP for highly indebtedpoor oil-importing countries versus �0.2 per-cent of GDP for high-income countries) be-cause of the relatively high share that energyrepresents in their imports. And sucheconomies tend to be more sensitive to agiven terms of trade shock because of theirlimited ability to attract capital flows thatwould offset any resulting increases in theirtrade deficits. In contrast to high-incomecountries, which can increase their externalborrowing to offset the real-side impact ofhigher oil prices, low-income countries areobliged to absorb most of the shock immedi-ately. As a result, they undergo a deprecia-tion and substantial reductions in consump-tion and investment spending—adjustmentmechanisms that ultimately reduce spendingon imports by almost the entire amount ofthe increased oil bill (figure 1.13). Their in-ability to defer adjustment (like high-incomecountries do) implies significant costs, bothto individuals who see their consumptionpossibilities reduced and to the economy, aslower levels of investment feed through toreduce the capital stock and diminish pro-ductive capacity.

Finally, a failure of current efforts to slowthe unsustainable pace of growth in China byengineering a soft landing could result inmajor disruptions. The Chinese authoritieshave put into place a number of specific—mainly command and control—measures,that restrict additional investment and lendingto the construction and heavy productionsectors. In the World Bank’s forecast, this isprojected to succeed in slowing overall growth


13

to some 7.1 percent in 2006, down from anestimated 8.8 percent this year.

So far, these steps have slowed import de-mand in a number of sectors, notably metalsand ores,14 while credit restrictions have dra-matically reduced the pace of money creation.In contrast, private consumption growthshows no sign of easing, and inflation haspicked up rapidly. For the moment increasedcosts have not found their way into wages,but such a possibility cannot be ruled out.Should overheating contribute to further in-creases in inflation, a stronger policy responsemay be required. Moreover, investment levelsremain very high, leaving open the possibilityof a very rapid correction, especially if badloans in the banking sector reveal themselvesto be a serious problem. Either eventualitycould provoke a more abrupt slowdown thanforecast.

Long-Term Growth, StructuralChange, and Poverty

This part of the report, as in years past, pre-sents a long-term growth scenario for

the global economy and its implications formeeting one of the MDGs: the halving of the

proportion of the population living on $1 orless a day by 2015 (compared to 1990 levels).The strong economic growth in developingcountries over the last 2 to 3 years, which isexpected to continue through 2006, albeit at asomewhat slower pace, is based on solid fun-damentals that are likely to carry forward andcontribute to long-term economic prospects.In our base scenario this leads to an annualgrowth of some 3.5 percent in per capita GDPbetween 2006 and 2015, and contributes toachieving the MDGs. The poverty MDG willbe met on a global basis, but a large numberof countries will not meet the goal, particu-larly those in Sub-Saharan Africa. And thoughgrowth is necessary to make progress towardachieving the MDGs, in most countries,growth is insufficient without more targetedpolicies.

At least four factors are responsible for therecent and prospective improvement ingrowth prospects. As outlined in the first sec-tion of this chapter, among the solid funda-mental changes in developing countries is animprovement in macroeconomic conditions(e.g., inflation and indebtedness). The recentWorld Development Report stresses theimportance of the investment climate, which


14

Low-income net oil importers

Percent of GDP

High-income countries

Percent of GDP

Figure 1.13 First year impacts of a $10 increase in oil prices

�0.8

�0.1

�0.2

�0.3

�0.4

�0.5

�0.6

�0.7

0

Source: World Bank.

Domesticdemand

Current accountbalance

�0.8

�0.1

�0.2

�0.3

�0.4

�0.5

�0.6

�0.7

0

Domesticdemand

Current accountbalance

vested interests protected by an RTA impedeprogress toward a globally more beneficialagreement? These questions will be addressedin chapter 6. The conclusion from this chapteris that the challenge for most developing coun-tries will be the creation of jobs for a risingwork force, rather than how to deal with em-ployment shifts across economic activities.

Long-term growth scenarioThe global economy is currently reboundingfrom the downturn suffered in 2001 and2002. Not all regions are benefiting equallyfrom the rebound—Japan and the UnitedStates are leading the way among industrialeconomies—but there is fairly solid progressin all the main developing regions on an ag-gregate basis. This year, 2004, is likely to bethe peak in the current upward cycle, witheconomies drifting toward long-term trendgrowth in 2005 and beyond. Table 1.3 re-flects a plausible long-term scenario for thehigh-income countries and the World Bank’ssix aggregate developing regions (see box 1.1for details concerning aggregation). The sce-nario reflects current views on potential trendgrowth over the 2006–15 decade. Betterpolicies, an acceleration in investment, andother factors could improve the prospects,particularly for the slower growing regions.There is still a considerable gap in the pro-ductivity levels between developing and in-dustrial economies, and a number of develop-ing countries—particularly in Asia—havedemonstrated, over the last 20 to 30 years, asustained ability for rapid growth.

The focus of this forecast section this year ison anticipated structural changes. These havemany dimensions—demographic, rural versusurban, sectoral, employment shifts, openness,and income distribution, among others. Whilemost of these shifts have long-term positiveimpacts, they can also be associated withshort-term transitional costs. Public policiescan limit the costs of transition, but they canalso be significantly reduced—at least in termsof duration—in a fast growing economy wherejob growth is robust.

has improved in many countries and has led toan acceleration in growth. A third factor,explored in more detail below, includessignificant structural changes—that is, eco-nomic diversification and a move away fromreliance on agriculture, and integration withthe global economy; both of these structuralchanges involve increased urbanization. Afourth factor, pursued in greater detail inchapter 6, is the reduction in trade barriers.

The special focus of the long-term scenarioin this report is on structural changes, partic-ularly as they affect employment. Rapidgrowth will, in and of itself, lead to structuralchanges; that is, a relative decline in agricul-ture and a rise in the demand for services.Countries need to think ahead, allocate scarcepublic investment in a rational manner, andpromote education to better position theirwork force for a changing environment.While structural changes are likely to be im-portant, many developing countries face anequal challenge in the sheer growth of thelabor force. Labor force growth rates arelikely to decline over the next decade, but inmany regions they will average between 1.5 to2.5 percent per annum. For the poor, bothgrowth and structural change are likely to bebeneficial. Growth, to the extent that it liftsall incomes, will inevitably lead to a fall inpoverty. Structural change can accelerate theprocess of poverty reduction. A decline in therural population could ease wage pressures.A rising urban population provides easier ac-cess to essential health and education servicesand can lead to a rise in transfers to ruralareas.

The focus on structural change also links tothe broader theme of the report—the shapeand impacts of regional trade agreements(RTAs). The RTAs will undoubtedly lead to ad-ditional structural shifts, and with associatedtransitional costs. How do RTAs compare withgrowth-induced structural shifts? Do RTAsproduce structural shifts that are broadly con-sistent with those induced by a truly openglobal economy (which would emerge from amultilateral agreement)? And, if not, would the


15

Structural Changes over TwoDecades

Looking back on the last 20 years of devel-opment, many developing regions have al-

ready witnessed significant structural shifts.Perhaps foremost is the decline of agricultureas a source of income and employment. InEast Asia and the Pacific, agricultural valueadded has declined from a 28 percent share in1982 to only 15 percent in 2002, and manu-facturing, other industrial, and services have

risen (see figure 1.14). Services, according tothese figures, still represented less than40 percent of GDP in 2002, well below thenearly 65 percent share in the high-incomecountries of East Asia. Thus there is still sig-nificant scope for further structural shifts.

The value added shares also belie the rela-tive employment share in agriculture, whichtends to be much higher. Take, for example,Thailand, where agriculture’s share of valueadded is below 10 percent, but still employsaround 50 percent of the total labor force. Inthe high-income countries, the relevant sharesare around 2 percent of value added and lessthan 4 percent of employment.15 Higher agri-cultural productivity and relative wage differ-entials will continue to drive an exodus fromagriculture into other sectors. And the changecan come rapidly. In the Republic of Korea,the percent of employment in agriculturedropped from 32 percent in 1982 to 10 per-cent in 2001. The agricultural transformationis present in some of the other developing re-gions as well; for example, in South Asia thepercent of employment in agriculture droppedfrom 40 percent in 1982 down to 27.2 percentin 2002, and in Latin America and theCaribbean, the percent of employment in agri-culture dropped from 14.4 percent down to10.6 percent over the same two-decadeperiod. There has been no significant shift ineither the Middle East and North Africa orSub-Saharan Africa regions. At the same time,neither of those two regions witnessed mucheconomic growth, with only 0.4 percent percapita growth per annum in the former, and aloss of 0.3 percent per annum in the latter.

In all regions, save East Asia, one can see aclimb in the share of services. This is not sur-prising because services are assumed to be in-come elastic and a relative rise in the con-sumption share of services is understood. Thiseffect is reinforced by the relatively high rate ofproductivity growth in manufacturing. All elsebeing equal, this reduces the price of manufac-tures relative to services and hence enhancesthe value share of services. Perhaps what ismore surprising is the variation across regions.


16

Box 1.1 Theaggregation paradoxThe per capita growth rate for the world re-flects the so-called aggregation paradox. Thelong-term per capita growth rates for high-income and developing countries are, respec-tively, 2.4 and 3.5 percent per annum, butthe global growth rate is only 2.1 percentand is not the average of the growth rates(weighted or un-weighted). The followingtable highlights the aggregation paradox.The paradox is explained by the relativelyhigh weight of high-income countries GDPin the world total, but their low weight inworld population.

High- Developincome -ing World

Population (million)2006 970 5,340 6,3202015 990 5,900 6,900Growth ratea 0.3 1.1 1.0

GDP ($billion)2006 31,200 8,200 39,4002015 39,500 12,300 51,800Growth ratea 2.7 4.6 3.1

GDP per capita ($)2006 32,090 1,530 6,2402015 39,700 2,080 7,510Growth ratea 2.4 3.5 2.1

a. Growth rates are percent per annum.


17

East Asia and Pacific

Structure of value added (percent)

South Asia


Middle East and North Africa


Figure 1.14 A rise in services

Agriculture Industry Services

10

0

20

30

40

50

60


10

0

20

30

40

50

60


10

0

20

30

40

50

60

Sub-Saharan Africa


Latin America and the Caribbean


Europe and Central Asia


Agriculture

Source: World Bank.

Industry Services

10

0

20

30

40

50

60


10

0

20

30

40

50

70

60


10

0

20

30

40

50

70

60

1982

1992

2002

Table 1.3 Long-term prospects: Forecast growth of world GDP per capitaReal GDP per capita, annual average percentage change

1980s 1990s 2000–06 2006–15

World total 1.3 1.1 1.6 2.1

High-income countries 2.5 1.8 1.7 2.4OECD 2.5 1.7 1.7 2.3

United States 2.2 1.9 1.8 2.5Japan 3.5 1.1 1.7 1.9European Union 2.1 1.8 1.5 2.3

Non-OECD countries 3.5 4.1 1.6 3.5

Developing countries 0.6 1.5 3.4 3.5East Asia and the Pacific 5.8 6.3 6.0 5.3Europe and Central Asia 1.0 �1.8 5.2 3.5Latin America & the Caribbean �0.9 1.5 0.8 2.4Middle East North Africa �1.6 1.1 2.4 2.6South Asia 3.3 3.2 4.2 4.1Sub-Saharan Africa �1.2 �0.5 1.2 1.6

Note: Aggregations are moving averages, reweighted annually after calculations of growth in constant prices.Source: World Bank.

In the high growth regions—East Asia andSouth Asia—there are two contrastingpatterns. In South Asia the employment inagriculture shifted mostly to services, with asmall increase in industrial output. In EastAsia, employment in agriculture shifted moreevenly between industry and services. Andthere appears to have been a structural breakin the 1990s with an acceleration of industrialoutput. This is consistent with the sharp rise inthe trade to GDP ratio doubling from 36 per-cent in 1982 to 72 percent in 2002, and withEast Asia as a hub of assembly and manufac-turing activities (see figure 1.15). There are, ofcourse, exceptions in each region. The Philip-pines, for example, has a sharp rise in servicesand a decline in manufacturing—perhaps as aresult of its regional comparative advantage inback office operations, call centers, and otherservices requiring specialized language skills.In South Asia, India’s services dominate, butgrowth is much lower in Bangladesh and Pak-istan, where textile and clothing exporters maybe taking advantage of their relatively gener-ous quotas to the main importing markets.

Three of the other regions—Latin Americaand the Caribbean, the Middle East and NorthAfrica, and Sub-Saharan Africa—show less

growth overall, but are also more dependenton natural resource production, and thoserelative prices have been declining over most ofthe period. Natural resources appear under in-dustrial production, so even if volume growthhas been positive, with declining relative pricesthe natural resources share in output could bedeclining. And apart from Latin America andthe Caribbean, these regions have also not re-ally benefited from global production sharingin the more integrated global economy. TheMiddle East and North Africa Region hasbarely seen any shift in its trade to GDP ratio.For both Sub-Saharan Africa and Latin Amer-ica and the Caribbean, however, the ratio hasincreased markedly, particularly in the1990s—from 50 percent to 69 percent for Sub-Saharan Africa, and from 27 to 47 percent forLatin America and the Caribbean. The morerecent rise in Latin America can be partly ex-plained by the implementation of a raft ofregional agreements, including NAFTA andMERCOSUR. At the same time Latin Amer-ica’s degree of openness is lower than that ofEast Asia, and in general it has been less co-opted into global production networks.

The transformation in the economies ofEurope and Central Asia over the last 15 yearsis a result of an abrupt structural shift. Thedominance of industry as part of an economicstrategy of planned economies was eliminated.Services in the transition economies quicklyfilled the gap, which led to significant disloca-tion for a period, but is now forming the basisof more rational and sustained growth.

Looking ahead it is clear that there is thepotential for significant change. While the rateof urbanization has been persistent over thelast two decades, there is a long way to go,particularly in Asia and Africa, before attain-ing the 80 percent level of the industrial coun-tries (figure 1.16). The income gap is alsohuge, even if incomes are measured in pur-chasing power parity (PPP) terms. In EastAsia, per capita incomes averaged just over$1,000 (1995 dollars) in 2002, compared withnearly $31,000 in the industrial countries—roughly a 30 to 1 differential. Even assuming


18

Figure 1.15 Rising openness to trade

0

30

20

10

50

40

60

70

90

80

Trade (exports and imports) to GDP ratio (percent)

1982

East A

sia

and

Pacific

South

Asia

Sub-S

ahar

an

Africa

Latin

Am

erica

and

the

Caribb

ean

Europ

e an

d

Centra

l Asia

Midd

le Eas

t and

North

Afri

ca

1992

2002

Source: World Bank.

a PPP exchange rate of 5 would still lead to asignificant 6 to 1 ratio in per capita incomes.In Latin America, the richest developingregion with a per capita average income of$3,700, would have a ratio similar to EastAsia using a PPP exchange rate of around 2.

Structural Change in the Future

Table 1.3 presents the long-term growthrates. This section focuses on some of the

consequences of growth and other underlyingassumptions of the long-term scenario onstructural changes, particularly regardinglabor shifts—both in volume terms and acrosssectors.16

In the aggregate, and assuming no change inlabor force participation rates, labor supplygrowth will slow down sharply in most regionsafter 2010—with the exception of the MiddleEast and North Africa and Sub-Saharan Africaregions (figure 1.17).17 In Western and EasternEurope, Russia, and Japan, the labor supplywould most likely shrink (even before 2010),putting additional pressure on underfinancedpension schemes. Additionally, it is the regionswith the highest labor force growth rates thatalso tend to have the lowest per capita growthrates, so these regions are on a knife-edge interms of their capacity to absorb high rates ofnew workers. These same regions typically haverelatively low labor force participation rates,particularly of females; thus increases in


19

Figure 1.16 Increased urbanization

0

30

20

10

50

40

60

70

80

Urban population as a share of total population (percent)

East A

sia

and

Pacific

South

Asia

Sub-S

ahar

an

Africa

Latin

Am

erica

and

the

Caribb

ean

Europ

e an

d

Centra

l Asia

Midd

le Eas

t and

North

Afri

ca

1982

1992

2002

Source: World Bank.

Population between the ages of 15 and 65 (percent)

Figure 1.17 Growth rate of labor supply declining

�1.5

0

0.5

1.0

1.5

2.0

2.5

�0.5

�1.0

3.0

Note: United States+ includes Australia, Canada, New Zealand, and the United States. NIEs are newly industrialized economies. XAS represents other major countries in East and South Asia.

Source: World Bank.

United

Sta

tes�

Europ

e

Japa

n

East W

est a

nd

Centra

l Asia NIE

China

India XAS

Latin

Am

erica

and

the

Caribb

ean

Sub-S

ahar

an

Africa

Midd

le Eas

t

and

North

Afri

ca

2005

2010

2015


20

Table 1.4 Labor market structure, 2005–15

Growth between 2005–15: percent per annum Growth decomposition

Agric Manuf Services Total Structure Expansion Total

Australia, Canada & New Zealand �0.7 �0.8 0.7 0.4 3.45 3.90 6.88United States �1.1 �0.6 0.8 0.5 2.95 5.62 8.34Japan �3.0 �2.2 �0.6 �1.0 3.62 9.35 8.70Korea and Taiwan �1.9 �0.6 1.0 0.5 4.18 5.65 9.17Hong Kong (China) and Singapore 0.0 �0.5 1.0 0.7 4.80 7.28 11.65EU with EFTA �2.6 �1.5 0.2 �0.2 4.92 2.28 3.12Brazil 0.8 0.5 1.4 1.2 3.52 13.09 16.28China 0.5 0.4 1.3 0.8 5.12 8.44 12.02India 0.1 1.6 2.4 1.7 9.27 19.89 28.02Indonesia 1.1 1.8 1.6 1.5 8.32 17.18 22.73Mexico �0.2 1.6 2.5 2.0 7.00 22.78 28.84Russia �1.0 �1.1 0.0 �0.4 4.97 4.35 2.43SACU �0.7 �0.2 0.9 0.6 3.82 6.87 10.36Vietnam 1.7 1.9 2.0 2.0 5.69 22.23 26.20Rest of East Asia 0.7 1.2 2.1 1.7 4.72 19.36 23.09Rest of South Asia 1.8 1.3 2.7 2.4 4.60 27.22 31.38EU accession countries �0.8 �1.1 0.2 �0.2 6.51 2.41 4.67Rest of ECA �0.3 0.3 1.1 0.8 6.15 8.88 14.43Middle East 2.3 1.6 2.5 2.3 5.26 26.43 30.79North Africa 0.8 2.0 2.5 2.0 7.27 23.74 29.83Rest of Sub-Saharan Africa 2.0 2.2 3.0 2.6 4.50 30.19 34.08Rest of LAC 2.1 0.9 2.2 1.8 5.36 20.74 25.41Rest of the world 0.8 1.2 2.1 1.7 4.15 19.45 23.12

Source: World Bank simulations.

participation rates will lead to additional labotmarket weakness.

With high-income demand elasticity for ser-vices and relatively higher labor productivity inmanufacturing, labor demand growth will tendto be higher in services than in manufacturingand/or agriculture (see table in endnote 17).This effect is quite pronounced in the industrialcountries, where labor demand growth be-tween 2005 and 2015 will be negative, on aver-age, in agriculture and manufacturing in allhigh-income regions, with all of the net growthoccurring in services (with the exception ofJapan, where labor force growth could poten-tially decline by 1 percent per annum on aver-age). The shift toward services also occurs in de-veloping countries, but with continued highgrowth in manufacturing and less growth inagriculture.

Table 1.4 also shows a summary measureof the structural changes. It decomposes thetotal change in the structure of the labor forceinto two components. The first is the

“structural” component, which measures thequantity of labor force movement across sec-tors, assuming no change in the volume oflabor. The second is an “expansion” compo-nent, measuring the overall growth in thelabor force. In the case of India, for example,the numbers suggest that the labor force willgrow by about 20 percent between 2005 and2015, or about 1.8 percent per annum. Andin each year, about 0.9 percent of the initiallabor force will move across sectors. Thus thetotal annual movement of 2.5 percent perannum is composed roughly of 2/3 expansionand 1/3 by intersectoral movements. It shouldbe clear from the decomposition that for mostof the developing regions, there will be morelabor movement from the expansion of thelabor force than from structural change, withthe notable exceptions of Russia and theother countries in Europe and Central Asia—and, perhaps somewhat more surprisingly,China. For the industrial regions with low ordeclining labor growth, clearly the structural


21

shifts will be relatively the same order of mag-nitude as the expansion component. But theshifts are relatively small on an annual basis,perhaps 0.3 to 0.5 percent of the labor force.

Chapter 6 of this report will re-address oneissue related to structural shifts in the contextof RTAs. Do RTAs lead to structural changesthat are inconsistent with the structuralchanges from a broad multilateral agreement?For example, a country signing an RTA mayhave a local comparative advantage in a givensector, but not a global comparative advantage.In this case, would the country need to undergotwo potentially costly adjustments, should amultilateral agreement be signed subsequent toan RTA? And would the vested interests thatbenefit from the RTA hamper the abilityto achieve a broader multilateral agreement,

with positive aggregate benefits, but hurt thesectors that thrived under the preferentialarrangement?

Poverty Forecast

Developing country economic performancehas been strong since 2002, and this is

projected to continue over the next two yearsand beyond (tables 1.1 and 1.3). This patternof high growth would in all likelihood lead toa halving of the number of poor (i.e., the per-centage of poor living on $1 or less a day) indeveloping countries between 1990 and 2015(table 1.5)—one of the key MDGs. At theglobal level, the target to be achieved in 2015is around 14 percent (one-half of 27.9), andthe forecast is for a headcount index of10.2 percent. This translates into a forecast of

Table 1.5 Regional breakdown of poverty in developing countries

Number of people living on less than $1 per day (millions)

GEP2004 GEP2005

Region 1990 2000 2015 1990 2001 2015

East Asia and Pacific 470 261 44 472 271 19China 361 204 41 375 212 16Rest of East Asia and Pacific 110 57 3 97 60 2

Europe and Central Asia 6 20 6 2 17 2Latin America and the Caribbean 48 56 46 49 50 43Middle East and North Africa 5 8 4 6 7 4South Asia 467 432 268 462 431 216Sub-Saharan Africa 241 323 366 227 313 340

Total 1,237 1,100 734 1,218 1,089 622Excluding China 877 896 692 844 877 606

$1 per day head count index (percent)

GEP2004 GEP2005

Region 1990 2000 2015 1990 2001 2015

East Asia and Pacific 29.4 14.5 2.3 29.6 14.9 0.9China 31.5 16.1 3.0 33.0 16.6 1.2Rest of East Asia and Pacific 24.1 10.6 0.5 21.1 10.8 0.4

Europe and Central Asia 1.4 4.2 1.3 0.5 3.6 0.4Latin America and the Caribbean 11.0 10.8 7.6 11.3 9.5 6.9Middle East and North Africa 2.1 2.8 1.2 2.3 2.4 0.9South Asia 41.5 31.9 16.4 41.3 31.3 12.8Sub-Saharan Africa 47.4 49.0 42.3 44.6 46.4 38.4

Total 28.3 21.6 12.5 27.9 21.1 10.2Excluding China 27.2 23.3 15.4 26.1 22.5 12.9


22

Table 1.5 Regional breakdown of poverty in developing countries (continued)

Number of people living on less than $2 per day (millions)

GEP2004 GEP2005

Region 1990 2000 2015 1990 2001 2015

East Asia and Pacific 1,094 873 354 1,116 864 230China 800 600 256 825 594 134Rest of East Asia and Pacific 295 273 98 292 271 95

Europe and Central Asia 31 101 48 23 93 25Latin America and the Caribbean 121 136 124 125 128 122Middle East and North Africa 50 72 38 51 70 46South Asia 971 1,052 968 958 1,064 912Sub-Saharan Africa 386 504 612 382 516 612

Total 2,653 2,737 2,144 2,654 2,735 1,946Excluding China 1,854 2,138 1,888 1,829 2,142 1,812

$2 per day head count index (percent)

GEP2004 GEP2005

Region 1990 2000 2015 1990 2001 2015

East Asia and Pacific 68.5 48.3 18.2 69.9 47.4 11.3China 69.9 47.3 18.4 72.6 46.7 9.7Rest of East Asia and Pacific 64.9 50.8 17.6 63.2 49.2 14.7

Europe and Central Asia 6.8 21.3 10.3 4.9 19.7 5.2Latin America and the Caribbean 27.6 26.3 20.5 28.4 24.5 19.6Middle East and North Africa 21.0 24.4 10.2 21.4 23.2 11.9South Asia 86.3 77.7 59.2 85.5 77.2 54.2Sub-Saharan Africa 76.0 76.5 70.7 75.0 76.6 69.2

Total 60.8 53.6 36.4 60.8 52.9 32.0Excluding China 57.5 55.7 42.0 56.6 54.9 38.6

Source: World Bank.

622 million persons living $1 or less a day in2015, compared with 1.2 billion in 1990 andan estimated 1.1 billion in 2001.18 With re-spect to the somewhat higher poverty line of$2 a day, the headcount should improve to32 percent in 2015—not quite a halving of theestimated 61 percent headcount index in1990—and corresponding to almost 2 billionpoor.

However, progress is highly uneven acrossand within countries. The global target willlargely be achieved because of the significantprogress on poverty reduction in China andIndia. Sub-Saharan Africa lags far behind, andthough poverty rates are much lower in some ofthe other regions, for example Latin Americaand the Caribbean, progress over the last

15 years has been insufficient to be on track toachieve the income poverty target in 2015 with-out more rapid growth or policies that are bet-ter targeted to the poor. Within regions,progress has also been uneven. Despite the hugeoverall reduction in East Asia, several countries,for example, Cambodia, Lao PDR, and PapuaNew Guinea, are off track to meet the goal. InSub-Saharan Africa, there are only eightcountries—representing 15 percent of the sub-continent’s population—that will potentiallymake significant progress toward achieving theincome poverty target. Within countries, suchas China, there are large pockets of poorpeople, and reducing poverty in these pockets isdifficult because they are often concentrated inremote, hard-to-reach locations. Links to the

national and/or global economy are weak, andprovision of public services—education, health,water and sanitation—is difficult andexpensive.

This year’s poverty forecast, as in yearspast, reflects changes in two key dimensions.First, new country surveys lead to a re-evaluation of the level of poverty in 1990 andin the most recent base year, 2001. At theglobal level, the $1/day headcount index for1990 has been shaved slightly from 28.3 per-cent in last year’s report, to 27.9 percent inthis year’s report. There is also a very modestdecline in the estimated level of poverty for2001. The new surveys also force a re-evalu-ation of the link between income growth andpoverty reduction. Using the latest survey in-formation and last year’s economic forecast,the forecasted decline in poverty is somewhatmore rapid, with the headcount index declin-ing to 10.4 percent (from 21.1 percent in2001), instead of 12.5 percent (from 21.6percent in 2000).19 The second key dimen-sion is the change in the long-term economicforecast. The changes overall are relativelymodest. However, the somewhat improvedperformance anticipated between 2003 and2006 generates better average growth for theforecast period 2001–15 and drops the head-count index for 2015 from 10.4 percent to10.2 percent.

While progress on income poverty in partsof the world, particularly East and South Asia,has been spectacular if not historic, there is noroom for complacency. As mentioned earlier,there are significant pockets of poverty evenwithin the more successful countries. More-over, there are other dimensions of poverty inwhich progress has been more limited, andalmost all developing countries are off track. InEast Asia, for example, the region scores rela-tively well for achieving 100 percent primaryschool completion rates, with China and Viet-nam already having achieved the target and thePhilippines on track.20 But Thailand and In-donesia are off track, as are some of the poorercountries in the region. For the child mortalityMDG, the situation is more worrying. Four

countries are on track to achieve the target—Indonesia, Lao PDR, Malaysia, and the Philip-pines. All other countries are off track, andtwo—Cambodia and Papua New Guinea—areseriously off track. The situation is also dire forbirths attended (linked to maternal mortality)and access to safe water. These examples alsoillustrate that the other MDG targets are lessdirectly correlated to income levels.21 For ex-ample, Lao PDR and Indonesia are on track forthe child mortality target, but Thailand is not.

Concluding Remarks

The rapid growth of developing economies,mostly concentrated in East and South

Asia, has produced a spectacular, if not his-toric, fall in poverty that will enable theachievement of the poverty MDG on a globalbasis, although many countries will be seri-ously off-target. The rapid growth has been as-sociated with large structural shifts—greateropenness, more urbanized populations, and asharp fall in agricultural employment. Thesetrends will persist in the future as growth ratesremain high, and incomes and productivity lev-els in developing countries are still well belowindustrial country averages—even taking intoaccount PPP adjustments. As an example ofpotential structural shifts, take China’s level ofurbanization. Its rural population may not ap-proach the 20 percent level of industrial coun-tries, but a 50 percent share in 2015 could leadto a cumulative migration in the range of 140to 175 million persons between 2005 and2015. Such large shifts will require consider-able public and private resources and their effi-cient allocation. Chapter 6 addresses a comple-mentary issue—structural changes induced bychanges in trade policies, notably the impactsof preferential trade agreements.

Notes1. The investment to GDP ratio in the United

States is currently 21 percent, close to its peak of21.5 percent during the Internet bubble, and wellabove historical peaks of less than 18 percent.


23

2. The weighted average of the dollar’s fluctuationsrelative to world currencies.

3. West-Texas Intermediate was much higher inOctober 2004 ($56). The overall average wasdepressed by the price of other oil (notably fromDubai), which was lower because producers increasedthe supply of lower quality oil.

4. Lau (2003) estimates that because of the re-export nature of its trade, the domestic value-addedcontent of Chinese exports may be as little as 20 percent.

5. Net equity and foreign direct inflows of foreignprivate investors declined by 73 percent between 2001and 2003. At the same time, net outflows by Americanprivate investors increased by 10 percent. As a result,total flows have reversed, from a significant inflow of$35 billion in 2001 to a $195 billion outflow in 2003.Since then, these trends have continued, with total out-flows representing $267 billion in the second quarterof 2004.

6. See endnote 5.7. Calculated as the change in U.S. t-bills held by

the central banks of these countries divided by the netincrease in t-bills held by official lenders (see http://www.treas.gov/tic/mfhhis01.txt).

8. Mussa (2004) suggests that a further 20 percentdepreciation might be required to bring the U.S. econ-omy into external balance.

9. These results are consistent with those publishedby the OECD for developing countries (see Dalsgaardand others 2001).

10. Even after Ricardian equivalence-basedchanges to private saving. Nevertheless, Brooks andothers (2003) show that, taken alone, neither a 2 per-centage point cut in fiscal spending, nor a 10 percenteffective depreciation would be sufficient to restore ex-ternal balance in the United States. They argue that acombination of depreciation, stronger world demand,and a larger fiscal contraction would be required.

11. Under higher interest rates, the desired stock ofcapital declines, which requires a prolonged period ofslower growth before the economy adjusts to the newlower levels of output and capital.

12. In October 2004, this average price was $46.8comprised of $53 for West-Texas Intermediate, $49.5for Brent, and $37.7 for Dubai oil.

13. Dalsgaard and others (2001) estimate similarimpacts for OECD countries.

14. Growth in steel demand fell 36 percent duringthe 3-month period ending in July, while copper im-ports were flat.

15. World Bank 2003b.16. Unlike the previous section, which focused on

the structure of value added, this section focuses onlabor. The focus on labor provides a better perspec-tive on the poverty dimension of structural shift. The

historical analysis focused on output because of thegreater availability and reliability of the data. Histori-cal data on employment patterns has many gaps, andthe data that does exist is often not compatible acrosscountries.

17. The baseline scenario and the induced struc-tural changes are predicated on a number of assump-tions. First, growth in the labor supply is equated withgrowth of the working age population. For all regions,this implies a slowing of labor force growth, albeit withhigh growth in some developing regions. At the sametime, the labor force is assumed to be flexible and thuswill reinforce anticipated structural shifts. Second, sav-ings are similarly influenced by demographics. In manydeveloping countries this will translate into a slightacceleration in savings as the ratio of youth to workersdeclines, and a decline in industrial countries as theratio of elderly to workers rise (explored in more detailin World Bank 2003a). Investment growth will largelybe driven by domestic savings, as it has in the past;however, with modest increases in net capital flows to-ward developing countries, with the exception of EastAsia, which has been a major source of internationalcapital over the last five years.

Third are the assumptions regarding productivitygrowth; based on previously observed trends, these aredivided into three broad economic sectors. In agricul-ture, it is assumed that the past growth of roughly2.5 percent per annum is maintained through 2015(see, for example, Martin and Devashish 1999).Maintaining this high rate of agricultural productivitywill require continued and perhaps increasing invest-ment in agricultural research and extension, combinedwith rising investment in agricultural infrastructure,particularly for water resource management. This rateof productivity growth in agriculture is consistent witha modest secular decline in agricultural prices, relativeto the general price trend, as observed in the past. Theother two broad sectors are manufacturing and services.Again, based on past trends, it is assumed that produc-tivity growth in manufacturing will be higher than inservices. This has two impacts: (1) it reduces the price ofmanufactures relative to services, all else being equal,and thus enhances the share of services in value terms;and (2) for the same level of output, it reduces the


24

Income elasticities in the Linkage model

Ag. and Industrialfood Energy goods Services

United States 0.01 0.58 0.78 1.14Japan 0.04 0.64 0.68 1.24Europe 0.08 0.72 0.71 1.29Rest of high-income 0.16 0.86 0.80 1.26Low-income 0.52 1.40 1.08 1.41

demand for employment in the manufacturing sectors,and thus allows for a shift of labor toward services.

Fourth are the demand assumptions—the other sideof the coin regarding structural changes. High-incomecountries have already witnessed a large decline inthe demand for agriculture and food relative to income.Demand for services has increased relative to income andthe demand for other goods. And there is no reason forthese trends not to continue in the future. Thus the for-ward-looking scenarios assume that income elasticitiesover the next 10 years will largely reflect their currentlevels, though highly differentiated across commoditiesand regions (see table).

18. The absolute number of poor won’t necessarilybe halved due to population growth.

19. A more subtle change in the methodology hasalso been incorporated in this year’s poverty forecast.The poverty forecast is based on the growth of thesurvey-based per capita consumption, assuming distri-bution neutrality (with some exceptions). However, ithas been observed in the past that survey-based con-sumption growth deviates from consumption growthas measured in the national accounts. A conversionfactor has been used to adjust for this deviation, whichfor most countries implied an elasticity of 0.9. In otherwords, if national income consumption grows at10 percent, the assumed growth in survey-based con-sumption is 9 percent. More recent econometric evi-dence suggests that the long-run elasticity is 1, but thatthere are short-term deviations from the long-runelasticity. Because of the robustness of the long-run re-lationship, the new forecast assumes an elasticity of 1.Thus, all else being equal, this year’s forecast will belower than in the past because of higher implied con-sumption growth.

20. See World Bank 2004.21. The World Bank, in its effort to improve its abil-

ity to monitor and forecast the other dimensions of theMillennium Development Goals, is developing and test-ing a new tool to forecast some of the MDGs. The toolwill link economic growth with expenditures on health,education, and infrastructure. It will also capture some

of the complementarities across targets, for example thedegree to which improvements in access to safe watercan improve health outcomes. A pilot study is currentlybeing undertaken for Ethiopia and first results will bedescribed in the Global Monitoring Report 2005.

ReferencesBergsten, Fred C. and John Williamson, eds. 2004.

Dollar Adjustment: How far? Against What? InSpecial Report, No. 17, Institute for InternationalEconomics, Washington DC.

Brook, Anne-Marie, Franck Sédillot, and Patrice Olli-vaud. 2004. Channels For Narrowing The USCurrent Account Deficit And Implications ForOther Economies. OECD Economics DepartmentWorking Papers, No. 390, OECD, Paris.

Dalsgaard, Thomas, Christophe André, and PeteRichardson. 2001. Standard Shocks In TheOECD Interlink Model OECD EconomicsDepartment Working Papers, No. 306, OECD,Paris.

Lau, Larence. 2003. Is China Playing by the Rules?Free Trade, Fair Trade, and WTO Compliance.Testimony at a Hearing of the Congressional Ex-ecutive Commission on China, Washington, DC.

Martin, Will, and Devashish Mitra. 1999. ProductivityGrowth and Convergence in Agriculture andManufacturing. Working Papers, No. 2171,World Bank, Washington, DC.

Mussa, Michael. 2004. Exchange Rate AdjustmentsNeeded to Reduce Global Payments Imbalances.Special Report, No 17, Institute for InternationalEconomics, Washington DC.

World Bank. 2003a. Global Economic Prospects: In-vesting to Unlock Global Opportunities.Washington, DC: World Bank.

———. 2003b. World Development Indicators.Washington, DC: World Bank.

———. 2004. Global Monitoring Report. Washington,DC: World Bank.


25

In the last four decades, developing countrieshave burst onto the global marketplace. Theirshare of global trade increased from aboutone-fifth in 1960 to about one-third in 2004—at a time when global trade as whole wasincreasing to unprecedented levels. In everyregion, exports have outpaced the growth ofoutput and increased as a share of GDP. Threerounds of multilateral trade negotiations com-bined with structural economic reforms un-dertaken throughout the world ushered in thesustained reduction in border protection thatmade this growth possible. The World TradeOrganization (WTO), formed in 1994, con-solidated an evolving system of rules based onnondiscrimination among trading partners—a cornerstone of the multilateral system.

Today a second trend in the trading systemis rapidly gaining momentum and establishinga very different set of rules. This new trend isthe proliferation of regional and bilateraltrade agreements (RTAs)—agreements amonga group of countries that reduce barriers totrade on a reciprocal and preferential basis forthose in the group. The number of these agree-ments has more than quadrupled since 1990,rising to around 230 by late 2004.1 Trade be-tween RTA partners now makes up nearly 40percent of total global trade, and new agree-ments increasingly address issues beyondtrade. The value of preferences has steadilyfallen, however, as most countries have beenreducing tariffs across the board to all

partners on a most favored nation, or nondis-criminatory (MFN) basis, at the same time asthey have been eliminating barriers preferen-tially through RTAs. In fact, roughly 66 per-cent of the decline in average tariffs in devel-oping countries during the last two decadeshas come from unilateral reductions, as dis-tinct from 25 percent coming out of theUruguay Round and around 10 percent fromRTAs. Moreover, product exclusions and re-strictive rules of origin further limit the trade-expanding effects of preferences. Nonetheless,the result of this proliferation is an increas-ingly complex global trading system wheredifferent countries’ access to a given marketare often governed by very different sets ofrules.

This chapter charts the rise of RTAs, exam-ines the different motivations countries havefor pursuing RTAs, and draws attention tothe complexity they generate. It then describesthe evolution of regional trading patterns andshows how the major developing regions dif-fered strikingly in their timing of integration,their pace of export growth, their policiestoward import competition and foreign in-vestment, and the impact of regional tradingarrangements. It concludes that those regionsthat aspired to trade most with the globaleconomy became the most regionally inte-grated as well. Further, regional trade tends toprecede preferential trade agreements ratherthan the other way around.

27

Regional Trade and PreferentialTrading Agreements:A Global Perspective

2

The Proliferation of RegionalPreference Systems

More agreements are being signed. Since1990, the number of RTAs in force rose

from 50 to nearly 230 (figure 2.1). The WTO

estimates that another 60 agreements are invarious stages of negotiation. The boom inRTAs reflects changes in certain countries’trade policy objectives, the changingperceptions of the multilateral liberalization


28

Following WTO convention, the term regionaltrade agreement encompasses both reciprocal

bilateral free trade or customs areas and multicountry(plurilateral) agreements. Regional and bilateral tradeagreements provide for one type of trade liberaliza-tion, and they must be seen in a broader context ofalternative methods of liberalization. Members ofRTAs liberalize trade on a reciprocal and preferentialbasis. While programs such as the U.S. AfricanGrowth and Opportunity Act (AGOA) and the EU’sEverything But Arms (EBA) also liberalize tradepreferentially (i.e., different trade partners receivedifferent treatment), the United States and EU extendthese preferences unilaterally rather than reciprocally.In contrast to both of these types of preferential liber-alization, countries often lower trade barriers in anondiscriminatory fashion for all trade partners. Theymight do so multilaterally—through GATT/WTOnegotiating rounds—or autonomously, as in the caseof Pakistan in the late 1990s. The matrix belowillustrates this taxonomy of liberalization methods.

RTAs are commonly divided into several basiccategories, according the degree of economic integra-tion they provide. The canonical taxonomy of RTAscontains the following four levels of integration:

1. In a Free Trade Area, members eliminatebarriers to trade in goods (and increasingly services)among members, but each member is free to maintain

Box 2.1 RTAs and types of trade liberalizationdifferent MFN barriers on nonmembers. This lattercharacteristic requires members to develop rules oforigin to prevent imports from third countries frombeing transshipped through the member country withthe lowest tariffs.

2. A Customs Union moves beyond a free tradearea by establishing a common external tariff on alltrade between members and nonmembers. Customsunions typically contain mechanisms to redistributetariff revenue among members.

3. A Common Market deepens a customs unionby providing for the free flow of factors of produc-tion (labor and capital) in addition to the free flowof outputs.

4. In an Economic and Monetary Union, membersshare a common currency and macroeconomicpolicies.

The international experience with RTAs is muchricher than this simple taxonomy suggests. NAFTAandother more recent agreements establishing freetrade areas contain provisions governing domesticlabor standards and other regulatory issues, whichone traditionally associated with agreements fordeeper integration. On the other hand, many freetrade agreements exclude important categories ofgoods (notably agriculture) from trade liberalization.In some cases customs unions still levy tariffs ontrade between members.

Source: World Bank staff.

Method of implementation

Scope of beneficiaries Reciprocal Unilateral

Preferential: selected countries NAFTA, EU, COMESA, GSP, AGOA, EPAs, and other RTAs EBA, Cotonou

Nondiscriminatory GATT/WTO (MFN): all countries multilateral agreements Autonomous liberalization

process, and the reintegration into the globaleconomy of countries in transition from social-ism. This last category accounts for many of thenew agreements signed in the early 1990s, whencountries in Eastern Europe and the formerSoviet Union negotiated RTAs with WesternEurope [both the EU and the European FreeTrade Association (EFTA)] and with each other.

Some of the RTAs included in figure 2.1have never been reported to the WTO, for anyof several reasons. One reason is that theWTO does not enforce notification (the sameis true of notification requirements in otherWTO agreements). Another is that severalcountries that have yet to join the WTO havebeen quite active in forming RTAs. Russia, forexample, is in the process of joining the WTOand has signed bilateral free trade agreements(FTAs) with other members of the Common-wealth of Independent States (CIS). It is also

pursuing two regional arrangements that aredesignated to become customs unions: theEuroasian Economic Community and theSingle Economic Space. Because a consistentdata source covering all RTAs is lacking, dataare based on the information contained in theWTO database, supplemented by data fromthe major unreported agreements.

Most countries are participatingNearly all countries belong to at least oneRTA,2 and some are party to numerousagreements (table 2.1). On average, eachcountry belongs to six RTAs, though there isconsiderable variation across regions andlevels of development. East Asian countriessign fewer agreements than countries inother regions. Northern countries have par-ticipated to the greatest extent, each signing,on average, 13 agreements. A substantial

R E G I O N A L T R A D E A N D P R E F E R E N T I A L T R A D I N G A G R E E M E N T S

29

Sources: WTO data and WTO staff.

Figure 2.1 Number of RTAs exploded in the 1990s

300

250

200

150

100

50

01958 1989198419761969 1994 20041999

0

5

10

15

20

25

30

Cumulative number of agreements(EU-15 counted as single country)

Agreements notnotified to the WTO

Agreements notifiedto the WTO

Cumulative number of agreements(EU-25 counted as single country)

RTAs represent a fundamental departure from thecore WTO principle of nondiscrimination.

Nonetheless, the WTO affords its members a largedegree of flexibility in entering new RTAs. Withinthe WTO mandate, countries may join agreementsby meeting the requirements of the General Agree-ment on Tariffs and Trade (GATT) Article XXIV

Box 2.2 Reporting RTAs to the WTOcovering the formation of customs unions and freetrade areas in merchandise trade; the General Agree-ment on Trade in Services (GATS) Article V onagreements in services; or the Enabling Clause (the1979 Decision on Differential and More FavorableTreatment, Reciprocity, and Fuller Participation of

(Box continues on next page)


30

Developing Countries) dealing with trade in goodsbetween developing countries.

Countries are required to notify the WTOsecretariat of any RTAs they enter into, and their pro-visions are subject to review by the WTO. However,the review process in practice ends with the commit-tee’s queries of the parties and has not led to a subse-quent report to the membership or formal WTOendorsement in any case. Furthermore, even the notifi-cation requirement, though a formal rule, has enjoyedonly inconsistent compliance. An examination of theWTO RTA database reveals very large gaps betweenthe date agreements are signed and the date they arereported to the WTO. Several agreements with WTO

Box 2.2 (continued)

members have yet to be reported to the organization.These include the Greater Arab Free Trade Area, theAghadir Agreement in the Middle East and NorthAfrica, the India-Nepal and India-Bhutan agreementsin South Asia, and several agreements in Africa. Someagreements, such as EU accessions, are reported morethan a year in advance, while other agreements arenotified six or more years after entry into force. Theaverage gap is 354 days. Excluding agreements noti-fied before signature, the average gap rises to 446. Themedian delays between entry into force and notifica-tion are 135 and 188 days, respectively.

Sources: World Bank and WTO staff.

Table 2.1 Most countries belong to more than one RTA

Europe Latin Middle East Asia and America East

and Central and the and North Sub-Saharan Pacific Asia Caribbean Africa South Asia Africa North Total

Number of countries 32 36 39 21 8 48 25 209North-South bilateral

Countries belonging to at least one RTA 4 12 6 10 0 2 10 44Average number of RTAs per country 2 1 2 1 1 4 2Maximum number of RTAs per country 4 4 4 3 0 1 24 24

All othersCountries belonging to at least one RTA 24 22 33 20 8 47 10 164Average number of RTAs per country 2 6 8 5 4 4 8 5Maximum number of RTAs per country 3 12 17 12 9 9 15 17

TotalCountries belonging to at least one RTA 26 26 35 20 8 48 11 174Average number of RTAs per country 2 6 8 5 4 4 11 5Maximum number of RTAs per country 7 12 19 13 9 9 29 29

Note: Bilateral agreements are defined as an RTA with two members. North is OECD 24 plus Lichtenstein, and South is all othercountries.Source: Published WTO data, World Bank staff.

number of developing countries (45) havesigned bilateral preferential agreements witha Northern partner. However, this activity isnot spread evenly across regions. Most ac-tivity has been in Eastern Europe, NorthernAfrica, and Latin America. There are nocountries in South Asia that have signed a

bilateral agreement with a Northern partner.The enlargement of the EU in May 2004 ledto a fall in the number of North-South RTAsin Europe.

Since 2000, several major new trends haveemerged in the pattern of regional tradeagreements. One unifying characteristic is

that these take RTAs well beyond agreementsbetween adjacent countries. For example,

• The EU’s move toward bilateral marketaccess FTAs and Economic PartnershipAgreements (EPAs) with the ACPcountries;

• The shift in the U.S. position towardbilateral preferential agreements; and

• The effort of a handful of developingcountries to open markets through RTAs.

We turn now to a more in-depth investiga-tion of these trends.

EU Preferential Trade ArrangementsDuring the 1990s, the EU was an active spon-sor of bilateral arrangements with individualcountries and groups of countries and was themajor player in the RTA game. Prior to therecent accession of 10 new members, the EUhad bilateral or regional agreements with 111countries. Trade agreements became an inte-gral instrument of European foreign policy,particularly in the aftermath of the collapse ofthe Soviet Union.3

Three types of agreements were intended tostabilize the region after 1989. Europe Agree-ments were intended to prepare borderingEastern European countries for eventual acces-sion into the EU. They involved bilateral agree-ments between each other and with the EU toreduce tariffs, develop uniform rules of origin,EU-consistent regulatory approaches to ser-vices, and common treatment of standards aswell as transition rules in sectors such as agri-culture. These efforts culminated with the fulladmission of 10 new countries into the EU in2004—which is why the number of RTAs reg-istered with the WTO fell for the first time ever.

Euro-Mediterranean Agreements were in-tended to build bilateral trade relations be-tween neighbors, with the objective of form-ing a NAFTA-like free trade area by 2010.Launched in 1995, the EU and 12 countrieshave been involved in talks on “associationagreements” that would subsume some exist-ing bilateral arrangements. To date, bilateralagreements have been signed with Tunisia

(1995), Israel (1995), Morocco (1996),Jordan (1997), the Palestinian Authority(1997), Algeria (2001), Egypt (2001), andLebanon (2002). In general, services liberal-ization provisions are limited to the restate-ment of WTO GATS commitments with nonew liberalization or with preferential accessreserved for suppliers based in member coun-tries. Dispute settlement is state-to-state basedon ad hoc arbitration.

Partnership and Cooperation Agreements(PCAs) with the Western Balkans, Russia, andthe CIS were designed to help promote stabil-ity on the border of the EU, and in the case ofRussia, expand trade. The EU has been pro-viding technical assistance to these govern-ments to help implement the institutional re-forms that are part of the PCAs.

Two new agreements have been added tothis list since 2000.

• Economic Partnership Agreements(EPAs) are designed to replace the pref-erential systems embodied in the Conto-nou Agreement (the successor to theLomé Convention), which had received awaiver under the enabling clause fromGATT Article XXIV, a waiver thatexpires in 2007. EPAs are designed topromote trade and development in theACP 77 countries in a WTO-consistentfashion by establishing agreements be-tween large groups of countries formingcustoms unions (box 2.3).

• Free Trade Agreements with SouthAfrica (which entered into force in2000), Mexico (2000), and Chile (2003)are designed to open markets and securetrade. Agreements with the Gulf Cooper-ation Council (GCC) and the CommonMarket for the South (MERCOSUR) areunder active negotiation. These embodyfree trade provisions for a range of prod-ucts as well as provisions to liberalize atleast some services (Ullrich 2004).

The EU agreements govern services trade inaddition to trade in goods. The agreements


31

with Mexico and Chile provide for specificliberalization commitments in the financialsector over and above those included inGATS, with the Chilean agreement addingtelecommunications and maritime services(see Ullrich 2004). The South African agree-ment alludes to possible services liberaliza-tion, but without commitment. The EU agree-ments differ in important respects from theU.S. agreements in that they are generally lesscomprehensive, provide less market access inagriculture, and do not provide for investor-state dispute resolution (see chapter 5).

The U.S. embraces bilateralismPrior to the present administration, the U.S. hadgenerally eschewed reciprocal preferentialtrade agreements, whether regional or bilateral.

Exceptions included only Canada and Israel inthe 1980s and NAFTA in the early 1990s. In-deed, many U.S. trade observers contend thatopening NAFTA talks was designed primarilyto support multilateral trade negotiations—tospur the Europeans and others into acting ontheUruguayRound.Twoyears later, theClintonadministration announced its desire to form aFree Trade Area of the Americas (FTAA), and itsigned an FTA with Jordan in 2000.

Since the approval of trade promotion au-thority in 2002, however, the United States hasgiven much greater emphasis to securing bilat-eral FTAs in tandem with its efforts to achievemultilateral liberalization through the WTO.Since 2002 the United States has signed bilat-eral accords with Australia, Bahrain, CentralAmerica plus the Dominican Republic, Chile,


32

EPAs are the most ambitious attempt to harnesstrade, development resources, and technical-legal

assistance to the cause of integration-led develop-ment. The objective is to promote development,strengthen regional integration, and ensure compati-bility with WTO principles. By negotiating reciprocalliberalization with existing South-South regionalgroupings and by providing common rules of originwith cumulative provisions, participants hope to pre-vent the hub-and-spoke effects that plague manybilateral North-South agreements. The EPAs will alsoencourage liberalization of services, provide for com-mon product standards, and set up the negotiation ofinvestor protections, based on state-to-state ad hocarbitration of disputes.

After a one-year clarification phase by the AfricanCaribbean and Pacific states (ACP), the first negotia-tions were launched in October 2003. The EU initi-ated discussions with the Economic Community ofWest African States (ECOWAS) plus Mauritania, theEconomic and Monetary Community of CentralAfrica (CEMAC) plus São Tomé, Eastern and

Box 2.3 EPAs become the EU’s trade anddevelopment instrument: An experiment in “North-South-South” integration

Southern Africa (16 countries), the Southern AfricanDevelopment Community (SADC), the CaribbeanACP countries, and the Pacific states (Kiener 2004).

The content for the agreements is currently openfor discussion. Reciprocal trade liberalization wouldbe the centerpiece under the terms of the EPA pro-gram . . . (Most of the EPA countries already enjoypreferential market access that the EU grants unilat-erally under this program.) In addition, the EU hasstated that it would like to have services liberaliza-tion, investment, competition, government procure-ment, and trade facilitation covered in the agree-ments (Falkenberg 2004).

Several issues will determine the ultimate effec-tiveness of the EPAs in promoting development: thedegree of additional MFN liberalization in goodsand services markets in both the RTAs and in theEU; the restrictiveness of the rules of origin forgoods; and the extent of trade diversion that couldoccur in the event that there are no concomitant re-ductions in MFN border protections (see Hinkle andSchiff 2004).

Morocco, and Singapore. The United Statesappears to have intensified its pursuit of RTAssince the Cancun WTO Ministerial (September2003). Negotiations are officially4 under waywith Colombia, Ecuador, Panama, Peru,SACU, and Thailand. Other economies deemedto be in the queue are Bolivia, Egypt, NewZealand, Pakistan, the Philippines, SouthKorea, Sri Lanka, and Taiwan (China), andUruguay (Schott 2004). This intensified pacemay reflect the intention to prod both the mul-tilateral negotiations and the FTAA, as well asto respond to U.S. businesses that fear beingshut out of export markets by a growing num-ber of RTAs in which the United States is not amember.

In the broadest of terms, developing coun-tries seek to provide access to their servicesmarkets and guarantees in many nontradeareas in exchange for assured access to U.S.goods markets. Key facets of these agreementsinclude:5

• Tariff rates on most nonagriculturalproducts are bound at zero; for example,the U.S.-Chile FTA will bind duties atzero for 85 percent of trade.

• Exclusion or delayed liberalization ofsensitive products, commonly includingagricultural products such as dairy prod-ucts, cotton, ethyl alcohol, peanuts and

peanut butter, sugar, and tobacco for theUnited States. Some exclusions are dueto be phased out according to lengthytimetables; in the Chile-U.S. FTA, for ex-ample, all duties will be phased out in12 years (USTR 2004).

• Intellectual property rights are conven-tionally accorded stronger protectionsthan under the WTO’s TRIPS agreement,with investor-state suits permitted in theevent of disputes.

• Investment protections, with provisionsfor national treatment and nondiscrimi-nation in pre-establishment provisionsfor companies based in each othersmarkets (though liberal rules of originindicate foreign subsidiaries located inmember countries qualify for eligibility).

• Services trade are to be open except forthose excluded in a negative list; notablyexcluded are labor service providers, ex-cept for the provisional visas held by pro-fessionals associated with investing firms.

• Labor and environment issues are in-cluded in recent agreements, with signa-tory countries undertaking commitmentsto enforce their own environmental andlabor laws. Dispute settlement panels areempowered to impose monetary finesrather than using trade sanctions to forcecompliance (box 2.4.)


33

Until NAFTA, the United States did not attemptto include provisions on labor in trade agree-

ments that it negotiated. As a presidential candidate,Bill Clinton promised to negotiate new side agree-ments to NAFTA on labor in order to secure suffi-cient political support for NAFTA. Since then laborissues have featured prominently in Congressionaldebates on granting the president negotiatingauthority and the resulting trade agreements.

All recent FTAs negotiated with the United Statescontain provisions requiring parties to enforce theirown labor laws. These are premised on the assump-

Box 2.4 Labor in U.S. FTAstion that each member’s existing laws are satisfactoryand therefore any trade distortions that might ariseare caused by a lack of enforcement. The agreementsenumerate five core standards: the right of associa-tion; the right to organize and bargain collectively;prohibitions on forced labor; a minimum age for em-ployment of children; and acceptable working condi-tions. The FTAs establish a procedure for makingcomplaints, encouraging resolution first through con-sultation and, if this fails, by establishing a panel ofexperts to hear the dispute.


The United States indicated it would not nego-tiate changes in its antidumping statutes or onits agricultural subsidies, insisting on address-ing both through the WTO’s multilateral nego-tiations. In chapter 5, we return to a deeper dis-cussion of provision for services, investment,and intellectual property rights (IPR).

Developing countries actively pursuemajor marketsThe launching of NAFTA spawned a newflurry of interest among developing countrieseager to use RTAs to secure market access.Mexico and Chile have been at the forefront ofthese developments. Mexico, having created aworld-class trade negotiating team for NAFTA,turned its attention to Central America andother countries in Latin America. It establishedarrangements with Costa Rica (1995), Bolivia(1995), Nicaragua (1998), the EU (2000),EFTA (2001), and Japan (2004). After NAFTAwas signed, Chile immediately solicited entryinto the accord. Rebuffed initially, the countryembarked on a wider strategy. Chile estab-lished agreements with MERCOSUR (1996),Canada (1997), Peru (1998), Mexico (1999),Central America (2002), the United States andEU (2003), and EFTA (2004). By 2004, Chilehad signed free trade agreements that provided

over 60 percent of its exports with duty-freeaccess to markets around the world (see Devlinand Estevadeordal 2004).

Many existing regional organizations inAfrica also moved aggressively to intensifypreferential trade liberalization during the1990s. For example, the treaty establishingthe Common Market for Eastern and South-ern Africa (COMESA), which was signed in1993 to replace the Preferential Trade Area,called for a free trade area by 2000 and acustoms union by 2004. The East AfricanCommunity was formed in the mid-1990s toaccelerate economic integration among threeCOMESA members (Kenya, Tanzania, andUganda). The SADC Trade Cooperation Pro-tocol was signed in 1996 as part of an effort toreintegrate South Africa into the regional econ-omy after the end of apartheid.

Asian countries have launched similar nego-tiations since 2001. India has concluded or isnegotiating limited arrangements with MER-COSUR and Thailand; MERCOSUR is negoti-ating with the Andean countries; China haslaunched bilateral accords with members of theAssociation of Southeast Asian Nations(ASEAN), to mention a few. In 2004, India,Pakistan, and other South Asian countries an-nounced the South Asian Free Trade Agreement


34

The labor provisions break new ground in how adispute settlement panel’s decisions are enforced.Rather than using trade remedies (i.e., granting theinjured party the right to withhold trade conces-sions), a panel can impose a monetary fine of up to$15 million per year (adjusted for inflation). Pay-ments of the fines would go into a fund to supportappropriate labor initiatives, which may includeefforts to improve enforcement of labor laws. Thismechanism appears in the agreements that theUnited States has signed with Australia, Bahrain,Chile, Central America and Dominican Republic,Morocco, and Singapore.

Box 2.4 (continued)

Using fines rather than trade sanctions has severaladvantages: while trade sanctions penalize both pol-luting and clean exporters, fines target the polluters;increased trade sanctions hurt all workers in exportindustries, but fines help restructure plants andmaintain employment; and fines build in targetedsolutions to the problem rather than present pro-tracted trade disputes.

Sources: Destler and Balint 1999, texts of FTAs on the USTRweb site (www.ustr.gov), and Weintraub 2004.

(SAFTA), which is intended to encompass all ofthe countries of the region (see Baysan 2004;Newfarmer 2004).

Many RTAs, diverse provisionsRTAs have increasingly been designed to covermuch more than liberalization of tariffs andquotas. New provisions on enforcement of do-mestic labor and environmental laws have al-ready been mentioned. Table 2.2 gives a flavorof the range of services and intellectual prop-erty rights issues that are addressed in currentagreements. Many of these issues, which aredealt with in more detail in later chapters,have implications for trade, although the pre-cise mechanisms by which trade is affected arenot always well defined.

Many RTAs, many rationalesThese recent trends highlight different ratio-nales that drive the quest for preferential

agreements, but in nearly all cases politics is asimportant as economics. The classic North-North agreement, the European Union, had itsorigin in politics (see Schiff and Winters2003). The fathers of the European Commu-nity, Robert Schumann and Jean Monnet,clearly believed that Franco-German integra-tion through trade and investment wouldproduce a new constellation of common eco-nomic interests that would attenuate historicmilitary animosity. As a first step, they feltthat placing French and German coal and steelindustries under a single authority, theEuropean Coal and Steel Community, wouldmake it impossible for either of these histori-cal enemies to use these resources for militarypurposes against the other.

Today, for North-South agreements,Northern partners often have a complex mixof rationales—rooted in foreign policy, com-mercial diplomacy, and development policy.


35

Table 2.2 RTAs cover many topics besides merchandise trade

Customs Intellectual DisputeStandards Transport Cooperation Services Property Investment Settlement Labor Competition

U.S.U.S.-Jordan No No Yes Yes Yes Yes Yes Yes YesU.S.-Chile Yes No Yes Yes Yes Yes Yes Yes YesU.S.-Singapore Yes No Yes Yes Yes Yes Yes Yes YesU.S.-Australia Yes No Yes Yes Yes Yes Yes Yes YesU.S.-CAFTA Yes No Yes Yes Yes Yes Yes Yes NoU.S.-Morocco Yes No Yes Yes Yes Yes Yes Yes NoNAFTA Yes No Yes Yes Yes Yes Yes Yes Yes

EU†

EU-South Africa No No No No Yes No Yes No YesEU-Mexico Yes* Yes Yes* Yes Yes Yes Yes No YesEU-Chile Yes* Yes Yes Yes Yes Yes Yes No YesEuro-Med.

Agreements No No No No Yes No Yes No Yes*South-South

MERCOSUR Yes Yes Yes Yes No Yes Yes YesAndean Community Yes Yes Yes Yes No Yes Yes Yes YesCARICOM Yes Yes Yes Yes No Yes Yes Yes YesAFTA Yes Yes Yes Yes No Yes No No NoSADC Yes Yes Yes No YesCOMESA Yes Yes Yes Yes No Yes Yes Yes Yes

OtherJapan-Singapore Yes No Yes Yes Yes Yes Yes Yes YesCanada-Chile No No Yes Yes No Yes Yes Yes YesChile-Mexico Yes Yes Yes Yes Yes Yes Yes Yes

Source: World Bank staff.†While EU agreements mention cooperation in most of the subject areas, only those in which specific commitments are under-taken receive a “Yes” rating.*Implementation steps are to be agreed on at a later date.


36

It is important to establish coherent relationshipsbetween environmental policies and the trade

obligations set out in various RTAs. The followingexamples illustrate the various ways that environ-mental issues are handled in these trade agreements.

WTO. Within the WTO, environmental provisionsare limited to the adoption of product-related mea-sures as “necessary to protect human, animal or plantlife or health,” or “relating to the conservation of ex-haustible natural resources.” Process-related require-ments continue to remain outside the scope of the

Box 2.5 Trade agreements and the environmentWTO. However, in the absence of agreed-on interna-tional standards (e.g., fisheries), the risk of disguisedprotectionism has prevented further consensus on theway forward. Long-standing disputes between theUnited States and other countries on tuna fishing anddolphin or turtle protection are cases in point.

NAFTA. The environmental agreement underNAFTA created the Commission for EnvironmentalCooperation to promote environmental cooperationamong the three members. The commission itself does

“Trade policy has always been the principalinstrument of foreign policy for the EuropeanUnion” (Sapir 1998). The United States nowappears to be using preferential agreementsfor reasons that are similarly broad. Both theEU and the United States seek trade agree-ments that go beyond simple tariff removal toinclude rules governing services, protection ofintellectual property, and adherence to health,labor, and environmental standards.

One goal of developing countries seekingan RTA with a large market, such as the EU6

or the United States, is simply to secure marketaccess. One should note, however, that mostdeveloping countries, especially the least de-veloped countries (LDCs), already enjoy con-siderable access to these markets for mostmanufactured products (whether through uni-lateral preference programs or because MFNtariffs are already quite low), and RTAs withthese countries often exclude agriculture andother politically sensitive products. Neverthe-less, RTAs provide some insurance againstfuture protectionist policies, and by reachingan agreement “preemptively,” they seek toavoid being left out of a future agreement.

A second objective is to reinforce internalregulatory reforms through external treatyobligations and visible political commitments.Locking in domestic reforms through a for-eign trade agreement with the EU clearly

motivated countries making the transitionfrom socialism in the 1990s. Mexico underNAFTA was motivated by a similar objective.Guaranteed market access combined withcredible domestic reforms can attract foreigndirect investment (see chapter 5).

South-South agreements often reflect a po-litical desire to form or join a broadly basedregional initiative, such as ASEAN, COMESA,or MERCOSUR. The drive for economicintegration often begins with political objec-tives. Like France and Germany in the 1950s,the newly established democracies of theSouthern Cone formed MERCOSUR in themid-1980s in the hopes of damping the tradi-tional military hostility between major re-gional powers—Argentina and Brazil. SADCoriginated in the 1980s as a coalition opposedto apartheid in South Africa and has morerecently turned to creating a free trade area.Some observers note that African customsunions and free trade areas are as active inareas such as conflict resolution as in tradeliberalization. Finally, many see relaxed ten-sions between India and Pakistan as the realpayoff from the proposed SAFTA agreement,regardless of what happens to trade barriersin the region. The tentative conclusion of ex-isting studies is that RTAs that expand tradeflows appear to have a substantial dampeningimpact on conflict (box 2.6).


Not all political objectives involve war andpeace issues; some South-South agreementsare designed to pool resources for trade nego-tiations and trade policymaking. Much as theEuropean Union established a common tradepolicy with a common commissioner in chargeof trade (in part to negotiate more forcefullywith the United States in the GATT), so too adriving force for MERCOSUR was to estab-lish a common trade policy relative to the mul-tilateral and hemispheric system.

Entering into a regional agreement may alsoreflect a desire to deal with region-specific is-sues—such as transit, water, energy, migration,

movement of labor, customs, and standards—that are difficult to broach at the global level.RTAs among CIS countries are arguably anattempt to reconstruct some of the economiclinkages that were severed with the disintegra-tion of the Soviet Union and the disorganiza-tion caused by the collapse of central planning.Although many of these regional externalitiescan be handled without a trade agreement,RTAs may provide institutions and a frame-work through which to make progress on theseissues (see chapter 4).

The wide variation in RTAs flows from thevery different motivations countries have for


37

not set standards in the various countries, though partof its mandate is to help harmonize them upward. If acountry persistently fails to enforce environmentallaws that have conferred a trade benefit, dispute settle-ment provisions can be invoked. The commission’srole in the disputes is to see that enforcement of exist-ing laws takes place. In addition, it is charged withmonitoring the environmental effects of NAFTA. Arti-cles of agreement also dictate that countries will nottry to attract investment by relaxing or ignoring do-mestic health, safety, or environmental regulations. In-ternational environmental agreements recognized bythe three parties take precedence over national rules.

MERCOSUR. Environmental concerns are cur-rently being dealt with in MERCOSUR by a workinggroup. This group has discussed issues such as theenvironment, competitiveness, non-tariff barriers totrade, and common systems of environmentalinformation. A draft agreement from this workinggroup provides for upward harmonization of envi-ronmental management systems and increased coop-eration on shared ecosystems, in addition to mecha-nisms for social participation. It also includesprovisions on instruments for environmental man-agement, including quality standards, environmentalimpact assessment methods, environmental monitor-ing and costing, environmental information systemsand certification processes, provisions for protectinghealth and quality of life, and other general

Box 2.5 (continued)

mechanisms for implementing the protocol. Theregime is still evolving, and the challenge at hand isto ensure that the promise of the protocol leads toeffective regional cooperation and action.

Bilateral agreements. A number of recently con-cluded bilateral FTAs, including the U.S.–SingaporeFTA and the Japan–Singapore Economic Agreementfor a New Age Partnership, contain environmentalprovisions. The U.S.–Singapore FTA establishes animportant precedent for dealing with environmentalissues by including a chapter specifically on the envi-ronment. As discussed in box 2.4 on labor laws, thisagreement ensures that countries effectively enforcetheir environmental laws, and it provides for en-forcement mechanisms, including fines.

Even in the absence of such special provisions,however, trade agreements can contribute to acleaner environment simply by making trade moreresponsive to market forces. In general, countriesthat are more open to trade adopt cleaner technolo-gies more quickly, and increases in real income areoften associated with greater demand for environ-mental quality (WTO 1999). Opening up domesticmarkets also encourages cleaner manufacturing,because protectionist countries tend to shelterpollution-intensive heavy industries. The incentivesto over-exploit or deplete resources are more directlyrelated to policies and institutions within the sectorthan to trade openness per se (World Bank 1999).


38

Does trade inhibit or increase hostilities betweenstates? Greater contact among traders and con-

sumers across borders may stimulate mutual respectand more harmonious relations, and high levels oftrade can create economic interdependence, which, inturn, raises the cost of political disputes and militaryconflict.

In 1889, Wilfred Pareto suggested that “customsunions and other systems of closer commercial rela-tions [could serve] as means to the improvement ofpolitical relations and the maintenance of peace.” In1919, John Maynard Keynes wrote that “a FreeTrade Union, comprising the whole of Central,Eastern and South-Eastern Europe, Siberia, Turkey,and (I should hope) the United Kingdom, Egyptand India, might do as much for the peace and pros-perity of the world as the League of Nations itself.”

RTAs also can provide institutions and a forumfor bargaining and negotiation—to address tensionsbefore they erupt in conflict. European integration,ASEAN, and MERCOSUR are often cited as venuesfor improving political-military relations. Regionaltrade agreements do not ensure positive political out-comes, however. The U.S. civil war (1861–65) wasfought—at least in part—over high protection ofnorthern manufactures and trade restrictions on cot-ton. Similarly, the Central American soccer war of1969 emerged out of lingering hostility over tradearrangements that created advantages for El Salvadorat the expense of Honduras. And one reasonBangladesh seceded from Pakistan was the common

Box 2.6 Can RTAs prevent conflict?external tariff structure that deprived it of access tocheaper inputs from the global market and divertedtrade to Pakistan (Schiff and Winters 2003).

Mansfield and Pevehouse (2000) attempt to iden-tify empirically the role of RTAs in amelioratingconflict. They find that, on average, the likelihoodthat a pair of states will see the outbreak of a mili-tarized interstate dispute declines by around 50 per-cent if both belong to the same RTA. However,only RTAs that expand trade flows appear to havea substantial impact on conflict. When evaluated atthe lowest level of trade between partners, it ap-pears that membership in a RTA reduces the chanceof dispute by just 15 percent. Other studies havesuggested that RTAs that have little impact on trademay actually exacerbate conflict (see Powers 2003).If the gains from trade are not distributed evenly,for example, then the subsequent change in inter-state power relations can be a source of increasedtension. Also, rising interdependence may be seen asa source of increasing vulnerability, making expan-sion through military force appear more attractive.

These results, which suggest that RTAs could con-tribute to a reduced risk of military conflict, should betreated with a high degree of caution, due to problemsof causality and omitted factors, such as the broaderinstitutional framework governing relations betweenparticular pairs of countries. In Africa, for example,RTAs that address the management of cross-borderresource issues (such as water) are more effective inreducing military conflict than other RTAs.

entering into the arrangements. As we will seein subsequent sections, these motivationscontribute to greater complexity in rulesgoverning world trade.

Many RTAs can complicateadministrative proceduresAn important feature of the rise in the num-ber of RTAs is the growing number of over-lapping agreements and the so-called“spaghetti bowl” that has emerged from theproliferation of bilateral agreements (fig-ure 2.2). The associated myriad of rules

strains institutions charged with administer-ing trade agreements. A web of differing tradearrangements can tangle administrative pro-cedures—customs procedures, technical stan-dards, rules of origin, and so on—and therebyraises the costs for both enterprises and gov-ernments. This complexity undermines worktoward greater trade facilitation in developingcountries.

Many agreements between country pairsare duplicated by other agreements to whichthe same two countries are parties. In Sub-Saharan Africa, for example, about one-half of


39

AMU:CBI:CEMAC:CILSS:COMESA:EAC:ECOWAS:IGAD:IOC:SACU:

Arab Maghreb UnionCross Border InitiativeEconomic and Monetary Community of Central AfricaPermanent Interstate Committee on Drought Control in the SahelCommon Market for Eastern and Southern AfricaEast African CooperationEconomic Community of Western African StudiesInter-Governmental Authority for GovernmentIndian Ocean CommissionSouthern African Customs Union

Figure 2.2 Spaghetti and rigatoni: Multiple, overlapping RTAs, 2004a. African agreements are overlapping

AlgeriaLibyaMoroccoMauritaniaTunisia

AMU

GhanaNigeria

Cape VerdeGambia

ECOWAS

BeninTogoCôte d’Ivoire

NigerBurkina Faso

Conseil deL’Entente

Guinea-Bissau MaliSenegal

WAEMU

LiberiaSierra Leone

Guinea

Mano River Union

CILSS

CameroonCentral African Rep.GabonEquat. GuineaRep. Congo

Chad

São Tomé & Principe

ECCAS CEMAC

Angola

Burundi*Rwanda*

Egypt

DR Congo

DjiboutiEthiopiaEritreaSudan

Kenya*Uganda*

Somalia

Tanzania*

EAC

South AfricaBotswanaLesotho

Namibia*Swaziland*

Mozambique

SACU

Malawi*Zambia*Zimbabwe*

Mauritius*Syechelles*

Comoros*Madagascar*

Reunion

IOC*CBI

SADC

COMESA

Nile River Basin IGAD

b. “Spaghetti Bowl” of RTAs in the Americas and Asia-Pacific (Agreements signed and in force in Latin America as ofMay 2004)

Source: Devlin and Estevadeordal 2004.

Source: Schiff and Winters 2003.

FTAA Trans-Pacific signedIntra-Asia-Pacific signedIntra-Asia-Pacific in force Intra-Americas in forceAPEC

Canada

USA

Mexico

Chile

Uruguay

Paraguay

BrazilArgentina

MERCOSUR

Bolivia

Colombia

Venezuela

Peru

Ecuador

CostaRica

Nicaragua

El Salvador

HondurasGuatemala

CACM

Dominica Trinidad and Tobago

Guyana Antigua and BarbudaJamaica St. Vincent and Grenadines

Suriname Grenada Barbados

St. Kitts and Nevis Belize

CARICOM

Panama

DominicanRepublic

AndeanCommunity

Bahamas

Haiti St. Lucia

Brunei

Cambodia

Thailand

Laos

Malaysia

Philippines

Myanmar

Singapore

Indonesia

Vietnam

Japan

New Zealand

Australia

ASEAN

Korea

PR China

Hong Kong

Taiwan

Russia

Papua New Guinea APEC

FTAA

SADC:WAEMU:* Indicates membership in CBI regional grouping

Southern African Development CommunityWest African Economic and Monetary Union


40

BAFTA:CEFTA:CIS:* Prior to EU expansion.

Baltic FTACentral European FTACommonwealth of Independent States

Figure 2.2 (continued)

c. In Eastern Europe and Central Asia, bilateral agreements burden customs officials*

Ukraine

Moldova

Latvia BAFTA

Turkmenistan

Georgia

Russia

Kyrgyz Republic

Kazakhstan

Tajikistan

Belarus

CIS

CEFTA

AzerbaijanBosniaCroatia

ArmeniaMacedoniaSlovenia

UzbekistanHungary

Romania

EstoniaPoland

LithuaniaSlovak Republic

Czech Republic

TurkeyBulgaria


the pairwise trade relationships covered by anRTA are also covered by another agreement.In other regions, overlapping agreements alsocomprise a substantial share of the total num-ber of agreements. There would be significantbenefits, in terms of lower administrative costsand more effective implementation, from arationalization of the current structure ofoverlapping agreements.

Uneven terms—hub-and-spoke integrationThe substantial number of bilateral agree-ments involving large northern countries, mostof which have been signed since 1990, suggeststhat a hub-and-spoke structure in world tradeis emerging. Of the 109 North-South bilateralagreements, 86 have been created since 1990.In a hub-and-spoke trading system, the largestmarkets sign individual agreements with awide range of peripheral countries amongwhich market access remains restricted. Suchagreements can marginalize the spokes, wheremarket access conditions are usually less

advantageous than in the hub, which enjoysimproved access to all of the spokes. In com-parison with a broad preferential trade agree-ment, a hub-and-spoke approach in theorygenerates lower gains, which accrue mainly tothe hub (Wonnacott 1996). Hubs and spokesare already clearly discernible as the EU andUnited States extend restrictive rules of originfrom one bilateral agreement to another.7

Trade within RTAs is rising butpreferential trade is less importantTrade between RTA members is growing asthe number of agreements increases, and one-third of world trade now takes place betweenRTA members (figure 2.3). (Here we coveronly reciprocal agreements and exclude tradeunder the Generalized System of Preferences,Cotonou Agreement, and AGOA.) Disregard-ing intra-EU trade, bilateral flows betweenRTA members have been growing at a ratesimilar to the growth rate of agreementsthemselves, as shown in figure 2.3. This figure


41

Examining total trade flows between RTApartners overstates the amount of trade thattakes place on a preferential basis. However,tariff schedules of many RTA membersincreasingly contain duty-free MFN rateson which no preference can be given. Weestimate that the amount of preferential tradeamong RTA members, after accounting forMFN rates of zero, is much lower at 21 per-cent of world trade (figure 2.3). Furthermore,it is often more profitable for enterprises topay a low MFN tariff when there are highcosts to satisfy rules of origin or otheradministrative procedures that a trader mustfollow to qualify for preferential treatmentunder an RTA. If we exclude trade coveredby facing tariffs of 3 percent or less, we con-clude that, at present, the amount of globaltrade taking place under an economicallymeaningful tariff preference is around 15percent. An earlier estimate on a similar, butnot directly comparable, basis suggests thatin the mid-1990s, trade on a preferentialbasis amounted to 27 percent of globaltrade.8

While the number of RTAs has been in-creasing, the importance of preferential tradehas been falling, reflecting lower tariff barri-ers, especially in OECD countries. Since 1996the number of zero duty lines in the EU tariffschedule has increased from 13 to 21 percentof the total number of tariff lines and from 18to 32 percent for the United States. In 2002about 45 percent of the tariff lines in the EUand United States schedules had duties of3 percent or less. This reflects the impact ofmultilateral liberalization under the Uruguayand earlier trade rounds. Thus a large andgrowing proportion of EU and U.S. importsfrom preferential trade partners is unlikely toactually receive preferential access relative toother countries.

For many developing country agreements,the situation is different because the number oflow-duty tariff lines is small. In 2002, 6 percentof Brazilian tariff lines had MFN tariff rates ofzero, as did 1 percent of Indian tariff lines. Weestimate that 88 percent of trade between

1990 1996 20020

15

10

05

25

20

30

35

1990

a. Number of RTAs signed

b. Share of world trade covered

c. RTAs may be less than they appear

1996 2002

Figure 2.3 Trade within RTAs

0

100

50

150

200

250

USA

S-S

EU

USA

S-S

EU

Includingall trade

Note : EUN is EU-15.

Share of world trade covered (2002), %

Excluding0% MFN

Excluding�3% MFN

0

5

10

15

20

25

30

35


also reveals that the rapid increase in the num-ber of South-South RTAs signed in the pastdecade has not been matched by much changein trade flows among parties to these agree-ments. This discrepancy highlights a pointmade earlier: many new South-South agree-ments overlap with existing agreements.


42

countries in Latin America is potentially eligi-ble for preferential treatment under an RTA.9

For the Middle East and North African coun-tries it is 83 percent of the total. The newSAFTA will lead to three-quarters of the tradebetween members taking place on a preferen-tial basis (assuming all products are included).East Asia is an exception, where, for example,22 percent of Indonesian and 59 percent ofMalaysian tariffs are zero. Thus the amount oftrade between East Asian countries receivingtariff preferences is very small. Like OECDcountries, however, developing countries havetaken great strides to reduce MFN tariffs dur-ing the past two decades. Most of this liberal-ization has come from autonomous reductionsand not through trade agreements—eitherRTAs or multilateral trade negotiations (seebox 2.7).

Trends in Trade and Growthby Region

These agreements were superimposed in acontext of deep changes in global trading

patterns.10 The postwar period has seen majorglobal shocks and changes in the economicenvironment, including oil crises in the 1970sand financial crises in the 1980s and 1990s. Inthe past 20 years there have also been majorchanges in policy regimes. Socialist countriesacross the world restructured their economicsystems and started the process of reorientingtheir trade to the world economy. In the formerSoviet Union, this meant collapse and recon-struction; in East Asia it meant progressive,sustained, and profound institutional change.Latin America went through its own, if lessdramatic, transition from import-substitution

How does liberalization in RTAs compare to au-tonomous and multilateral liberalization? The

rapid expansion of RTAs has occurred during a pe-riod when developing countries were undertakingautonomous liberalization and also fulfilling commit-ments made during the Uruguay Round of theGATT. An examination of tariff reductions by devel-oping countries finds that neither RTAs nor multilat-eral negotiations represent the largest driver of liber-alization. Autonomous liberalization accounts for thelion’s share of trade liberalization since the 1980s.The trade-weighted average MFN tariff rate leviedby the 33 largest developing country importers(which collectively account for 90 percent of all de-veloping country imports) was 29.9 percent in the1980s. By 2003 the average MFN rate had droppedto 11.3 percent. Based on tariff concessions grantedduring the Uruguay Round, multilateral negotiationsaccount for 5.1 percentage points of the total declinein MFN tariffs, and the remaining 13.5 percentagepoints resulted from autonomous liberalization. Ifthe RTAs that these 33 countries have signed werefully implemented, the trade-weighted average ap-plied tariff would fall further to 9.3 percent. The

Box 2.7 Regional versus multilateral and unilateral liberalization:What’s more important?

chart below shows how trade liberalization is allo-cated according to these different sources.

Autonomous Liberalization66%

MultilateralAgreements

25%

Regional Agreements

10%

Share of total tariff reduction, by type of liberalization, 1983–2003

Source: Martin and Ng 2004.


43

1980

1990

2000

Non-oil export share of GDP (percent)

b. Exports became the driving force of growth everywhere

15

10

5

0

20

25

30

35

40

1990–2000

1961–1970

1970–1980

1980–1990

East A

sia

and

Pacific

Europ

e an

d

Centra

l Asia

Latin

Am

erica

and

Caribb

ean

Midd

le Eas

t and

North

Afri

ca

South

Asia

Sub-S

ahar

an

Africa

East A

sia

and P

acific

Europ

e and

Centra

l Asia

Latin

Amer

ica

and C

aribb

ean

Middle

East a

nd

North

Africa

South

Asia

Sub-S

ahar

an

Africa

GDP growth rates by region (percent)

a. East Asia and South Asia grew rapidly in 1980–2000,but other regions slowed

Figure 2.4 Trade performance has differed across regions

1

0

�1

�2

2

3

4

5

6

7

8

Source: World Bank staff using WITS for non-oil/total exports ratio.


Note: Data obtained from GTAP release 5.4 up to 1998, thefollowing years are projected from 1998 using regional growthrates calculated from COMTRADE.

Sources: GTAP 5.4 and COMTRADE.

1980

Regional exports as share of world exports (percent)(oilexcluded)

0

2

4

6

8

10

12

14

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

c. East Asian exports grew fastest, and all but South Asiaand Africa increased their share of the global market

East Asiaand Pacific

Sub-Saharan Africa

South Asia

Europe andCentral Asia

Latin Americaand Caribbean

Middle Eastand North Africa

Intra-regional trade as a share of GDP (percent), 2002

d. Integration with the world proceeded hand in handwith regional integration

15

10

5

0

20

25

26.5

15.3

6.4

3.5

0.8

5.3

30

Source: COMTRADE.

East A

sia

and

Pacific

Europ

e an

d

Centra

l Asia

Latin

Am

erica

and

Caribb

ean

Midd

le Eas

t and

North

Afri

ca

South

Asia

Sub-S

ahar

an

Africa

industrialization to a strategy of outward-oriented growth. Apartheid in South Africa,political strife in various parts of the conti-nent, and the struggle against HIV/AIDS de-layed the establishment of stable policies and

depressed growth throughout Africa. It hasbeen a period of major transitions.

The growth performance of developingcountries in the past 20 years reflects thisvaried experience (figure 2.4). Only the East


44

Asia and Pacific and South Asia regions expe-rienced higher GDP growth rates in the1980–2000 period than during 1960–1980.The other four regions fared worse in the lasttwo decades, with GDP growth rates thatwere one-half to two-thirds smaller thanbetween 1960 and 1980.

Nonetheless, trade grew. The share of tradein GDP grew in all regions in the 1990s. EastAsian exports grew faster than the other re-gions, and the region increased its share oftotal world exports throughout the 1980s and1990s. Latin American exports also grew con-sistently as a share of the world market duringthe 1990s, but not as steeply as for East Asia.In South Asia, however, although GDP growthincreased in the 1980–2000 period and the ex-port share of GDP rose in the latter part ofthat period, the trade growth is from a muchsmaller base. South Asia still has the lowesttrade shares of any region. Sub-SaharanAfrica has also had disappointing growth

performance. These trends reflect different ini-tial conditions and external shocks, changes indevelopment strategies, and policies towardtrade liberalization.

East Asia generally followed a strategy ofexport-led growth, through elimination ofanti-export bias and sustained nondiscrimina-tory (MFN) liberalization. This trend wasmost dramatic in China, where border barri-ers at the start of the decade were prohibitivefor most tariff lines, and by the end of thedecade averaged less than 12 percent, with aplan to go to less than 10 percent by 2004.But virtually all countries were liberalizing.Average tariffs for the region as a whole fellfrom about 23 percent in 1989–90 to 8 per-cent in 2003.

In Europe and Central Asia the disintegra-tion of the Soviet Union after 1990 caused thecollapse of growth and required profound re-structuring of the region’s economies and aredirection of its trade. Roughly half of the


1989–1990

2003

1996–1998

Applied average unweighted tariffs

e. One reason: East Asia liberalized early in trade

30

20

10

0

40

50

60

70

80

Source: TRAINS.

East A

sia

and

Pacific

Europ

e an

d

Centra

l Asia

Latin

Am

erica

and

Caribb

ean

Midd

le Eas

t and

North

Afri

ca

South

Asia

Sub-S

ahar

an

Africa

1980s

1990s

2000s

FDI, net inflows (percent GDP)

f. Strong FDI flows coincided with relocation ofproduction processes

1.0

0.5

0

1.5

2.0

2.5

3.0

3.5

Note: Period averages for 1980–1989, 1990–1999, and 2000–2002 calculated from annual data. For all regions except LAC,some observations are missing.

East A

sia

and

Pacific

Europ

e an

d

Centra

l Asia

Latin

Am

erica

and

Caribb

ean

Midd

le Eas

t and

North

Afri

ca

South

Asia

Sub-S

ahar

an

Africa

Sources: World Bank, WDI.


45

region was drawn to the magnet of the EU’slarge and stable market. The EU respondedwith technical assistance and a political will-ingness to admit its Eastern European neigh-bors as full members. The combination of apolitical framework, trade, investment, andtechnical assistance led to an unprecedentedpace of reforms and economic integration thatculminated with 10 states joining the EU onMay 1, 2004. With their eyes turned towardmarkets in the EU, the Central European andBaltic countries achieved more extensive inte-gration and higher trade and FDI flows, whichis evident in the rapid export growth of theregion as a whole.

The CIS has moved much more slowly inits process of reform and reorientation, partic-ularly in Central Asia and the Caucasus.Under the CIS-7 initiative, trade regimes havebeen generally liberalized, but have been lim-ited by regional trade and transit barriers.

Latin America reversed its trade policystance, and during the 1990s average tariffs inthe region declined from over 30 percent to12 percent. The region’s share of the worldmarkets increased and net inflows of foreigndirect investment (FDI) as a percent of GDPsteadily climbed, reaching 5 percent of GDP in1999—higher than East Asia. Overall, FDI netinflows in the latter half of the decade morethan doubled from an average of 1.4 percentof GDP in the first half to an average of 3.6percent in the second half.

In the Middle East, policy and economicbarriers, together with a reliance on oil forseveral countries, prevented rapid growth intrade. High tariff rates, restrictions on servicesentry, and controls on agriculture interactedwith poor investment climates to impede tradeand keep transactions costs high. A largestate-led sector also shaped a noncompetitiveindustrial policy that discouraged trade. Aver-age tariff rates were almost 30 percent in thelate 1990s, mirroring the import substitutionpolicies early in Latin America and morerecently in India. Flows of foreign direct in-vestment as a percent of GDP have recovered

in the last decade, but still remain quite low atless than one percent.

South Asian countries other than Sri Lankaneither liberalized trade rules nor the rules gov-erning inflows of foreign direct investmentuntil the 1990s. Removal of the most egregiousforms of anti-export bias and gradual domesticreforms, together with textile preferences,produced a rapid expansion in garment/textileexports, and led to high growth rates for ex-ports in the 1990–2000 period and an increas-ing share of exports in GDP. Since growth wasfrom a low base, South Asian exports as ashare of world trade have remained lowthroughout the 1980–2000 period. South Asiamaintained the highest levels of average ap-plied tariffs, even compared to the import-sub-stitution industrialization period of other re-gions. However, this is changing. Nepallaunched trade liberalization in the early1990s. Sri Lanka and then Pakistan in 1997began to reduce their border barriers and in-crease their trade with the world economy.India began to reduce border protection fromvery high levels in the early 1990s and has con-tinued doing so; in early 2004, India an-nounced tariff cuts of roughly one-third,reducing the average tariff rate to about 22 per-cent. Bangladeshi border protections are stillamong the highest in the world, but they tooannounced reductions in 2004.11 The regionremains only minimally integrated in worldcapital markets. Net inflows of FDI, althoughhigher than in the early 1980s, are less than 0.8percent of GDP—the lowest of all the regions.

As with the Middle East and North Africaand South Asia regions, Sub-Saharan Africaremains weakly integrated into the globalmarket. Although exports as a share of GDPin Sub-Saharan Africa increased in 2000, ex-ports as a share of world exports have re-mained flat throughout the last decade and arelower than in the early 1980s. GDP growthhas also been worse than in the earlierdecades. Many countries in Sub-SaharanAfrica are dependent on only a handful ofcommodities with highly volatile prices; most


46

face very high transport costs and have weakinstitutions to facilitate trade. These countrieshave also experienced a number of armedconflicts throughout the previous decades andare plagued by endemic diseases such asmalaria and HIV/AIDS, which have majorimpacts on their economies and societies. Allthese factors hobble trade performance.

Changing Export Compositionand the Rise of Global Production Networks

The differential in trade and growth per-formance reflects the fact that certain re-

gions have been better placed—in partthrough the policies they adopted—to takeadvantage of new technologies and changesin the nature of world trade. Not only has thevolume of international trade expanded in thepostwar period, but also its structure haschanged in three fundamental ways. First, ex-ports of manufactured products from devel-oping countries, and trade in manufacturesamong them, have become increasingly im-portant for all regions. Second, trade integra-tion has allowed developing countries to spe-cialize (most evident in the emergence ofproduction chains), with trade in intermedi-ates becoming more important. This trend isalso evident in the role that new productsplay in production. Finally, foreign direct in-vestment is playing an ever-increasing role inthe integration process. These developmentshave facilitated the integration of countriesthat have adopted relatively open trade poli-cies, and have increased the disadvantagesfacing countries that have segmented them-selves from global markets.

Specialization in manufacturesManufactured products as a share of exportsincreased strongly between 1981 and 2001 forall regions (figure 2.5). Countries in East Asiaand later, Latin America and Eastern and Cen-tral Europe, have followed open development

strategies that have led to increasing exports,especially of manufactures. The share of man-ufacturing in exports from East Asia, for ex-ample, increased from about 52 percent in1981 to 88 percent in 2001, while the share inLatin America tripled from about 20 percentto 60 percent.

Trade has allowed manufacturers to ex-ploit economies of scale, specialization, andscope. This is reflected in the growing shareof parts and components in total exports. Inthe three more open regions—East Asia andPacific, Europe and Central Asia, and LatinAmerica and the Caribbean—parts andcomponents trade has surged. This interna-tional segmentation of production—“pro-duction chains” in which intermediate in-puts are traded and transformed into moreprocessed intermediate inputs, which arethen moved across borders to the next stagein production—has been a major factor dri-ving the surges in intra-regional trade inthose areas.

One indicator of specialization is the im-port content of exports. To measure the roleof imported intermediates in trade, we calcu-lated an index of vertical specialization,which measures the share of the value addedof an export accounted for by imported inter-mediate inputs, either directly as imported in-puts in the exporting sector or indirectlythrough the use of imported inputs in the do-mestic production of intermediate goods usedby the exporting sector.12 Vertical specializa-tion is most important in East Asia, and leastimportant in South Asia and Sub-SaharanAfrica (figure 2.5c).

The evolution of production chains andfiner division of the production processesacross countries, including developing coun-tries, allows producers to exploit potentialefficiency gains from: (1) local increasingreturns to scale in the production of interme-diate inputs, (2) regional differences in factorcosts for different parts of the productionprocess, (3) increased competition from awider market, and (4) technology transfer


47

0

Manufactures: Share of exports by region, %

a. Manufactures have become increasingly important

10 20 30 40 50 60 70 90 100

Figure 2.5 Composition of trade has changed

80

Sub-Saharan Africa


South Asia

Latin America and Caribbean

Europe and Central Africa

East Asia and Pacific1981

2001

East A

sia

and

Pacific

Exports of parts and componentsas a share of total exports (percent)

Sources: Data from GTAP (ver. 6.03) and World Bank for 2001.

b. Parts and components have increased for all regions

0

4

2

6

8

10

12

14

16

Europ

e an

d

Centra

l Asia

Latin

Am

erica

and

Caribb

ean

Midd

le Eas

t

and

North

Afri

ca

South

Asia

Sub-S

ahar

an

Africa

1980 2002

Indirect

Direct

Share of 1980’s bottom 10% ofmanufactured exports in total intra-regional exports

d. New products propelled export diversification

10

5

01981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001

Europe and Central Asia


Latin America and Caribbean

Sub-Saharan Africa

South Asia20

15

25

30

35

40

45

50

East A

sia

and

Pacific

Import content of exports (percent)

c. Vertical integration with the global economy hasincreased

0

105

1520253035404550

Europ

e an

d

Centra

l Asia

Latin

Am

erica

and

Caribb

ean

Midd

le Eas

t

and

North

Afri

ca

South

Asia

Sub-S

ahar

an

Africa

East Asia and Pacific

Source: UN COMTRADE.



from developed countries embodied in im-ported intermediate inputs and backwardlinkages through exports. The magnitude ofthese links between increased trade in inter-mediates and productivity growth in develop-ing countries has been studied in both cross-country analysis and country case studies.13

While causation is difficult to establish, theevidence indicates that such links are impor-tant, and productivity growth associated withincreased trade in intermediates is a poten-tially important source of growth.

Trade in new productsA large part of the expansion in exports incountries undergoing liberalization and suc-cessful trade expansion comes from productsthat were not traded—or minimally traded—prior to liberalization (see Kehoe and Ruhl2002).14 Growth in trade in new products

may have the important advantage of allow-ing countries to escape the deterioration in theterms of trade that would come from trying toincrease market share in existing products.15

To assess this phenomenon, we reviewed thetrade performance of the least traded decile ofproduct categories. In East Asia and Pacific,those products that figured in the lowest 10percent of all EAP manufactured exports to theworld in 1981 grew to almost 40 percent by2001 (figure 2.5). For the other five regions,the performance of products among their low-est initial 10 percent was also noteworthy.Countries are building dynamic new marketsfor their existing exports and developing newvariations of old products to replenish theproduct cycle. This trend is also associatedwith increased trade in intermediates; detailedanalysis indicates that many of the new exportgoods are intermediate inputs. Increased trade


48


East A

sia

and

Pacific

Share of region’s net FDI inflowsin the region’s total investment (percent)

Note: Period averages for 1980–1989, 1990–1999, and2000–2002 calculated from annual data. Both net FDI (BoP)and gross fixed capital formation are in current US dollars. For1980–1989 in ECA, GFCF was estimated using a three-yearaverage share of GFCF in GDP over the 1990–1992 time periodand back-calculating GFCF from GDP for the entire 1980–1989period. GDP time series (for 1980–1989 ECA only) obtainedfrom UN national accounts.

e. FDI flowed mostly into East Asia and Latin America

0

4

2

6

8

10

12

16

14

18

Europ

e an

d

Centra

l Asia

Latin

Am

erica

and

Caribb

ean

Midd

le Eas

t

and

North

Afri

ca

South

Asia

Sub-S

ahar

an

Africa

Parts and components, FDI, GDP: 2000FDI, percent of GDP

f. A positive trend between FDI and parts exports for theregions

0

0.5

0 2 4 6 8 10

Parts and components, percent of exports

12 14 16

1.0

1.5

2.0

2.5

3.0

3.5

4.0

SSA

MNASAS

ECA

Y � 0.2333x � 0.3067R2 � 0.6801

EAP

LAC4.5

Sources: World Bank, WDI, and UN COMTRADE.Sources: World Bank, WDI.


49

in new products is thus part of the virtuous cir-cle linking trade and growth.

Investment, handmaiden of tradeForeign investment has been a driver ofintegration, increased trade in manufactures,and vertical specialization. As tariff barriershave come down in manufactures, market-seeking, horizontal FDI that once led the way inthe import-substitution process has faded in im-portance relative to efficiency-seeking, verticalFDI that looks to locate segments of productionin the lowest-cost site. This form of investmentis associated with the rise in production chainsand trade in components and parts.

FDI has increased as a share of GDP in all re-gions. This trend abated somewhat since theEast Asia crisis in 1997–98 and the global re-cession of 2001–02, but FDI growth is likely toresume with the recovery of the global econ-omy in 2004. East Asia and Latin America—the largest markets—have had and haveretained by far the largest shares of FDIthroughout the period (figure 2.5).

In East Asia and Pacific, the increase in FDIsupported the pattern of segmentation and re-location of production processes within the re-gion. In the 1990s, a large part of the FDI intoLatin America was due to the privatizationprocess the region underwent during this pe-riod. There is broad correspondence betweenFDI trends by region and the share of parts andcomponents intermediates in regional exports.

Technology transfer from developed to de-veloping countries is linked to trade, especiallythrough trade in manufactures and intermedi-ates, and also through foreign direct invest-ment.16 The better economic performance ofEast Asia and Pacific can be seen as resultingfrom the emergence of a “virtuous circle” orsynergy between increased specialization inproduction, increased trade in intermediates,increased foreign investment, increased factorproductivity, and increased growth. Thisregion started earlier and appears more suc-cessful than other regions in achieving and sus-taining this virtuous circle.

Preferential Trade and Regional Outcomes

Many historical factors, not just preferen-tial trading arrangements, contributed

to these trends. In the next chapter we providea more detailed analysis of the impact of RTAson trade. Here we simply highlight that the na-ture of RTAs and the context in which theyhave been applied have varied enormouslyacross regions and that regional agreementsoften follow, rather than determine, changes inregional trade patterns. This suggests that pref-erential trade agreements are just one of manyfactors affecting trade outcomes and that whenimplemented in a highly restrictive economicenvironment, they are usually inconsequential.

History shapes trading patternsDiffering regional performances in trade andgrowth have roots that go deeper than justboundaries on a map. Trade patterns—whotrades with whom—have grown out of longpolitical and economic histories that precededthe trends evident in the last two decades. Theclusters of trading partners often bear little re-lationship to arbitrary definitions of regions(see Anderson and Blackhurst 1993, for an ear-lier analysis). Major trading blocs—that is,those countries that trade more with each otherthan with those outside their group—emergefrom a cluster analysis. These blocs are not de-fined as traditional geographic political re-gions, but rather by statistical patterns in tradeflows over decades.

The bipolar world of the 1960sComing out of the postwar period, the struc-ture of world trade by the 1960s reflected abipolar world, in which Europe and theUnited States had effectively formed blocswith some of their close neighbors, formercolonies, and/or cold war partners; and withhub-and-spoke links to most of the developingcountries. The two leading world trade blocseffectively accounted for 80 percent of globaltrade (figure 2.6). The European bloc was

50

In the 1960s, the European Union and United States dominate trade . . .

. . . but by the 1970s, Japan and Korea begin to lead an East Asian bloc . . .

. . . a decade later, the East Asian Tigers, ASEAN countries, and Australia consolidate the East Asia bloc . . .

. . . and in the 1990s, ECA emerges and East Asia trades more with itself than with the U.S. and EU.

Figure 2.6 Evolving trading blocs

Europe �

Percent

U.S. � Asia-U.S.Asia-UK

52.7

29.9

7.110.3

44.9

19.7

2.7

24.1

48.3

18.8

2.8

23

45.8

19.9

1.6

27.2

0

30

20

10

40

50

60

Europe �

Percent

NorthAmerica �

SouthAmerica

East andSouth

East Asia

0

20

10

30

40

50

Europe �

Percent

NorthAmerica �

MERCOSUREast andSouth

East Asia

0

20

10

30

40

50

Europe �

Percent

NorthAmerica �

SouthAmerica

East andSouth

East Asia

0

20

10

30

40

60

50

Source: Robinson and Diaz Bonilla 2004.

Asia-U.S.

Rest

Rest

1960s

1970s

1980s

1990s

Europe �U.S. �Asia-UK

Europe �

North America �East Asia

South America

Rest

Europe �North America �East Asia

South America

Europe �North America �

MERCOSURSouth Africa �

Andean

East Asia


51

closely linked with countries in Africa and for-mer colonies. The United States was closelylinked to Latin America.17 Britain still re-tained leadership in a small Asian cluster con-sisting of former colonies, China, and the restof the Middle East, much as the United Statesled a cluster of countries closely linked to thepost-WWII political order in the Pacific—Indonesia, Japan, Korea, and Taiwan. The de-pendent developing countries traded muchmore with Europe or the United States, andnot as much among themselves.

The realignment and the emergence of East Asia in the 1970s and 1980sIn the 1970s and the 1980s, a realignment ofworld trade began. The European and U.S.blocs fragmented in the 1970s, and their dom-inance dissipated to 65 percent of global trade.The East and Southeast Asian countries left theEuropean bloc and effectively formed a newbloc (East Asia).18 This bloc consolidated inthe 1980s, increasing the share of within-bloctrade and expanded membership to includeAustralia and New Zealand. Its export shareshifted toward the United States (36.2 percentin the 1980s compared to 26.4 percent in the1970s). It also represented a growing share oftotal world trade—23 percent in the 1980scompared to 16 percent in the 1970s.

Consolidation and diversification in the 1990sBy the 1990s, earlier trends blossomed into atri-polar world. One new element was thebreakup of the Soviet Union and Moscow’scentral management of trade in EasternEurope. Trading patterns began gravitatingtoward the EU. A second new element wasthe emergence of a new grouping: Argentina,Brazil, Paraguay, and Uruguay (i.e., MERCO-SUR).19 Finally, detailed analysis indicates anascent bloc forming around South Africa,including the SACU countries.

These global trading patterns are revealing:First, today’s tri-polar world of global com-merce does not signify that the world is evolv-ing into three disparate, autarchic trading

blocs. To the contrary, trade among these blocsis intensifying, becoming more diversified, andlinked with a web of business ties acrossoceans that bind the world market together.Second, in the 1990s, the emergence of newpoles of commerce—MERCOSUR and SouthAfrica—indicates that the process of segmenta-tion and new bloc formation in world trade isstill evolving. Third, it is clear that many of theemerging preferential arrangements have deeproots in historical trading patterns, but thatsome of the more recent bilateral FTAs aregoing beyond these historical patterns.

RTAs in different regions have different impactsPatterns of regional integration tend to confirmthe view that usually trade has preceded tradeagreements. For East Asia, integration with theglobal economy was a strong impetus for re-gional integration. Exports to the world pro-duced demand for imports from neighboringcountries. As Korea matured and Chinaopened, internal regional growth assumed itsown dynamic. As a share of GDP, intra-regionaltrade in 2002 was 26.5 percent, twice as high asin the next highest region, yet regional tradepreferences were very modest at best.

In fact, regional preferential tradingarrangements followed rather than precededthis regional integration. The ASEAN FreeTrade Area (AFTA), established in 1992, ac-celerated these tendencies and contributed toSoutheast Asia’s integration, but much of thetariff reduction has accompanied rather thanpreceded these patterns of regional integra-tion. However, ASEAN leaders accepted inthe Bali Declaration the need to pursue deeperintegration and to create a single market toenhance the competitiveness of the region.The importance of preferential trade in the re-gion was dramatically increased by the sign-ing of a FTA between ASEAN countries andChina.

Regional arrangements were critical for thesuccessful integration of Eastern Europe intothe world economy, but have not been assuccessful in the CIS countries. With the

assistance of the EU in its Europe Agreementsand with the aspiration of WTO membership,the Eastern European countries have movedswiftly toward integration. The CIS has beenburdened with incomplete reforms, a poor in-vestment climate, and a plethora of tradingarrangements that have been implementedonly partially. The combination has weigheddown the subregion’s performance.

In Latin America, the intra-regional shareof exports in GDP in 2002 remains only one-fourth of East Asia’s share. Since the 1960s,with the formation of the Latin America FreeTrade Area (LAFTA), the region has struggledto expand trade. This proved futile as long asimport substitution policies were in place andstate enterprises were used as instruments ofindustrial policy. Early attempts in CentralAmerica and the Andean countries failed be-cause of the inherent difficulty in managingpotential trade diversion and location of in-dustries within the regions behind high exter-nal protection barriers.20

This situation changed with the wave ofunilateral reforms in the 1985–95 period.Mexico’s reforms paved the way for the latercreation of NAFTA in 1994, and reforms inBrazil and Argentina led to the creation ofMERCOSUR in the same year. Similarly, theCentral American countries, with a second go,managed to put in place a successful commonmarket in the 1990s. As a result, intraregionalregional trade has proceeded pari pasu withgrowth in the external markets.

In the Middle East, intra-regional trade hasfailed to gain dynamism. Because countriesbegin with broadly similar production andexport structures, the scope for using regionaltrade to establish patterns of specializationand diversification in manufacturing produc-tion is limited. Intra-regional trade cannot bea substitute for extra-regional trade.

Several countries signed bilateral tradeagreements with the EU as part of the Euro-Mediterranean agreements. Jordan andMorocco also signed agreements with theUnited States. All countries in the region

entered into the Pan-Arab Free Trade Agree-ment and most participated in the sub-re-gional customs union, the Gulf CooperationCouncil. Even so, these agreements have notbeen sufficient to overcome the effects of highborder barriers and restrictions on services.The Euro-Med agreements with the EU havefallen short in their aspirations because of re-strictions on trade in agriculture, services, andlabor; lack of harmonization of standards;and stringent rules of origin.21

Regional agreements in South Asia, as withLatin America in the 1960s and 1970s, floun-dered on the shoals of high protection. The1993 South Asia Preferential Trade Agreement(SAPTA) was stillborn, given continuing highlevels of protection, a lack of meaningful con-cessions, domestic political problems, hostilitybetween India and Pakistan, India’s ban onimports of all consumer goods (from SAPTAcountries until 1998 and from the rest of theworld until 2001), and India’s control overmajor primary goods (Baysan 2004).

Recently, however, unilateral trade reformsin India and Pakistan, political rapprochement,as well as concerns about rising preferentialtariff arrangements in other parts of the world,led to the formation of the SAFTA Agreementin January 2004 (Newfarmer 2004).

In Sub-Saharan Africa, regional tradeagreements are common and reflect an aspi-ration to overcome the limits that small sov-ereign states impose. These include SACU—one of the oldest customs unions in theworld—CEMAC, COMESA, ECOWAS, andthe East African Community. Although aver-age applied MFN tariffs were cut by half be-tween the 1990s and 2003, non-border barri-ers restrict internal trade. The recent regionaltrade agreements have had more impact onoutward-looking MFN trade liberalization,and thus on external trade, than on intra-regional trade. The economic impact of theseagreements appears to have been small, espe-cially compared to pre-independence arrange-ments that essentially validated existing eco-nomic links (SACU, the West African


52


53

Economic and Monetary Union, andCEMAC). The EPAs under negotiation withthe EU may reinforce this outward-lookingpattern of trade integration, but the hope isthat they will also aid Africa’s own regionalintegration if they succeed in fostering eco-nomic reform and performance.22

The situation is different for agricultureand food products. Here margins of prefer-ence are more substantial in all regions exceptSouth Asia. The average margin of preferencein the high-income region is similar to that inEast Asia, Eastern Europe, and Latin America.Again, preferences are greatest in those re-gions showing the lowest degree of regionalintegration—the Middle East, and Sub-Saharan Africa.

Trade preferences have had very little im-pact on the high levels of intra-regional tradein manufactures in East Asia and EasternEurope. Regions that offer substantial tradepreferences behind high external barriers havenot fared well in stimulating the growth ofintra-regional trade.

Conclusion

Developing countries have increased theirshare of global trade as multilateral trade

negotiations have led to sustained reductionsin border protection for manufactured prod-ucts. At the same time, and for a variety of rea-sons, the preferential trade agreements haveproliferated. While the number of preferentialagreements has increased rapidly, their tradecoverage is substantially less than their officialspan of influence. Because many tariffs havecome down close to zero, rules of origin re-strict preferential access, and many productswithin agreements are excluded. Nonetheless,RTAs are leading to a more complex tradingsystem and inefficiencies in customs adminis-tration; high tariffs in certain regions still risksignificant trade diversion.

Notable differences are emerging betweenNorth-South bilateral agreements and South-South arrangements. North-South agreements

are considerably more ambitious in contentand coverage than South-South arrangementsand reach deep behind the border to includeservices, protection of investment rules, andintellectual property rights.

In general, the wave of preferential tradingarrangements followed—rather than preceded—an intensification of regional trade. Regionswith the lowest external (MFN) border barri-ers ironically have developed the deepestintra-regional links and have been best posi-tioned to diversify and exploit the emergenceof global production chains in the manufac-turing sector. East Asia is the starkest exam-ple, but Eastern Europe, in the wake of thedissolution of the Soviet Union, and LatinAmerica, with the end of import-substitutionindustrialization, are not far behind.

What are the economic consequences ofthese arrangements? That is the subject of thenext three chapters.

Notes1. The recent accession of 10 new members to the

European Union reduced the total number of RTAs inforce from 285 to 229.

2. In fact there are only 12 countries that are notrecorded as being party to a RTA, and many of theseare small islands and principalities. The 12 are Ameri-can Samoa, Bermuda, Channel Islands, Guam, Isle ofMan, Monaco, Mongolia, N. Mariana Islands, Palau,Puerto Rico, Timor-Leste, and the Virgin Islands.

3. We are indebted to Gaspar Fontini of the Euro-pean Commission’s Directorate General for Trade forthis formulation.

4. The U.S. Trade Representative is required to of-ficially notify the U.S. Congress of its intent to negoti-ate FTAs.

5. See Weintraub (2004) and Schott (2004) for amore detailed discussion.

6. In the course of this discussion, the EU is treatedas a single entity. For example, an EU-Mexico agree-ment is classified as bilateral.

7. The provision for regional cumulation in therules of origin, particularly full cumulation, will tendto offset the hub-and-spoke system. See Brenton andImagawa (2004). The EU, following substantial criti-cism of the hub-and-spoke system that emerged inEurope in the late 1990s, moved to create pan-European cumulation, although in terms of the morelimited partial cumulation.

8. See Grether and Olarreaga (1999), who includeGSP preferences but a smaller sample of countries.

9. In other words, 88 percent of intra-regionaltrade takes place among RTA partners.

10. This section benefited from the World Bank’sregional chief economists at a workshop titled Region-alism, Trade and Development, May 5, 2004. SadiqAhmed, Harry Broadman, Alan Gelb, Homi Kharas,Mustafa Nabli, and Guillermo Perry made presenta-tions that became the basis of this discussion.

11. The announced reforms were to reduce “sup-plementary duties” that were, in legal terms, additionalto ad valorem tariffs; the economic effect is to reduceeffective tariffs.

12. Hummels, Ishii, and Yi (2001) define an indexof vertical specialization (VS). For direct effects forcountry k:

VSk � �uA

X

M

k

X�

where u is a 1 x n vector of 1’s, Am is the n x n im-ported coefficient matrix, X is an n x 1 vector of exports,n is the number of sectors, and Xk is the sum of exportsacross the n sectors. When indirect effects are added,

VSk��uAM[I�

XA

k

D]�1X�

the index becomes: , where I is the identity matrix andAD is the n x n domestic coefficient matrix. Typically,given data limitations, measures of vertical specializa-tion are imperfect. For example, a sector which pro-duces both for exports and domestic markets, which iscommon given the aggregation available for sectoraldata, is assumed to have the same production technol-ogy, particularly use of imported intermediates, forgoods sold in either market. Even with these limita-tions, the measures provide a good picture of thechanging role of trade in intermediates.

13. Winters (2004, 10–11) reviews the literatureon links between trade in intermediates and productiv-ity growth.

14. See Kehoe and Ruhl (2002). They present lim-ited empirical evidence and then suggest a theoreticalmodel that captures the phenomenon. Increased tradein new products is difficult to capture in standard mod-els of world trade, so such models will tend to over-state terms-of-trade effects from changes in trade.

15. See Hummels and Klenow (2002). 16. Winters (2004) surveys work on the links be-

tween trade, productivity, and growth in developingcountries. See also Nishimizu and Robinson (1984),and Esfahani (1991) who consider the links in semi-industrial countries. Schiff and Wang (2004) considerthe link between TFP growth and regional tradelinks. Keller (2002) considers the mechanismsinvolved.

17. The definition of “closely linked” is countriesthat have such large trade shares with their bloc part-ners, particularly the United States and Europe, thatadding them to the bloc increases the average within-bloc trade share.

18. Plus one region, “rest of Middle East and NorthAfrica” (MENA), which is closely linked to Europe, buthas significant trade with East and Southeast Asia.

19. The within-bloc trade share for MERCOSURof 23 percent is lower than its trade share with Eu-rope (27 percent), but given the increasing trend of itswithin-cluster trade share and the legal formation ofMERCOSUR in 1995, it makes sense to designate itas a bloc. Latin American countries fall into threegroups. One group is linked closely with North Amer-ica, including RB de Venezuela, Colombia, and rest ofCentral America. The MERCOSUR countries define asecond group, a trade bloc with diversified exports(a slightly higher share to the EU than to the UnitedStates). Finally, the third group is a heterogeneous col-lection of countries (Peru, Chile, and Other AndeanPact) with exports largely to the EU, partly to Eastand Southeast Asia, and with little trade amongthemselves.

20. IDB, Beyond Borders, 2002. 21. Chief Economist, MENA, 2004. 22. Chief Economist, Africa, 2004.

ReferencesAcemoglu, D., S. Johnson, and J. A. Robinson. 2001.

The Colonial Origins of Comparative Develop-ment: An Empirical Investigation. AmericanEconomic Review 91, no. 5: 1369–1401.

Anderson, Kym, and Richard Blackhurst, eds. 1993.Regional integration and the global trading sys-tem. New York: St. Martin’s Press.

Baldwin, R. 2003. Openness and Growth: What’s theEmpirical Relationship? National Bureau of Eco-nomic Research, Working Paper no. w9578.

Baldwin, R., and A. Venables. 1995. Regional Eco-nomic Integration. In Handbook of InternationalEconomics, vol. 3, eds. Grossman and Rogoff,Amsterdam: North Holland Press.

Baysan, T. 2004. South Asia: Lessons and the WayForward. Mimeo, World Bank, Washington, DC.

Bhagwati, J., and A. Panagariya, eds. 1996. TheEconomics of Preferential Trade Agreements.Washington, DC: American Enterprise InstitutePress.

Brenton, Paul, and Hiroshi Imagawa. 2004. Rules ofOrigin, Trade, and Customs. In Customs Mod-ernization Handbook, Conference edition, eds.Luc De Wulf and Jose Sokol. Washington, DC:World Bank.


54

Burfisher, Mary, Sherman Robinson, and KarenThierfelder. 2004. Forthcoming. Regionalism:Old and New, Theory and Practice. In Agricul-tural Policy Reform and the WTO: Where AreWe Heading? Eds. G. Anania, M. E. Bohman,C. A. Carter, and A. F. McCalla. Cheltenham,UK & Northampton MA, US: Edward Elgar.

Destler, I. M., and Peter J. Balint. 1999. The New Poli-tics of American Trade: Trade, Labor, and theEnvironment. Washington, DC: Institute for In-ternational Economics.

Esfahani, H. S. 1991. Exports, Imports, and EconomicGrowth in Semi-Industrialized Countries. Journalof Development Economics 35.

Falkenberg, Karl F. 2004. EPAs and Doha Develop-ment Agenda—Parallelism or CrossroadsTrade Negotiations Insights 3, (July) (4).www.ictsd.org.

Frankel, J. A., E. Stein, and S-J. Wei. 1996. RegionalTrading Arrangements: Natural or Supernatural?American Economic Review 86, (2): 52–56.

Grether, Jean-Marie, and Marcelo Olarreaga. 1999.Preferential and Nonpreferential Trade Flows inWorld Trade. In Trade rules in the making:Challenges in regional and multilateral negotia-tions, 159–79. Washington, DC: BrookingsInstitution Press and Organization of AmericanStates.

Hinkle L., and M. Schiff. 2004. Economic PartnershipAgreements between Subsaharan Africa andthe EU: A Development Perspective on the TradeComponents, Africa Region Working Paper(draft). February, World Bank, Washington, DC.

Hummels, David, Jun Ishi, and Kei-Mu Yi. 2001.The nature and growth of vertical specializationin world trade. Journal of InternationalEconomics (Netherlands) 54 (June) (1): 75–96.

Hummels, David, and Peter J. Klenow. 2002. Thevariety and quality of a nation’s trade. NationalBureau of Economic Research, Working PaperSeries 8712, January.

IDB. (Inter-American Development Bank). 2002. Be-yond Borders: The New Regionalism in LatinAmerica. Economic and Social Progress in LatinAmerica. Inter-American Development Bank,Washington, DC.

Kehoe, Timothy J., and Kim J. Ruhl. 2002. HowImportant is the New Goods Margin in Interna-tional Trade? University of Minnesota.

Keller, Wolfgang. 2002. Trade and the Transmission ofTechnology. Journal of Economic Growth 7, (1):5–24.

Kiener, Christophe. 2004. Economic PartnershipAgreements: A New Approach to ACP-EU Eco-nomic and Trade Cooperation. Paper presented to

the EU-World Bank Retreat on Preferential TradeAgreements, July.

Mansfield, Edward D., and Jon C. Pevehouse. 2000.Trade Blocs, Trade Flows, and InternationalConflict. International Organization 54 (Au-tumn) (4): 775–808.

Martin, W., and F. Ng. 2004. Sources of Tariff Reduc-tions. Background paper.

Melo, J. de, and A. Panagariya, eds. 1993. NewDimensions in Regional Integration. Cambridge:Cambridge University Press.

Newfarmer, R. 2004. SAFTA: Promise and Pitfalls ofPreferential Trade Arrangements. Backgroundpaper for Global Economic Prospects 2005, May,World Bank, Washington, DC.

Nishimizu, Mieko, and Sherman Robinson. 1984.Trade Policies and Productivity Change in Semi-Industrialized Countries. Journal of DevelopmentEconomics 16, (1–2): 177–206.

OECD. 2003. Regionalism and the MultilateralSystem. OECD, Paris.

Robinson S., and C. Diaz-Bonilla. 2004. The evolutionof global trading blocs in the past forty years:Who trades what with whom, and why. Back-ground paper.

Robinson, S., and K. Thierfelder. 2002. Trade liberal-ization and regional integration: the search forlarge numbers. Australian Journal of Agriculturaland Resource Economics. 46, (4) 585–604.

Sapir, Andre. 1998. The Political Economy of EC Re-gionalism. European Economic Review. 42,(3–5): 717–32.

Schiff, M., and L. A. Winters. 2003. Regional Integra-tion and Development. Washington, DC: WorldBank.

Schiff, M., and Y. Wang. 2004. North-South Technol-ogy Diffusion, Regional Integration, and theDynamics of the ‘Natural Trading Partners’Hypothesis. World Bank, June, Washington, DC.

Schott, Jeffrey J. 2004. Free Trade Agreements: Boonor Bane of the World Trading System. In FreeTrade Agreements: US Strategies and Priorities,ed. J.J. Schott. Washington: Institute for Interna-tional Economics; Song and Xu. 2000.

Ullrich, Heidi. 2004. Comparing EU Free TradeAgreements–Services. European Centre for Devel-opment Policy Management. July.

USTR. 2004. Free Trade with Chile, Trade Facts.www.USTR.gov.

Weintraub, Sidney. 2004. Lessons from the Chile andSingapore Free Trade Agreements. In Free TradeAgreements: US Strategies and Priorities, ed. J. J.Schott. Washington: Institute for InternationalEconomics.


55

Winters, L. Alan. 2004. Trade Liberalization andEconomic Performance: An Overview. The Eco-nomic Journal, 114 (February): F4–F21.

Winters, L. Alan, Neil McCulloch, and AndrewMcKay. 2004. Trade Liberalization and Poverty:The Evidence So Far. Journal of EconomicLiterature, XLII (March): 72–115.

Wonnacott, Paul. 1996. Beyond NAFTA—The Designof a Free Trade Agreement of the Americas. InThe economics of preferential trade agreements,79–107. College Park: University of Maryland,Center for International Economics; Washington,DC: AEI Press.

World Bank. 1999. Trade, Global Policy, and theEnvironment. World Bank Discussion Paper No.402. Washington, DC: World Bank.

World Bank. 2003. Trade, Investment, and Develop-ment in the Middle East and North Africa: En-gaging the World. Washington DC: World Bank.

. 2004. East Asia Integrates: A Trade PolicyAgenda for Shared Growth. East Asia and PacificWorld Bank Region, World Bank, Washington,DC.

WTO (World Trade Organization). 1999. Trade andEnvironment. Geneva: WTO.


56

Regional trade agreements (RTAs) can havepositive or negative effects on trade dependingon their design and implementation. Analysisin this chapter confirms that gains from a pref-erential trade agreement cannot be taken forgranted; moreover, even in agreements withpositive impacts on average incomes, not allmembers are assured of increases. The inter-esting policy question then is not whetherRTAs are categorically good or bad, but whatdetermines their success?

The broader policy context in which anRTA is designed and implemented is crucial.Agreements that have been designed to com-plement a general program of economic re-form have been most effective in raising trade.When RTAs have tended to be fruitless, it isoften because of the lack of a coherent pro-gram of reform.

For an RTA itself, the most importantingredient for success is low trade barrierswith all global partners. Most-favored-nation(MFN; i.e., nondiscriminatory) liberalization,which creates more trade, is the fastest andmost efficient way to increase intraregionaltrade. In addition, agreements that minimizeexcluded products expand the scope for posi-tive net benefits through competition andtrade creation.

Recent research has added nonrestrictiverules of origin to the list of successful factors;local firms must be able to effectively sourcematerials at the lowest cost. Such rules of ori-gin are an essential element of agreements that

expand both regional exports and exports tothe rest of the world.1

RTAs can be a springboard to global mar-kets, but here too, low MFN trade barriers arenecessary for success. RTAs can help countriesintegrate with global markets, but no agree-ment provides guarantees, so design and im-plementation matter.

The Impact of RTAs onMerchandise Trade and IncomesRTAs cover much more than trade barriers RTAs have increasingly been designed to covermuch more than formal trade policies (seechapter 2), and RTAs are signed for a varietyof reasons. The impact of these agreements ontrade determines the extent to which broaderpolitical and social objectives are achieved. Itis difficult to identify an agreement that hasfostered wider political objectives withoutachieving economic integration. It is clear thatthe political context and broad economic en-vironment in which integration takes place arecrucial for determining the trade impact. Suc-cess derives from a strong willingness to liber-alize and to accept the subsequent economicadjustments, accompanied by intense mutualeconomic dialogue and communication andgenuine efforts toward mutual understanding.Severe macroeconomic disturbances and a tur-bulent investment climate can easily disrupttrade and derail an agreement.

57

Regional Trade Agreements:Effects on Trade

3

The simplest measure of integration is thetrend in the share of imports from regional part-ners in the total imports of a region. Successfulregional agreements might be expected to in-crease trade between partners relative to thosecountries’ trade with the rest of the world. Butthree important caveats need to be understood.

First, successful regional integration is typ-ically accompanied by reductions in tariffs forall partners. Hence, regional trade shares maynot rise even though the volume of regionaltrade is increasing. Second, regional tradeagreements that provide for the removal orreduction in trade costs other than those asso-ciated with formal trade policies (such as im-proved customs procedures), may stimulatetrade from all sources. Third, many agree-ments cover nontrade issues such as invest-ment, services, and labor, and these can haveimportant consequences for growth andincomes. These are analyzed in subsequentchapters, but it is important to bear in mindhere that an agreement may be successful evenif the propensity for members to trade amongthemselves does not increase markedly.

Trade performance in several regional tradeagreements shows that the increase in intra-regional trade shares of agreements signed inthe 1990s has been substantial (figure 3.1).The share of intra-NAFTA (North AmericaFree Trade Agreement) trade rose from lessthan 35 percent in the late 1980s to almost

50 percent in 1999. Over the same period, theimportance of trade between MERCOSURmembers doubled from 10 to 20 percent.

For many of the agreements signed in the1990s, intra-regional trade shares weregrowing strongly before the agreements weresigned (NAFTA, MERCOSUR, SAPTA,SADC). There may have been some anticipa-tion effect in the year or two before signing, butthis doesn’t explain trend increases in sharescommencing five or more years previous, as inthe case of MERCOSUR. In many cases this in-crease in regional trade reflects the impact ofunilateral, multilateral, as well as regional tradeliberalization and the fact that agreementsoften follow growing trade relationships.

In Africa, the picture is mixed. The extentof regional integration among the CommonMarket for Eastern and Southern Africa(COMESA) members has been relatively staticover the past two decades. In contrast theshare of intra-area trade has increased sub-stantially for Economic Community of WestAfrican States (ECOWAS) since the early1980s and for SADC since the late 1980s. InEast Asia, a region that has experienced sub-stantial economic progress over the past 20years, there has been little increase in intra-regional trade shares.

Given these disparate results, it is necessaryto go beyond simple trade shares to identifythe economic impact of regional trade


58

Figure 3.1 Evolution of the share of intra-regional imports in total imports, 1960–2000

Percent Percent

0

10

20

30

40

60

50

70

ASEAN

MERCOSUR

CACM

NAFTA

EC

1960

1964

1968

1972

1976

1980

1984

1988

1992

1996

2000

1960

1964

1968

1972

1976

1980

1984

1988

1992

1996

2000

0

2

4

6

8

12SADC

ECOWAS

COMESA

10

14

SAPTA


R E G I O N A L T R A D E A G R E E M E N T S: E F F E C T S O N T R A D E

59

Revised

1991

Year ofentryintoforce

1994

Year of entryinto force

1992

ASEAN intra

Year of entryinto force

1975


1994


1991

ASEAN extra

0

2

4

6

10

14

12

8

16

CACM intra

CACM extra

0

5

10

15

20

30

35

25

40

ECOWAS intra

ECOWAS extra

0

5

10

15

20

25

30

35

COMESA intra

COMESA extra

Figure 3.2 The ratio of external and intra-regional trade to GDP

Percent

ASEAN

05

101520

353025

50

4045

1960

1964

1968

1972

1976

1980

1984

1988

1992

1996

2000

Percent

CACM

1960

1964

1968

1972

1976

1980

1984

1988

1992

1996

2000

Percent

COMESA

1960

1964

1968

1972

1976

1980

1984

1988

1992

1996

2000

Percent

ECOWAS

1960

1964

1968

1972

1976

1980

1984

1988

1992

1996

2000

0

2

4

6

10

14

12

8

16

Percent

MERCOSUR

1960

1964

1968

1972

1976

1980

1984

1988

1992

1996

2000

Percent

NAFTA

1960

1964

1968

1972

1976

1980

1984

1988

1992

1996

2000

NAFTA intra

NAFTA extra

0

1

2

3

4

5

6

7

MERCOSUR intra

MERCOSUR extra

Original


1961

agreements. Because a decline in the share ofextra-regional trade in total trade will be ofless significance if the total value of trade isincreasing, a logical (and commonly used)measure is the share of extra- and intra-regional trade in regional GDP (figure 3.2).

With the exception of MERCOSUR, allregions that have experienced an increasingshare of intra-regional trade in total tradehave also seen the ratio of extra-regionaltrade in GDP increase. The Association ofSoutheast Asian Nations (ASEAN) is an


60

interesting example. The share of intra-regional trade remained fairly flat during the1990s. However, the ratios of intra-ASEANtrade to GDP and ASEAN imports from therest of the world to GDP have both increasedstrongly. ASEAN appears to have been verysuccessful.

In general, this suggests that external open-ness and the expansion of intra-regional tradego together. To take this analysis a little further,we plot the estimated relationship between an-nual changes in intra-regional trade and annualchanges in the total volume of world trade, andwe find a positive association in all cases (fig-ure 3.3). Although crude, this analysis suggeststhat the successful expansion of trade amongthe members of a regional trade agreementtends to be associated with increasing extra-regional imports as a share of GDP and withthe growth of world trade.2

Do regional trade agreementsstimulate trade?The analysis just discussed provides useful in-formation, but it does not directly measurethe impact of regional trade agreements. Toisolate the role of policy—that is, RTAs—from other factors influencing trade patternsrequires more sophisticated economicmodeling. Different, yet complimentary,

approaches are available that we can crudelyseparate into ex ante general equilibrium sim-ulation studies and ex post econometricanalyses by using the gravity model (box 3.1).

The broad results3 from general equilib-rium exercises are that, first, excludedcountries almost always lose. Second, for de-veloping countries the bottom line determi-nant of positive income effects is the increasein market access. Third, in Free Trade Areas(FTAs) each country can always lower its tar-iff to ensure gains. This may be more difficultin a customs union. Finally, regional tradeagreements are typically expected to createmore trade than they divert, although this isnot always the case.

These points are highlighted in figure 3.4,which summarizes model estimates of the im-pact of Chile signing FTAs with different re-gional groupings. Excluded countries lose inevery case. Chile loses from an FTA withMERCOSUR. FTAs with larger marketsbring bigger gains for Chile but also tend toentail larger losses for excluded countries.Large northern countries gain little fromFTAs with substantially smaller southernpartners.

A number of analysts have concluded thatthe numerous estimates from the gravitymodel generally support the contention that

Annual growth of intra-regional exports (percent)

ECOWAS

SADC

COMESA

MERCOSUR

ASEAN

NAFTA

Annual growth of intra-regional exports (percent)

Figure 3.3 Intra-regional trade grows faster when world trade growth is positive

0 5 10 15 20 25

Annual growth of world exports (percent)

5

0

�5

�10

10

15

20

25

30

35

0 5 10 15 20 25

Annual growth of world exports (percent)

10

5

0

�5

15

20

25

30

35

40

CACM



61

A. Simulation studies: Looking forward to potentialgainsThe ex ante studies are based on a specific generalequilibrium model structure that allows a rich analy-sis of the impact of RTAs at both the aggregate andsectoral levels. A key strength of this approach is itsability to highlight which sectors may expand andwhich may contract in the face of given resourceconstraints. The richness of the model structure,however, requires that many key parameters beselected, (often on the basis of an extensive literaturesearch), with others being derived by a process ofcalibration to a single base- year observation; that is,the remaining parameters are derived such that themodel replicates the situation in the base year. To alarge extent the results of the impact of RTAs aredetermined by the choice of value for key relevantparameters (in this case the price elasticity ofdemand for exports). Also, given that parameters arechosen and not estimated, the statistical properties ofthe results are unknown.

The characterization of RTAs is often simple,with most studies focusing on the removal of tariffsbut ignoring issues such as the rules of origin, prod-uct exclusions, and services. These simulation exer-cises answer the question, “What would be the im-pact of the preferential removal of tariffs against alimited set of trade partners, given the assumedmodel structure?” But they do not tell us whetherparticular agreements have actually created ordiverted trade.B. Econometric studies using the gravity model:Looking back at actual performance

The gravity model provides a useful frameworkfor assessing the impact of policy variables on thebehavior of bilateral flows between countries. Itsname is derived from its passing similarity toNewtonian physics, in that flows between twocountries increase in proportion to their economicmass (as measured by GDP) and are constrained by

Box 3.1 A primer on modeling of RTAsthe friction between them (due to trade and othercosts, which is proxied by distance). It is also com-mon to use so-called dummy variables to capture ge-ographical effects (such as whether the two countriesshare a border, or if a country has access to the sea),cultural and historical similarities (such as if twocountries share a language or were linked by pastcolonial ties), and regional integration (such as be-longing to a free trade agreement or sharing a com-mon currency). A disadvantage of using dummy vari-ables is that they may capture the impact of a rangeof other effects that occurred during the same timeperiod as the RTA. For example, most applicationsdo not distinguish the extent of multilateral trade lib-eralization. Ideally, specific trade policy variableswould be included in the estimating equations, suchthe level of multilateral and preferential tariffs. How-ever, the complexity of preferential trade arrange-ments precludes such an approach. A notable excep-tion is the study done by Estevadeordal andRobertson (2004), who included a measure of prefer-ential tariffs in their analysis of the impacts of RTAson regional trade in Latin America.

Although widely used because of its empiricalsuccess, the gravity model had lacked rigorous theo-retical underpinnings and was long criticized forbeing an ad hoc model. Recent theoreticallygrounded gravity equations are derived from modelswith strong constraints on preferences and technol-ogy, which undermines a straightforward interpreta-tion of some of the estimated coefficients. Andersonand van Wincoop (2004) provide a good overview ofthis debate. Another weakness of many applicationsof the gravity model is the proxying of trade costs bydistance, and the implicit assumption that cargoestraveling 1,000 miles in Africa face exactly the sametrade costs as similar cargoes traveling 1000 miles in,say, Europe.

Sources: Inter-American Development Bank 2002 and Bank staff.

RTAs create trade.4 This merits further analy-sis. Differing studies have produced sharplydifferent results for the same agreement. Forexample, Bayoumi and Eichengreen (1997)find no evidence of trade diversion from

enlargement of the European Union (to in-clude Greece, Portugal, and Spain), whereasWei and Frankel (1995) find “massive tradediversion.” One way to digest this contradic-tory literature is to combine and assess these

results in a single statistical analysis, calledmeta-analysis (box 3.2). This meta-analysis ofthe literature on the impact of regional tradeagreements on intra- and extra-regional tradeindicates that although agreements typicallyhave a positive impact on intra-regional trade,their overall impact is uncertain. Actualexperience reinforces that there can be nopresumption that a preferential trade agree-ment will be trade creating.

Do regional trade agreements benefitall members?The attention in most of the econometric stud-ies is on the impact of particular RTAs. Fewstudies have sought to estimate the impact ofRTAs on individual members. This is despitethe fact that studies of agreements that failedin the 1960s typically identify the lack ofmechanisms for redistribution in the presenceof asymmetric impacts as a crucial factor cre-ating political tension and undermining com-mitment to the agreement (Greenaway andMilner 1990). We estimated gravity equationsthat identify impacts for individual membersfor each of 17 different regional trade agree-ments to determine whether the statistical evi-dence suggests that the agreement has createdtrade, diverted trade, or had no significant neteffect on trade for each country.

For none of the agreements do we find un-ambiguous evidence of a net trade-creatingeffect extending to all members.5 Thus even ifan agreement as a whole creates trade, it isimportant that there are mechanisms to ensurethat all members benefit.

Regional trade agreements and exports tothe worldSo far the analysis has concentrated onwhether increases in intra-regional trade fol-lowing the signing of a RTA are associatedwith falling imports from the rest of the worldrelative to a scenario in which the RTA wasnot signed. It is equally important to ask howregional agreements can be used as part of abroad approach to openness and especiallywhether they can provide a springboard toglobal markets for local exporters.

Applying the gravity model with an addi-tional variable to capture overall exports of amember of a particular set of RTAs, we canassess whether these countries tend to exportproportionately more than would “normally”be the case for a similar country that was notparty to the agreements.6

These results, based on a sample period of1948 to 2000, show that different agreementsare associated with different propensitiesfor higher-than-“normal” overall exports(figure 3.5). AFTA, EC, GCC, MERCOSUR,NAFTA, and SACU all appear to export signif-icantly more than they would have done in theabsence of the agreement. The countries thatcomprise these regional groups appear to haveadopted policies that led them to be moreexport-oriented than they otherwise wouldhave been. We cannot say, however, that it wasthe RTA alone that led to these policies. Thevariables that pick up changes in trade flowsmay be capturing the effects of unilateral andmultilateral trade policies. Other agreements—CEMAC, CIS, COMESA, EAC, ECOWAS,and WAEMU—show a propensity to exportsignificantly less than “normal.” The AndeanCommunity and SADC appear to export lesswhen the whole sample period is considered,but not when the analysis is confined to themore recent sub-period, from 1980 to 2000. In


62

Change in welfare ($)

NAFTA

Nafta

�

Mer

cosu

rNaf

ta �

Mer

cosu

r � E

U

Figure 3.4 Simulated welfare impact ofvarious FTAs involving Chile

�1000

1000

0

500

�500

1500

2000

2500

Otherincluded

Chile

Sum for excluded

Mer

cosu

r

Source: Harrison et al (2004).

that period, the Andean Community alsoappears to have been more export-orientedthan they otherwise would have been, perhapsreflecting substantial trade policy revisions.

Most of the agreements in which exportpropensities are lower also appear to generatefewer imports than would “normal” countriesnot participating in the agreements (CEMAC,CIS, COMESA, EAC, WAEMU). At the sametime, those agreements that appear to be moreexport-oriented tend to be more open to im-ports (AFTA, EC, GCC, MERCOSUR since1980, NAFTA, SACU). In many cases, therehas been a strong impact on intra-regionaltrade. In general, members of regional agree-

ments that have been relatively open to im-ports have shown higher propensities to exportto the global market than would otherwise beexpected. Elsewhere, intraregional trade hasbeen initiated, but imports have been divertedand exports suppressed.

The potential gains from larger marketsand higher growthTrade of RTA members will be affectedthrough the changes in trade policies that takeplace, but will also change if there is an im-provement in technology, higher investment,and a higher rate of growth. By crudely usingdummy variables, gravity models provide a


63

Meta-analysis provides a means of assessing andcombining empirical results from different

studies. The approach takes as individual observa-tions the point estimates of relevant parameters fromdifferent studies. This set of observations is then usedto test the hypothesis that the relevant coefficient isstatistically different from zero. Here we are con-cerned with two parameters. The first measures theimpact of the agreement on total imports (which welabel overall impact); a negative value for this para-meter suggests that for the agreement concerned, thelevel of trade between a member and any other coun-try is less than the normal level of trade that onecould expect. Thus a negative value is evidence oftrade diversion. The second parameter captures theimpact of a regional trade agreement on the level oftrade between partners (internal impact). In ouranalysis we have included 254 estimates of overallimpact and 362 estimates of internal impact from

Box 3.2 Regional trade agreements in gravity models:A meta-analysis

17 research studies. The table below reports themean value of the overall and internal impacts, thestandard deviation, the number of statistically signifi-cant estimates, and the total number of estimates ofeach impact.

Of the estimates of the overall impact, 76 percentare statistically significant, 42 percent are negativeand significant, and 34 percent are positive and sig-nificant. For the internal impact, 66 percent of theestimates are statistically significant, 54 percent arepositive and significant, and only 12 percent are neg-ative and significant. The mean estimate of the over-all impact is negative. The most robust estimates ofthe overall impact are negative. The mean value ofthe internal impact is positive. For both parametersthere is a high degree of variance about the meanvalues. Within this analysis the estimates of 19 re-gional agreements were assessed; 10 exhibited onaverage net trade diversion.

Summary of the estimates by regional trade agreement

Overall impact Internal Impact

Mean Standard Significant Total Mean Standard Significant TotalValue error estimates estimates value error estimates estimates

Total �0.31 1.12 194 254 0.79 1.30 238 362


measure of RTAs, which catches all of thesefactors through their impact on trade but can-not distinguish the precise mechanisms. Com-plementary approaches look at the impact ofRTAs on these other factors.

Berthelon (2004), using cross-country regres-sions to estimate the effects of RTAs on growthduring 1960–99, found that RTAs that enlargedthe market substantially had substantial posi-tive effects on growth. The results suggest, forexample, that the FTA signed between Chileand the EU might be expected to increase thegrowth rate in Chile by 0.6 percentage pointsand in the EU by 0.005 percentage points. Thelarger market permits wider competition, largerscales, and greater specialization, all of whichincrease productivity and growth. South-Southagreements face an uphill struggle in two re-spects: they generally entail much smaller mar-kets, and they have less scope for realizing thegains from comparative advantage that differ-ent factor intensities would otherwise bring.

RTAs can also affect growth through tech-nological transfer. Trade raises total factorproductivity by providing access to a widerand more advanced range of technologies. Theproductivity of an importing country can in-crease through the importation of intermedi-ate goods, which as a result of R&D in theexporting country, are either new and/or ofbetter quality relative to existing products. Inthis way a country that is open to trade canbenefit from R&D activities undertaken over-seas. RTAs will have a positive effect if theystimulate imports from technological leaders.On the other hand, if the trade agreementleads to trade diversion away from moretechnologically advanced sources of inputs,then there could be a negative impact onproductivity growth.

Schiff and Wang (2003) found that, for Mex-ico, trade with NAFTA partners had a large andpositive impact on Mexico’s total factor pro-ductivity (TFP), while trade with the rest of the


64

�4 �2 0 2 4 6 8

Estimated exponential impact on trade

Note: The bars show the magnitude of the dummy variables capturing respectively the extent to which intraregional trade, overallimports, and overall exports differ from the “normal” levels predicted by the gravity model on the basis of economic size, proximity,and relevant institutional and historical variables, such as a common language.

10

Figure 3.5 RTAs that divert imports tend to export less to global markets

GCC

AFTA

Overall exports Overall importsIntra-regional trade

ANDEAN

CEMAC

Mercosur

SADC

SAPTA

CIS

EAC

WAEMU

COMESA

CACM

ECOWAS

EC

NAFTA

SACU


OECD did not. They suggest that this is becauseMexico not only benefited from the content oftrade with the NAFTA partners, the countryalso experienced closer contact and more infor-mation exchanges, especially among subcon-tracting firms, which are more integrated intothe production networks of their Northern part-ners than was the rest of the OECD. They simu-late the impact of NAFTA as a consequence and

find that it has led to a permanent increase inTFP in Mexican manufacturing of between5.5 percent and 7.5 percent.

In a later study, Schiff and Wang (2004)look at the dynamic impact of North-Southtrade on technology diffusion to Korea,Mexico, and Poland from the EU, Japan, andNorth America. Using industry level data,they found that technology diffusion and


65

The European Union and agricultureThe founding treaty (the Treaty of Rome), and subse-quent replacements, commit the European communi-ties to “the harmonious development of world trade,the progressive abolition of restrictions on interna-tional trade, and the lowering of customs barriers.”The European Union (EU) has failed to meet theseobjectives for agricultural products. There is littledoubt that EU agricultural policy has been the sourceof considerable disharmony among trading partners.

Movement of Moldovan wine through Ukraine Moldova is a major producer of wine. Although ithas a free trade agreement with Russia, its mainmarket, it costs more to ship a case of wine fromChisinau to Moscow than from Australia to Moscow(UNECE 2003). Why? Moldovan wine must passthrough Ukraine, usually by rail. Although the twocountries are party to the CIS free trade agreement,which provides for fair treatment in transit, theUkrainian authorities, in addition to imposing delaysand requiring unofficial payments, recently intro-duced an additional requirement that bulk winesmust be transported in specially heated railwaywagons, although a clear rationale for this is difficultto ascertain (World Bank 2004).

ASEAN and exclusions from preferences ASEAN members initially were allowed to excludecertain products from tariff reduction, a right thatthey exercised liberally. In many cases the tariffreductions offered were of very limited value to othermembers. Thailand’s offers, for example, includedwood products that it did not import and that othermembers did not produce. Malaysia’s list of products

Box 3.3 Implementation mattersfor tariff cuts included a number of rubber productsof which it was a major exporter. Indonesia, whichlies on the equator, offered a 10 percent cut in theduty on snow plows (Balasubramanyam 1989).More recently, the trade elements of the agreementhave been intensified with the launching of the AFTA(ASEAN Free Trade Area), the aim being to create agenuine free trade area. In 1995 the deadline forfully implementing AFTA was reduced from 15 to10 years, although there has been some backslidingrecently from agreed tariff-reduction schedules.

SADC and rules of originSADC initially agreed to simple, general, and consis-tent rules of origin similar to those of neighboringand overlapping COMESA. The initial rules requiredeither a change of tariff heading, a minimum of35 percent of value-added within the region, or amaximum import content of 60 percent of the valueof total inputs. Subsequently, however, the rules wererevised to include more restrictive sector- andproduct-specific rules. The requirement concerningchange of tariff heading has been supplanted by de-tailed technical-process requirements, a much higherdomestic value added requirement, and lowerpermitted import contents. The rules became muchmore similar to those of the EU and of NAFTA,reflecting in part the influence of the recently negoti-ated EU-South Africa agreement and the rules oforigin governing EU preferences to ACP countries(Flatters and Kirk 2003). This example illustrateshow sectoral interests and misperceptions of the roleand impact of rules of origin can undermine RTAs.


productivity gains tend to be regional. Apossible reason for this being that knowledgediffusion is also governed by close contactsand the hands-on relationships that are morelikely with neighbors. Nevertheless, for allcountries the biggest impact of trade onTFP can be guaranteed by removing tradebarriers on knowledge-intensive goods fromall countries.

Ingredients of SuccessOpen regions do betterRTAs are only effective for developing coun-tries if implemented in conjunction with morecomprehensive domestic reforms. At the sametime, a successful RTA will contribute to theoverall economic impact of that reform pro-gram. In Europe, the eight Central and East-ern European countries that recently joinedthe EU experienced strong growth in tradeand investment inflows during the 1990s; yettwo countries in the region, Bulgaria andRomania, having almost identical trade agree-ments with the EU but much less extensive do-mestic reform programs, saw a much weakertrade and investment response. Regional inte-gration initiatives in Latin America in the1990s have been much more effective thanearly efforts, reflecting broad and crediblestructural reforms in many countries (Devlinand French-Davis 1999). Given this context,there are a number of key features of RTAsthat are likely to contribute to favorable tradeoutcomes.

The external trade regime is a crucial de-terminant of the success of RTAs for severalreasons. First, trade diversion tends to fallwith the level of the external tariffs main-tained by member countries after they form apreferential trade agreement. The negativeeffects of trade diversion are offset or over-come if the preferential removal of trade bar-riers against some countries is accompaniedby a degree of liberalization to all countries,whether undertaken unilaterally or throughmultilateral negotiations. If a country that en-ters into a free trade agreement increases itsimports from all countries, not just its

partners in the agreement, then it will experi-ence an improvement in economic welfare.Therefore, countries forming preferentialtrade areas should simultaneously reduce thelevel of external protection facing nonmembertrading partners. Risks of trade diversion areparticularly high in the newly proposed SouthAsian Free Trade Area (SAFTA); (figure 3.6).

Second, where there are asymmetries in thelevel of external protection, it is importantthat the high-duty country reduce tariffs toavoid an adverse terms-of-trade shock. This isparticularly relevant for developing countriesseeking to sign agreements with the EU or theUnited States. In developed countries wheretariffs on manufactured products are ratherlow (and high-duty agricultural products aretypically excluded from regional preferences),trade diversion and trade creation are lesslikely to be significant. Thus with no tradebeing created in the developed market, the de-cline in domestic sales by firms in the high-tariff developing country may not be offset bya rise in exports to the developed country.Overall, the demand for goods produced inthe high-tariff country may fall, and its termsof trade could worsen.

Third, low MFN tariffs (and nonrestrictiverules of origin) ensure that producers withinthe regional trade agreement will have accessto competitively priced inputs. In today’s


66

0 5 10 15 20

Average tariff

Note: Tariffs are import-weighted at the country level toarrive at PTA averages.

Source: UN TRAINS, accessed through WITS.

25

Figure 3.6 Not all regions are open

SAPTA

COMESA

ECOWAS

EAC

MERCOSUR

AFTA

NAFTA

SADC

globalized market, policies that significantlyraise the input costs of producers will con-strain their exports to both regional andglobal markets. Regional integration is morelikely to be successful if it is achieved on thebasis of strong competition and ease of accessto low-cost inputs.

Trade liberalization is a crucial mechanismfor increasing competition in domestic mar-kets. Where it is not politically feasible toopen up broadly to all external suppliers, a re-gional approach can provide a stepping stonetoward the benefits of comprehensive liberal-ization. However, it is important to take thesecond step: Even in a large region such as theEU, competition from within the region hasbeen found to be much weaker than that pro-vided by external imports. Jacquemin andSapir (1991), for example, found that profitmargins in European countries were signifi-cantly dampened by external imports but notby intra-regional imports. And collusiveagreements are more difficult to enforce forcompanies based in distant locations. Firmsthat face little competition in local and re-gional markets will have low incentives toachieve the efficiency necessary to compete inworld markets.

Clearly RTAs may affect the setting of ex-ternal tariffs. This is true by definition in thecase of a customs union and indirectly true inthe case of a free trade area. Recent researchfinds that World Trade Organization (WTO)members do not appear to have more liberalexternal trade policies than non-WTO mem-bers (Rose 2004), and that membership in aRTA has, on average, no clear effect on a coun-try’s trade policy (Nitsch and Sturm 2003).Foroutan (1998), on the other hand, concludesthat countries in effective regional groupings,distinguished by the growth of intra-areatrade, have undertaken more far-reachingtrade liberalization. However, there are casesof liberalizing countries that did not belong toan RTA and of countries in an effective RTAthat did not liberalize trade policy. The conclu-sion is that the acceptance of a liberal tradepolicy may be a requirement for the survival

and deepening of a meaningful RTA, whereasbelonging to a regional scheme constitutes nei-ther a necessary nor a sufficient condition foran open and liberal trade regime.

In the 1960s and 1970s, preferential agree-ments among developing countries were typi-cally accompanied by high external tariff bar-riers as part of an import substitution strategy.In contrast, agreements among more devel-oped countries in the same period were moreoften associated with declining external barri-ers. For example, the simple average externaltariff of the original six members of the Euro-pean Union fell from 13 percent in 1958 to6.6 percent after the Kennedy Round of Gen-eral Agreement on Tariffs and Trade (GATT)negotiations. Agricultural products were ex-cluded from these reductions, reflecting theirexclusion from GATT negotiations until theUruguay Round. The failure to reduce agricul-tural tariffs in Europe led to substantial tradediversion in agriculture with significant wel-fare losses for European consumers, especiallythe poorest, and a considerable hardship forpoor farmers in developing countries.

Many developing countries have since re-duced external tariff barriers both unilaterallyand through multilateral negotiations. As aresult, recent preferential agreements amongmany developing countries have been intro-duced or revamped with lower external barri-ers. This is particularly true in Asia and LatinAmerica, where preferential and MFN tariffsdeclined in tandem after 1985, so that marginsof preference remained stable or were slightlycompressed (figure 3.7).

Paper agreements are not enough Important aspects in the assessment of RTAsare whether their members have implementedtheir objectives under the agreement and theextent to which the objectives in the agree-ment have been met. Often the objectives in anagreement are defined by foreign ministers oreven prime ministers, while the way that thoseobjectives are to be carried out is determinedlater in negotiations between ministries. If tar-iff concessions are subsequently negotiated


67


68

Average tariff rates (percent)

Note: MFN tariffs represents the simple average of themost-favored-nation tariffs applied by the following 11 LatinAmerican countries (based on country averages): Argentina,Bolivia, Chile, Colombia, Ecuador, Mexico, Paraguay, Peru,Uruguay, and Venezuela. Preferential tariffs represent theaverage preferential tariff that each country applies to theother countries in this sample under different regional tradeagreements. Calculations include only ad valorem tariffs.

Source: Estevadeordal and Robertson 2004.

MFN and preferential tariff liberalizationLatin America, 1985–1997

19971985 1987 1989 19931991 1995

Figure 3.7 Preferential tariffs in tandemwith all tariffs in Latin America

0

20

10

30

40

50

Preferentialtariffs

MFN tariffs

sector by sector or item by item, the processbecomes cumbersome and open to capture bydomestic interests. The distinction is oftenmade between agreements that reduce dutiesonly on products specified in a positive listand other agreements, typically more liberal,implemented on the basis of a negative list ofproducts excluded from tariff reduction.

Sectoral accords within RTAs can curbmarket forces and limit the benefits fromcompetition. For example, Ozden and Parodi(2003) found that the auto agreement embed-ded in MERCOSUR between Argentina andBrazil compelled companies in both countriesto balance trade, ensuring that productionwould not be reallocated to the lowest costproducer (Brazil); this move secured thesupport of the companies for the agreement.Because a new entrant would have to buildplants in both countries (not just one), theagreement acted as a barrier to competitionthat favored insiders.

North-South agreements appear to have abetter track record than South-South agree-ments. The comprehensive tariff objectives ofmost North-North agreements signed beforethe mid-1980s were implemented on or aheadof schedule (table 3.1). In contrast, South-South agreements reached during thisperiod—most based on limited positive lists ofproducts for tariff liberalization—had a veryweak record of implementation. The delays inimplementing initial regional tariff commit-ments “generally reflected a basic incompati-bility between the inward-oriented develop-ment strategies of most members and regionalliberalization”(De la Torre and Kelly 1992).

A larger number of South-South agree-ments signed or substantially revised in thelate 1980s and early 1990s have sought amuch broader degree of internal tariff liberal-ization, have been more effective in imple-menting agreed-on tariff reductions, and havetended to reduce external tariffs. For exam-ple, the GCC, launched in 1982, was origi-nally intended to become a free trade area—agoal achieved by 1983. By the late 1980s,however, the objective evolved into formationof a customs union, which was established in2003 (see World Bank 2003). However,table 3.1 also reports that substantial prob-lems with implementation remain in many ofthe regional agreements involving developingcountries.

Nonrestrictive rules of origin are integralto successPreferential rules of origin are integral to pref-erential trade agreements. However, it has be-come increasingly clear that rules of origincan be designed in a way that restricts tradebeyond what is necessary to prevent tradedeflection or the transshipment of productsfrom third countries through a member forthe purpose of obtaining preferential duties.In addition, the proliferation of free tradeagreements with accompanying rules of originis increasing the burdens on customs servicesin many countries, and these burdens haveconsequent implications for trade.

Table 3.1 Implementation of tariff commitments by type of agreement, 1960–1999

Agreement Objective on intra-bloc tariffs Implementation record

North-North agreements reached from 1960–89

ANZERTA (signed 1983) Eliminate all tariffs by 1988 On scheduleEuropean Economic Eliminate all tariffs by 1968 Ahead of schedule

Community (signed 1957)U.S.-Canada FTA (1988) Eliminate all tariffs by 1999 Ahead of scheduleEFTA (1960) Eliminate all tariffs on manufactures by mid 1967 On schedule

South-South agreements, 1960–89Andean Pact (1969) Eliminate tariffs on positive list Postponed several timesCentral American Common Elimination of tariffs Initially on schedule, most duties

Market (1960) removed in the early 1970s, butrestrictions reintroduced in the 1980s

EAC (1967–1977) Establishment of a common market The Community was dissolvedLatin American Integration Liberalization of common lists of products by 1972 Common lists not liberalized on

Association scheduleECOWAS Tariff liberalization by 1990 Progress negligibleASEAN (1967) FTA based on positive lists Repeatedly postponedGCC (1982) FTA Virtual elimination of all tariffs in 1983

South-South agreements, 1990–99AFTA (1992) Gradual reduction of tariffs over 12–15 years Liberalization took place ahead of

according to member-specific schedules original scheduleCACM—Revised (1991) Customs union Implementation postponed; progress

uneven among membersGCC Customs union begins in 2003; completed by 2005 Customs union established on scheduleCOMESA Progressive tariff elimination to be completed Implementation varies by country, 9

by 2000 out of 20 members have moved toduty-free trade

MERCOSUR Elimination of all tariffs by 1995 All lines, with the exception of sugarand automobiles, have been liberalized

SAPTA Limited tariff concessions from a country-specific No formal schedules have been adoptedpositive list

SADC Tariff liberalization by 2008, with sensitive lists Implementation delayed in some sectors eliminated by 2012 due to lack of agreement on rules of

originCEMAC (1999) Economic Union Tariffs liberalized according to schedule

in nearly all linesWAEMU (2000) Economic and monetary union Tariff liberalization mostly on schedule

North-South agreements, 1990–99Europe Agreements Country-specific tariff removal schedules in Bulgaria mostly on schedule, Romania

(Bulgaria, Romania) preparation for the EU membership continues to have some unresolvedissues

NAFTA Tariff elimination in stages to be complete by 2008 On schedule EU-Mexico Progressive tariff elimination by 2010 On schedule EU-South Africa FTA establishment by 2012 Partial implementation pending

official ratificationU.S.-Chile Progressive tariff elimination by 2015 N/A



69

In general, the rules of origin in North-South agreements are more restrictive thanthose adopted by South-South agreements(Figure 3.8). A feature of both EU and NAFTAagreements is the high degree of variation inrules of origin across product categories. Dif-ferent rules are specified for different products:

sometimes the rule may be a change of tariffheading, sometimes a change of tariff chapter;for other products there will be a value-addedrequirement; and in others the rules of originmay specify a particular technical process.

The amount of the required value addedcan vary across products. The change of tariff

Anson and others (2004) and Carrere andde Melo (2004) estimate that the administra-tive costs of providing the documentary evi-dence to support the certificate of origin underNAFTA are in the region of 1.8 percent of thevalue of exports. The distorted impact of therules, resulting from the need to use local andhigher cost inputs to qualify, may be equiva-lent to an average duty of around 4.3 percent.Thus, restrictive rules of origin can very easilywipe out any margin of preference generatedby a trade agreement. Other things beingequal, compliance costs are lowest for rules in-volving a change of tariff heading, followed byvalue-added rules. Rules requiring a specifictechnical process have the highest compliancecosts.

Estevadeordal and Suominen (2004)introduce a synthetic measure of therestrictiveness of rules of origin (the basis forfigure 3.8) into a standard gravity model of bi-lateral trade flows. Their econometric analysisleads them to conclude that restrictive prod-uct-specific rules of origin undermine overalltrade between preferential partners and thatprovisions such as cumulation7 and de minimisrules,8 which increase the flexibility of apply-ing a given set of processing requirements,boost intraregional trade. Applied at the sec-toral level, this approach yields support for thehypothesis that the restrictiveness of rules oforigin for final goods stimulates trade in inter-mediate products between preferential part-ners and diverts trade away from nonmem-bers. Cadot and others (2002) find that forsectors where tariff cuts are larger than aver-age, the rules of origin are more restrictive andthe rate of use of preferences by Mexican ex-porters lower than average. They concludethat rules of origin are the “prime culprit” forthe very modest impact of NAFTA on Mexicanexports identified by other researchers.

Deeper agreements can lead to larger tradeand income effectsIn principle, agreements that address a widerrange of barriers can have a greater impact ontrade flows and incomes.


70

classification can be used to provide a positivetest of origin by stating the tariff classificationof imported inputs that can be used in theproduction of the exported good. Or it maybe defined to provide a (more restrictive) neg-ative test by stating cases where a change oftariff classification will not confer origin. Forexample, in the EU rules of origin, bread, bis-cuits, and pastry products (heading 1905 ofthe Harmonized System) can be made fromimported products of any other tariff headingexcept those of chapter 11, which includesflour, the basic input to these products.

Specifying rules of origin on a product byproduct basis offers opportunities for sectoralinterests to influence the specification of therules in a protectionist way. The outcome ofhighly detailed product-specific rules of originis typically a complex set of rules, which canbe highly restrictive. Box 3.5 provides an ex-ample of the sort of complexity that can arise.Many agreements involving developing coun-tries, on the other hand, tend to specify gen-eral rules that apply to all products. TheAFTA, COMESA, and ECOWAS, for exam-ple, have a single value-added rule applicableto all products.

Index of restrictiveness

Note: Higher values of the index equal more restrictiverules of origin [derived from Estevadeordal and Suominen(2004)].

NAFTA

EU-Mex

ico

EU-Chil

e

SADC

Chile-

CACMAFTA

COMESA

ECOWAS

Figure 3.8 Rules of origin in North–Southagreements are more restrictive than inSouth–South agreements

0

2

1

3

5

4

6


71

Subsequent chapters elaborate on the po-tential economic impacts of dealing withmany of the regional agreement issues intro-duced in chapter 2. Here we simply ask

whether deeper agreements have a signifi-cantly greater impact on aggregate merchan-dise trade than more narrow trade agree-ments. Two studies9 assume a productivity

Here is an example of what rules of origin looklike; the following pertains to men’s or boys’

overcoats made of wool (HS620111).

A change to subheading 620111 from any otherchapter, except from heading 5106 through 5113,5204 through 5212, 5307 through 5308 or5310 through 5311, Chapter 54 or heading 5508through 5516, 5801 through 5802 or 6001through 6006, provided that: The good is bothcut and sewn or otherwise assembled in theterritory of one or more of the Parties.

The basic rule of origin stipulates change of chap-ter but then provides a list of headings and chaptersfrom which inputs cannot be used. Thus in effect,the overcoat must be manufactured from the stage ofwool fibers forward, because neither importedwoolen yarn (HS5106-5110) nor imported woolenfabric (HS5111-5113) can be used. However, therule also states that imported cotton thread(HS5204) or imported thread of man-made fibers(HS54) cannot be used to sew the coat together. Thisrule in itself is very restrictive; however, the rule isfurther complicated by requirements relating to thevisible lining:

Except for fabrics classified in 54082210,54082311, 54082321, and 54082410, the fabricsidentified in the following subheadings and head-ings, when used as visible lining material in cer-tain men’s and women’s suits, suit-type jackets,skirts, overcoats, car coats, anoraks, windbreak-ers, and similar articles, must be formed fromyarn and finished in the territory of a party: 5111through 5112, 520831 through 520859, 520931through 520959, 521031 through 521059,521131 through 521159, 521213 through521215, 521223 through 521225, 540742through 540744, 540752 through 540754,540761, 540772 through 540774, 540782

Box 3.4 Restrictive rules of origin under NAFTA—the case of clothing

through 540784, 540792 through 540794,540822 through 540824 (excluding tariff item540822aa, 540823aa or 540824aa), 540832through 540834, 551219, 551229, 551299,551321 through 551349, 551421 through551599, 551612 through 551614, 551622through 551624, 551632 through 551634,551642 through 551644, 551692 through551694, 600110, 600192, 600531 through600544 or 600610 through 600644.

This stipulates that the visible lining used must beproduced from yarn and finished in either party’slocation. This rule may well have been introduced toconstrain the impact of the tolerance rule, whichwould normally allow 7 percent of the weight of thearticle to be of nonoriginating materials. In overcoatsand suits, the lining is probably less than 7 percentof the total weight. Finally, it is interesting to notethat the rules of origin also provide very specificexemptions for materials that are in short supply orare not produced in the United States. In this regard,the rule reflects firm-specific lobbying to overcomethe restrictions of these rules of origin when the orig-inal NAFTA rules were defined. The most extremeexample is the following, where the apparel will bedeemed eligible for tariff preferences if assembledfrom imported inputs of:

Fabrics of subheading 511111 or 511119, ifhand-woven, with a loom width of less than76 cm, woven in the United Kingdom in accor-dance with the rules and regulations of the HarrisTweed Association, Ltd., and so certified by theAssociation.

Clearly, the job of the relevant official to checkconsistency and compliance with such rules is not asimple one.



72

response to trade liberalization when othermeasures are included and ascribe the resultsto the deep integration measures. The calcula-tions of these ex ante simulation studies illus-trate the potential that deeper agreementsmay hold when they produce a productivityresponse, with changes in trade flows and in-comes being a multiple of that under pref-erential tariff removal. However, becausethis result is inevitable from the way thatdeeper integration is modeled (i.e., inducingeconomy-wide increases in productivity),these results from one or another deep inte-gration measure should be seen as indicativeof potential rather than evidence of success.

Ex post exercises based on the gravitymodel will tend to capture all of the policy re-lated impacts of a regional trade agreementon trade, not just the removal of trade policyvariables. Several authors have tried tocapture in an index the differences of depthbetween agreements; it is then used as thedummy variable in the gravity model to cap-ture the impact of RTAs [Li (2000), Adamsand others (2003)]. However, this approachpresents the issue of how to weight differentpolicy measures—for instance, should ser-vices liberalization get more weight than cus-toms cooperation? Thus the value of theindex would be dependent on the subjectiveweights that are assigned. The weights chosenby Adams and others lead to EFTA beingranked as much more restrictive than the An-dean Pact or NAFTA. Further, many agree-ments appear extensive on paper but haveaccomplished little in practice.

Extensive monitoring of agreementsis crucial to ensure effectiveimplementationIn order to assess the impact of RTAs, infor-mation is needed on the extent to which theagreement’s provisions are being implementedand how they are affecting decisions by pro-ducers and consumers. Given the need formonitoring, an implementation scorecardwould be useful—such an approach has beenadopted by the EU Commission which, as

part of its monitoring of the implementationof the single market, has introduced the“Single Market Scoreboard” (box 3.5). Inaddition to providing vital information, thescorecard is useful as a disciplinary measure—to shame governments with a record of poorimplementation into action and to empowergovernments with good records of implemen-tation to challenge those members who arenot meeting their commitments.

More extensive monitoring could make animportant contribution to the implementationof many trade agreements. Lack of effectiveimplementation has been a major factor lim-iting the impact of many trade agreements inAfrica, South America, South Asia, NorthAfrica, and the CIS.

Conclusions: Preferential TradeAgreements and EconomicDevelopment

This review of the experience of preferen-tial trading agreements over the past 40

years offers the following conclusions:

• There is no strong evidence to supportthe claim that a preferential trade agree-ment will be net trade creating or that allmembers will benefit. Positive outcomesdepend on design and implementation.

• When embedded in a consistent andcredible reform strategy, the key deter-minant of regional trade agreements’success is low levels of external tradebarriers. While many developing coun-tries have reduced tariffs, they remainhigh in many countries and regions, andthe risk of trade diversion remains sig-nificant. Further reductions in appliedMFN tariffs will be required to ensurethat regional agreements are beneficialfor those participating in them and tominimize the impact on the countriesthat are left out.

• Trade agreements that provide for com-prehensive liberalization of trade across


73

all major sectors and nonrestrictive rulesof origin are more likely to be successful.Agreements that devote considerable re-sources to negotiating limited positivelists or large negative lists and detailed

product specific rules of origin limit thescope for gains.

• Effective implementation is crucial to pos-itive outcomes, yet implementation iscompromised by proliferation. If different

The Single Market Scoreboard measures (1) theextent to which Single Market directives have

been transposed into national law by each memberstate, (2) the average time it takes each member totranspose directives, and (3) the extent to whichmembers are cooperating with enforcement andproblem solving. This analysis by the EuropeanCommission is supported by regular surveys of busi-nesses and individuals on perceptions of the SingleMarket and where it is not working. The Commis-sion also monitors differences in prices of identicalgoods for indications that integration is leading toconvergence.

The left figure below shows the implementationdeficit for each member; that is, the proportion ofdirectives that have not been notified as having beentransposed into national law. The figure shows a

Box 3.5 Monitoring implementation of preferential tradeagreements: “Single Market Scoreboard” in the EuropeanUnion

substantial improvement in implementation since1997; it also shows that the original six members ofthe EU and Greece are currently the worst offenders.Effective monitoring of implementation also requiresthat clear targets be established. In 2001, the EUHeads of State established an interim target of a1.5 percent implementation deficit. As of July 2004,only five members had achieved this target.

A further measure of implementation is the extentto which agreed-on rules are being properly applied.In Europe, the Commission is charged with monitor-ing when Single Market rules are not being appliedcorrectly; the Commission also takes infringementcases against member countries that are breaking EUlaws. In terms of the number of infringement casesopen in May 2004, Italy and France are the worstoffenders (lower right figure).

Percentage rate of nonimplementation

Source: http://europa.eu.int/comm/internalmarket/score/indexen.htm#score.

Italy

Spain

Germ

any

Fran

ce

Greec

e

Belgium

Irelan

dUK

Nethe

rland

s

Austri

a

Portu

gal

Luxe

mbo

urg

Sweden

Finlan

d

Denm

ark

Single Market Scoreboard

0

4

2

6

10

8

12

Open infringement case (May 2004)

Italy

Spain

Germ

any

Fran

ce

Greec

e

Belgium

Irelan

dUK

Nethe

rland

s

Austri

a

Portu

gal

Luxe

mbo

urg

Sweden

Finlan

d

Denm

ark

0

60

40

80

20

120

100

160

140

b. EU law breakersa. Implementation of single market directives by EU members

1997

2004

agreements have different product cover-age, different liberalization schedules,and different rules of origin, the ability ofagencies such as customs to apply theagreements is severely undermined. Thecapacity to effectively implement is acrucial issue that countries should con-sider before signing an RTA.

• Monitoring can play an important rolein providing for effective implementa-tion, but often there is insufficient moni-toring as well. Technical reviews are fre-quently not done, and when reports aremade, senior officials fail to act on theirrecommendations.

Notes1. Flatters and Kirk (2003).2. The conclusion is unchanged if intra-EU and

intra-NAFTA trade are excluded from the total ofworld exports.

3. Drawn from Burfisher and others (2004) andHarrison and others (2004).

4. For example, Ghosh and Yamarik (2004) sug-gest that “a consensus has emerged among researchersthat RTAs are trade creating.”

5. In this exercise, where the counterfactual is basedon the historical pattern of trade flows, we assess howthe regional trade agreement affected trade flows afterits introduction. As a measure of robustness of the ef-fects, we used three different estimation methods.Effects are considered statistically robust only if allthree methods generate a significant impact of the samesign. The three methods are pooled OLS with robuststandard errors. The second estimation method includescountry-pair fixed effects using a specific OLS method.The third approach is a pooled Tobit estimation.

6. Here we follow Soloaga and Winters (2001),who include an additional dummy variable to assessthe impact on the exports of members of regional tradeagreements, although their focus is on the welfareeffects of RTAs. However, here we apply a panelapproach to a sample period of 1948 to 2000, cover-ing bilateral trade between 178 countries with country-pair fixed-effects, which a number of authors, althoughnot all, suggest is the preferred method. We apply theabove equation. The regional dummies are time sensi-tive; that is, they are relevant only after the agreementhas been signed. Using a different estimation tech-nique, such as Tobit and OLS, and a different sampleperiod can lead to different results for a particularagreement but the overall conclusion remains firm.

7. The basic rules of origin define the processing thathas to be done in the individual beneficiary or partner toconfer origin. Cumulation is an instrument allowing pro-ducers to import materials from a specific country or re-gional group of countries without undermining the originof the final product. In effect the imported materials fromthe identified countries are treated as being of domesticorigin of the country requesting preferential access. Thereare three types of cumulation, bilateral, diagonal (or par-tial), and full. The most basic form of cumulation is bi-lateral cumulation, which applies to materials providedby either of two partners of a preferential trade agree-ment. In this case originating inputs, that is materials,which have been produced in accordance with the rele-vant rules of origin, imported from the partner, qualify asoriginating materials when used in a country’s exports tothat Partner. Second, there can be diagonal cumulationon a regional basis so that qualifying materials from any-where in the specified region can be used without under-mining preferential access. Finally, there can be full cu-mulation whereby any processing activities carried out inany participating country in a regional group can becounted as qualifying content regardless of whether theprocessing is sufficient to confer originating status to thematerials themselves. Under full cumulation all of theprocessing carried out in participating countries is as-sessed in deciding whether there has been substantialtransformation. Hence, full cumulation provides fordeeper integration among participating countries.

8. De Minimis or tolerance rules allow a certainpercentage of nonoriginating materials to be usedwithout affecting the origin of the final product. Thus,the tolerance rule can act to make it easier for productswith nonoriginating inputs to qualify for preferencesunder the change of tariff heading and specific manu-facturing process rules. This provision does not affectvalue added rules.

9. For example, Hoekman and Konan (1999) findthat a free trade agreement between the EuropeanUnion and Egypt limited to goods (but with substantialprogress on removing regulatory barriers) could raisewelfare by around 4 percent while an agreement thatreduced barriers to services in Egypt could raiseeconomic welfare by over 13 percent. Similarly,Brenton, Tourdyeva, and Whalley (2002) find that anEU-Russia FTA limited to tariff removal would in-crease welfare by around one-tenth of one percentwhile a comprehensive agreement removing technicalbarriers to trade and barriers to trade in services wouldraise welfare by more than 13 percent.

ReferencesAdams, Richard, Philippa Dee, Jyothi Gali, and Greg

McGuire. 2003. The Trade and Investment Ef-


74


75

fects of Preferential Training Arrangements—Oldand New Evidence. Productivity CommissionStaff Working Paper, Canberra, Australia.

Anson, Jose, Olivier Cadot, Antoni Estevadeordal,Jaime de Melo, Akiko Suwa-Eisenmann,and Bolorma Tumurchudur. 2004. Rules of Ori-gin in North-South Preferential Trading Arrange-ments with an Application to NAFTA. ResearchUnit Working Paper 0406, Laboratoire d’E-conomie Appliquee, INRA, Paris.

Balasubramanyam,V. N. 1989. ASEAN and RegionalTrading Co-operation in Southeast Asia. InEconomic Aspects of Regional Trading Arrange-ments, eds. D. Greenaway, T. Hyclak, and R. J.Thornton. London: Harvester Wheatsheaf.

Berthelon, Matias. 2004. Growth Effects of RegionalIntegration Agreements. Processed.

Brenton, Paul, Natalia Tourdyeva, and John Whalley.2002. Economic Impact of a FTA between Russiaand the EU: Numerical Simulations Using aGeneral Equilibrium Trade Model. Report pre-pared for the European Commission, Brussels.

Brenton, Paul, and H. Imagawa. 2004. Rules of Ori-gin, Trade and Customs. In The Customs Mod-ernisation Handbook, ed. L. De Wulf, andJ. Sokol. Washington, DC: World Bank.

Burfisher, Mary, Sherman Robinson, and KarenThierfelder. 2004. Regionalism: Old and New,Theory and Practice. Forthcoming in Agricul-tural Policy Reform and the WTO: Where AreWe Heading? eds. G. Anania, M. E. Bohman,C. A. Carter, and A. F. McCalla. Cheltenham,UK: Edward Elgar.

Cadot, O., J. de Melo, A. Estevadeordal, A. Suwa-Eisenmann, and B. Tumurchudur. 2002. Assess-ing the Effect of NAFTA’s Rules of Origin.Processed.

Carrere, Celine. 2004. “Revisiting Regional Tradingwith Proper Specification of the Gravity Model.”European Economic Review (forthcoming).

Carrere, Celine, and Jaime De Melo. 2004. “Are Dif-ferent Rules of Origin Equally Costly? Estimatesfrom NAFTA.” CEPR Discussion PaperNo. 4437. London.

De la Torre, A., and M. Kelly. 1992. Regional TradeAgreements. Occasional Paper 93, Washington,DC: IMF.

Devlin, R., and A. Estevadeordal. Forthcoming. Tradeand Cooperation: A Regional Public Goods Ap-proach. In Regional Public Goods: From Theoryto Practice, eds. A. Estevadeordal, Brian Frantz,and Tam R. Nnguyen. Washington, DC: Inter-American Development Bank.

Devlin, R. and R. French-Davis. 1999. Towards andEvaluation of Regional Integration in Latin

America in the 1990s. The World Economy 22:261–90.

Estevadeordal, A., and R. Robertson. 2004. Do Pref-erential Trade Agreements Matter for Trade? InIntegrating the Americas: FTAA and Beyond,eds. A. Estevadeordal, D. Rodrik, A. M. Taylor,and A. Velasco. Cambridge, MA: Harvard Uni-versity Press.

Estevadeordal, A., and K. Suominen. 2004. Rules ofOrigin: A World Map and Trade Effects. In TheOrigin of Goods: Rules of Origin in PreferentialTrade Agreements, eds. A. Estevadeordal,O. Cadot, A. Suwa-Eisenmann, and T. Verdier.,DC: Inter-American Development Bank.

Evenett, S. J., and W. Keller. 2002. On Theories Ex-plaining the Success of the Gravity Equation.Journal of Political Economy 110 (2).

Foroutan, Faezeh. 1998. Does Membership in aRegional Preferential Arrangement Make aCountry More or Less Protectionist? The WorldEconomy 21(2):305–35.

Flatters, F., and R. Kirk. 2003. Rules of Origin asTools of Development? Some Lessons fromSADC. Paper presented at INRA conference onRules of Origin, Paris, May 2003.

Ghosh, S., and S. Yamarik. 2004. Are Regional Trad-ing Arrangements Trade Creating? An Applica-tion of Extreme Bounds Analysis. Journal ofInternational Economics 63: 369–95.

Glick, Reuben, and Andrew Rose. 2002. “Does aCurrency Union Affect Trade? Time Series Evi-dence.” European Economic Review 46:1125–51.

Greenaway, David, and Chris Milner. 1990.South–South Trade Theory, Evidence, and Pol-icy. The World Bank Research Observer 5(1):47–68.

Harrison, Glenn W., and David G. Tarr. 2003. Rulesof Thumb for Evaluating Preferential TradingArrangements: Evidence from ComputableGeneral Equilibrium Assessments. The PolicyResearch Working Paper 3142, World Bank,Washington, DC.

Hoekman, Bernard, and Denise Eby Konan. 1999.Deep Integration, Nondiscrimination, and Euro-Mediterranean Free Trade. Policy ResearchWorking Paper 2130, World Bank, Washington,DC.

IADB (Inter-American Development Bank). 2002. Be-yond Borders: The New Regionalism in LatinAmerica. Washington, DC: IADB.

Li, Quan. Institutional Rules of Regional Trade Blocsand Their Impact on Trade. Processed.

Ozden, Caglar, and Francisco Parodi. 2003. CustomsUnion and Foreign Investment: Theory and

Evidence from Mercosur’s Auto Industry. Work-ing Paper, Emory University.

Palmeter D. 1997. Rules of Origin in Regional TradeAgreements. In Regionalism and Multilateralismafter the Uruguay Round: Convergence, Diver-gence, and Interaction, eds. P. Demaret, J.F. Bellis,and G. Garcia Jimenez. Brussels: European In-teruniversity Press.

Pelkmans, J., and Brenton, P. 1999. Bilateral TradeAgreements with the EU: Driving Forces andEffects. In Multilateralism and Regionalism inthe Post-Uruguay Round Era: What Role for theEU? eds. O. Memdovic, A. Kuyvenhoven, and W.Molle. Boston: Kluwer.

Soloaga, Isidro, and Alan Winters. 2001. Regionalismin the Nineties: What Effect on Trade? NorthAmerican Journal of Economics and Finance 12 (1).

Schiff, Maurice, and Yanling Wang. 2003. Regional In-tegration and Technology Diffusion The Case ofthe North America Free Trade Agreement. PolicyResearch Working Paper 3132, World Bank,Washington, DC.

Volker Nitsch, and Daniel Sturm. The Trade Liberal-ization Effects of Regional Trade Agreements.

UNECE (United Nations Economic Commission forEurope). 2003. Economic Survey of Europe.Processed.

Wonnacott, R. 1996. Trade and Investment in a Hub-and-Spoke System Versus a Free Trade Area.World Economy 19: 237–52.

World Bank. 2003. Trade, Investment, and Developmentin the Middle East and North Africa: Engagingwith the World. Washington, DC: World Bank.

World Bank. 2004. DTIS Moldova, Case Study of theWine Sector.


76

The removal of tariffs and quotas is a key fea-ture of regional trade agreements (RTAs), butmodern RTAs can, and are, being designed toachieve much more than that. Trade policiesare only one element—and often a relativelyminor one—of the overall costs of trade. Be-cause logistical, institutional, and regulatorybarriers are often more costly than tariffs andgenerate no offsetting revenue, cooperativegovernmental efforts to improve customs pro-cedures, minimize the trade distorting impactof standards, and reduce transport costs mayhave a higher payoff than reciprocal reduc-tions in overt trade policy barriers.

When RTA membership is part of a broadprogram of economic liberalization in whichthe objective is to attract international invest-ment as much as to promote trade, a broad setof regulatory issues becomes paramount.Which are the most appropriate institutions toaddress these regulatory barriers? In certaincases, institutions at the regional level willprovide for the most effective solutions, rela-tive to both the multilateral and nationallevels.1 RTAs can effectively promote dialogueand implement coordinated responses.

However, most RTAs have contributed littleto reducing the associated trade costs, espe-cially RTAs among developing countries.Many regional policy initiatives havefoundered because of the lack of effective im-plementation, and crossing borders betweenmost developing countries is still a majorimpediment to trade.

This chapter focuses on three key issues re-lated to trade facilitation: customs clearance,transport, and standards and their conformityassessment. Coordinated action among devel-oping countries is likely to be greatest with thefirst two issues, customs clearance and trans-port; examples of best practices in these areasare available and can be followed in regionaltrade agreements. Progress in reducing barri-ers is likely to facilitate trade to and from alltrading partners with little or no scope to bediscriminatory. And while cheaper, faster, andmore predictable customs clearance and im-proved transport services have a direct impacton trade, they are also crucial elements of theinvestment climate.

Initiatives to deal with standards and con-formity assessment on a regional basis arescarce; the most successful agreementshave been between rich countries that areundertaking a deep integration process, as inthe case of the European Union (EU).Nonetheless, systems of standards, qualityassurance, accreditation, and measurementare crucial to competitiveness and sustainedgrowth. Regional interventions can be use-ful if developed in a transparent way andwith the participation of private groups (toensure that procedures are not manipulatedto serve a protectionist end). Initiativestargeted at a small number of key sectorsand toward improving the quality of confor-mity assessment are likely to be the mostuseful.

77

Beyond Trade Policy Barriers:Lowering Trade Costs Together

4

Agreements that involve large markets andhave differing levels of institutional capacitygenerally appear to have had the greatest suc-cess in dealing with these trade facilitationissues. This is because the more advancedpartner tends to drive institutional improve-ments among the less advanced partners.(However, a real danger is that, in seekinggreater access to industrial country marketsthrough bilateral trade agreements, develop-ing countries agree to apply a set of rules andregulations defined by the advanced countrythat are inappropriate for their level of devel-opment.) For many developing countries,agreements with industrial countries alonewill not be sufficient, because the main sourceof higher trade costs are the borders and theweak transport systems they share with theirdeveloping country neighbors.

Progress often requires coordinatedactions; for example, joint customs inspec-tions must be allowed, common rules fortransport must be established (including vehi-cle weight restrictions), and test results frompartners’ laboratories must be accepted. RTAscan provide a forum to enhance trust amongtrade partners that genuinely wish to moveforward on these and other fronts.

The Costs of Trade

Despite globalization and the rapid increasein trade over the past 40 years, the costs

of trading remain substantial—particularly fordeveloping countries (box 4.1). Because ofthose costs, the actual volume of internationaltrade is far less than economic theory wouldpredict in the absence of significant barriers totrade [the case of the “missing trade” (Trefler1995)]. And trade within countries is muchmore intense than between countries. If tradecosts were insignificant, the propensities totrade nationally and internationally would beequal. In fact, crossing a national borderappears to dampen trade flows even in regionssuch as the EU, where formal trade barriersand customs posts have been removed.2

Finally, the retail prices of particular goods

tend to diverge with distance, and this differ-ence is much higher when the two locationsbeing compared lie on either side of a nationalborder. If trade costs are low, then arbitrageshould constrain such price variation (Engeland Rogers 1996).

The tax equivalent of trade costs can rangefrom 30 to 105 percent, depending on the sec-tor, according to estimates for imports by theUnited States (Anderson and van Wincoop2004; Evans 2001). High trade costs discour-age investment and constrain the ability of localfirms to integrate into global production chains(Faini 2004). Given the magnitude of thesebarriers, ex ante simulation studies suggest thatthe benefits of lowering transaction costs, re-ducing insecurity, integrating services sectors,and increasing competition are multiples ofreducing tariffs (Hoekman and Konan 1999).However, there is very little convincing ex postevidence of significant returns to regional ini-tiatives that must deal with these issues, sug-gesting that substantial progress is difficult toachieve.

Ignoring institutional barriers during a tar-iff reform may undermine the objectives ofreform—and indeed produce perverse results.For example, tariff liberalization in the face ofborder delays and customs corruption mayhave no impact on imports and may even re-duce welfare if tariff revenues are replaced bylonger waits to clear customs (Cudmore andWhalley 2003). In the absence of competitionin the domestic transport sector, trade liberal-ization may simply lead to a transfer of rev-enue from the government to monopolistictransport owners. On the other hand, progresson many issues is not possible while high tariffbarriers remain in place.

Cost raising barriers may be linked in cir-cles of causation, with significant impacts dueto scale economies in transport. For example,a reduction in tariffs or a decline in costs at theport may stimulate trade that can offer oppor-tunities for transport companies to operate atmore efficient levels of scale. And if there is ef-fective competition in the transport sector, thiscould lead to lower transport prices and more


78

B E Y O N D T R A D E P O L I C Y B A R R I E R S

79

Box 4.1 Trading can be costly• Costs incurred when crossing a border due to

documentation, delays, and bribes to corruptofficials.

• Compliance with national product standardsand technical regulations.

• Insurance against risk, especially credit risk,and uncertainty associated with macroeconomicinstability, lack of effective institutions, andunpredictable politics.

Various policies and factors isolate nationaleconomies from world markets and thereby raise

the cost of international trade:

• Tariffs, quantitative restrictions, and other bor-der barriers—such as taxes on trade that raisethe prices of imported goods relative to thoseproduced domestically.

• Transport costs, both direct (freight andinsurance) or indirect (inventory costs).

trade, and so on.3 A reduction in corruptionand delays at the border may stimulate trade,add to government revenues, and allow for areduction in tariffs to achieve a given revenuetarget, which again stimulates trade.

Landlocked countries that face high barriersin moving their imports and exports throughneighboring countries have no choice but topursue bilateral or regional solutions.4 Theseneed not be embedded in a regional preferentialtrade agreement (PTA), but to be effective forsmall countries, agreements must provide forthe settlement of disputes. Such provisions arelikely to be more effective if they are part of abroad and comprehensive agreement.

Finally, removing institutional obstacles tocrossing borders has a more certain benefitthan reducing intra-regional barriers, becauseit saves real resources. Trucks that make moredeliveries to the port are more productive. In-terventions that lead to higher productivityhave the greatest impact on trade andwelfare—and on further increases in productiv-ity. In contrast, removing revenue-generatingtariff barriers on a preferential basis can leadto trade diversion and reductions in welfare.

Regional Agreements to FacilitateTrade and Transport

As countries develop their trade beyond theexport of basic agricultural and extracted

commodities, logistics requirements become

more important—and more costly. To com-pete in international markets and functionwithin global production chains, firms neednot only low transport costs and efficientports, but also short transit times, reliable de-livery schedules, appropriate storage facilities,and security (Carruthers and others 2003).

High transport costs, inefficient or corruptcustoms, and long delays at borders reducethe trading opportunities available to manydeveloping countries and can have significanteconomic and social costs (box 4.2). Con-versely, better conditions tend to be related tohigher levels of trade (Wilson and others2003). Increasing the efficiency of customs,for example, can reduce costs and increasetrade (figure 4.1). High transport and bordercrossing costs thwart, in particular, the poorlandlocked developing countries.5

Regional integration can help promotemore efficient and effective customsoperationsUnlike many other factors that raise tradecosts, there is broad agreement on what con-stitutes good customs procedures. Since itsinception, the World Customs Organization(WCO) has developed best practices of cus-toms policies and procedures. The Kyoto Con-vention commits its signatory members to im-plement these best practice principles andprovides them with guidance in their efforts toimprove national practices. While there is


80

much that countries can do individually to im-prove customs procedures, there is also scopefor regional initiatives to modernize customs.

Contacts fostered by regional agreementscan generate a mutual understanding of each

Box 4.2 Border delays tax trade• Crossing a border in Africa can be equivalent

to the cost of more than 1,000 miles of inlandtransportation; in Western Europe theequivalent is 100 miles (Arvis 2004).

Border delays are associated with other tradecosts as well, especially corruption in customs—andhave been linked to the spread of HIV/AIDS. TheWorld Bank has recently initiated a project in West-ern Africa to reduce, by the end of 2006, the averagetime for commercial vehicles to clear border formali-ties along the Lagos–Abidjan corridor by at least20 percent, and average delays at the Nigeria–Beninborder by at least 50 percent. These reductions arecritical for this project and its mandate to reduce theincidence of sexually transmitted infection amongcommercial vehicle drivers.


Each day lost in transport delays is equivalent to atax of about 0.5 percent (Hummels 2000). The

situation in crossing borders between developingcountries can be much worse.

• In Southern Africa, delays at the main border-crossing between South Africa and Zimbabwe(Beit Bridge) amounted to six days in February2003, leading to an estimated loss of earningsper vehicle of $1,750, equivalent to the costs ofa shipment from Durban to the United States.

• In Central Asia, on average, it takes morethan 100 hours to cross the border betweenUzbekistan and Turkmenistan. A truck travel-ing from Tashkent to Berlin, passing throughTurkmenistan, Iran, and Turkey, will spend, onaverage, a third of total transport time waitingat border crossings (UNESCAP 2003).

• In the Andean Community, trucks spend morethan half of the total journey time at bordercrossings (Pardo 2001).

Source: Investment Climate Surveys data and Global Trendsas cited in Subramanian and others 2003.

Figure 4.1 More efficient customs areassociated with more trade

Ratio of trade to GDP (percent)

00 5 10

Days through customs, imports

R2 � �0.338

15 20 25

50

150

100

250

200

other’s problems and difficulties and can en-gender the sharing of best practices and posi-tive experiences among members. This ex-change is likely to be more relevant and betteraccepted inside the regional group of develop-ing countries than examples from countriesthat are much more advanced and face verydifferent implementation issues.

On the other hand, when regional units aremade up of developed and developing coun-tries, there is scope for financial support andtechnical assistance for less developed coun-tries in their modernization efforts. The EU,for instance, provides assistance to theAfrican, Caribbean, and Pacific (ACP) coun-tries and incorporates customs technical assis-tance provisions in its Euro-MediterraneanInitiative. Such assistance will also be availableunder the Economic Partnership Agreements itintends to establish with regional groupings inAfrica, such as ECOWAS. Similarly, Japanprovides funding for capacity-building initia-tives in APEC member countries.


81

For a variety of reasons, tackling customsissues autonomously may be too daunting atask, and cooperation with trading partnersmay create the necessary momentum to over-come reluctance and opposition from domesticpolicymakers, customs officials, and traders.A review of a number of regional initiatives tomodernize customs suggest the following areasin which RTAs can lead to improvements:

• Align customs codes with internationalstandards. A good customs code sup-ports efficient customs operations. It es-tablishes the competence of the relevantauthorities, promotes transparency andpredictability of operational proceduresand enforcement, encourages coopera-tion with the private sector, provides foreffective appeals procedures, and en-hances integrity. It would be advanta-geous for all countries to align theircustoms codes with internationalstandards.

• Simplify and harmonize procedures. Therecommendation here is to introduce asingle customs document that limits thedata requirement to a single set andadopts e-commerce techniques.

• Bring all tariff structures in line withthe international harmonized tariff clas-sification (HS). Many disputes can beavoided if all members of the groupingadhere strictly to an identical tariffclassification.

• Strive for transparency. Increase theavailability and accessibility of the legaltext and regulations that traders andcustoms officials require and includeother relevant information such as tradestatistics.

• Adopt and effectively implement theWTO Valuation Agreement. Membercountries can assist each other througheffective mutual assistance agreementsand shared databases.

• Work together toward customs integrity. • Establishment of joint border posts.

Joint border posts preclude multiple

examinations and lengthy border cross-ing procedures. A simple first stagewould coordinate hours of operation andprovide compatible computer systems onboth sides of the border; these effortswould increase efficiency significantly.

• Joint training centers. Countries can joinforces to operate regional training cen-ters that can benefit from leveraged-upresources and can build cohesion be-tween the customs officers of differentcustoms services in the region.

Transport and trade facilitationinitiatives raise productivity In recent years there has been a developmentof a web of transport and trade facilitation(TTF) agreements aimed at easing the move-ment of goods and services across borders.Most of these agreements have been reachedas part of, or in parallel with, an RTA.6

Effectively implemented, TTFs can improveaccess to global markets for developingcountries with poor transport systems—particularly landlocked countries.

TTFs often contain provisions to standard-ize customs procedures at borders and to har-monize customs documentation. TTFs canfurther facilitate trade by providing for theinteroperability of transport resources and byfostering market access and competition inthe transport sector.

Divergent national regulations for truck sizeand weight require vehicle checks on both sidesof the border and often lead to overloadcharges and costly delays (box 4.3). InSouthern Africa, for example, axle-load regu-lations are different in Namibia, Botswana,and Zambia (Röschlau 2003). Truckers whoare in full compliance in one country can beprosecuted and fined across the border.Regulations concerning insurance, driver’slicenses, and other documentation provideample opportunities for cost savings. For ex-ample, the COMESA carrier’s license systemallows companies to operate regionally with-out having to pay for multiple licenses. AndCOMESA’s vehicle insurance scheme enables

transport operators to comply with insuranceobligations throughout the region with a singlepolicy. Similarly, ECOWAS’s brown card sys-tem, introduced in 1982, has helped to reducesettlement time significantly. The success ofsuch initiatives requires effective cooperationbetween different ministries (transport, inte-rior) in the member states.

The impact of harmonizing customs proce-dures and transport rules may be limitedunless there is competition in the domestictransport sector. In the extreme case of a

domestic transport monopoly, the gains fromlower operating costs and more efficient cus-toms procedures may accrue to the transportcompany in the form of greater monopolyprofits—with little impact on trade andpoverty. Equally important is competition be-tween routes and between different modes oftransport. For example, Lao goods werealmost exclusively exported through Vietnamuntil the Lao People’s Democratic Republicdeveloped alternative transit routes throughThailand. It now takes one day for an export


82

Box 4.3 Standardization and simplification can increase trade volumes:The case of the Trans-Kalahari Corridor

resulted in agreements (October 2000) to extend theoperating hours of customs at the Namibia/Botswanaborder from 22 to 24 hours to enable loading andunloading in Windhoek and crossing the border inthe same day.

In August 2003, the TKC started a pilot phase toreplace all existing transport documents with a singleadministrative document (SAD). To complement thiseffort, South African Customs developed a websitewith details on the SAD process. Border processingtimes were cut by more than half, from an averagetime of 45 minutes to 10–20 minutes. According tothe United States Agency for International Develop-ment (USAID) estimations, reduced border delayscreated savings of $2.6 million per year along thecorridor. As a result, the route had become economi-cal, and traffic flows increased. Operators were mov-ing about 620,000 tons annually along the TKC,about 65 percent of expected capacity, until theBotswanan government increased road user chargesin February 2004. In some cases, road charges weremultiplied by a factor of 10. The customs problemhad been settled, but following this unilateral deci-sion affecting the transport sector, traffic decreasedsignificantly.


The Trans-Kalahari Corridor (TKC), the roadroute between Gauteng province (South Africa)

and Walvis Bay (Namibia) via Botswana was openedin 1998, replacing the traditional longer routethrough western South Africa. Despite major roadrehabilitation in 1999, traffic reached only 15 per-cent of the expected capacity. The major obstaclesoccurred at the border crossings. This led the TKCCorridor Management Group to seek a partnershipwith the customs administrations of Namibia,Botswana, and South Africa. This partnership

Cape Town

Durban

Johannesburg Swaziland

Gaborone

Pionee

rs

Gate

Skilpadshek

Windhoek

Trans-KalahariMamuno

Ramatlabama

Ramatlabama Maputo

WalvisBay

Trans-Kalaharicorridor offices/border posts:

South Africa

Namibia

Botswana

Moz

ambi

que

Zimbabwe

Walvis BayWindhoekTrans-KalahariMamunoGaboronePioneers GateSkilpadshekRamatlabamaRamatlabamaJohannesburg

Lesotho

container loaded with garments fromVientiane to reach a main international trans-port node at Bangkok, compared with three tofour days to reach Danang (Banomyong2000).

Although competition is often included inregional treaties (such as the 1982 ECOWASconvention regulating interstate road trans-portation or the 1993 COMESA KampalaTreaty),7 effective implementation is rare. Be-cause national authorities often fear a loss ofsovereignty if they allow foreign operators inthe market, a regional legal framework and ef-fective enforcement mechanisms are necessaryfor successful implementation. In the case ofthe EU, full implementation was not achieveduntil 1985—28 years after the signing of theTreaty of Rome. Since then, the benefits havebeen substantial (box 4.4).

In several West African countries, transitregimes are governed by national, bilateral, or

customs frameworks, rather than regionalarrangements (UEMOA Commission 2000).8

Many transport companies oppose the adop-tion of regional frameworks for fear ofupsetting the “tour de role” system.9 Bilateraltransport treaties often predefine the transportshare of both countries (normally 50-50); theexporter therefore has no choice in selecting atransport operator. The effect of this system isto protect less efficient operators. Even themore sophisticated North America Free TradeAgreement (NAFTA) has experienced difficul-ties in implementing cross-border truckingcompetition, following protectionist pressuresfrom U.S. unions, concerns about trucksafety, and Mexican driver qualifications andcompetence.

MERCOSUR countries implemented the“International Common Manifesto Cargo andCustoms Transit Dispatch” (IMC/CTD) in1991. This form harmonized and unified all


83

Box 4.4 Logistics costs in Europe have fallen in the last two decades

Since the late 1980s the proportion of companyrevenues spent on logistics in Europe has declined

from 14.3 percent to 6.8 percent, a reduction thatfar exceeds the average level of EU external tariff onmanufactured goods (Mentzoni 2003).

This development reflects the following importanttrends in the logistics industry:

• Centralization of inventory through a smallernumber of warehouses;

• Increased outsourcing of logistics services tospecialized companies; and

• Just-in-time supply policy (Ruijgrok 2001).

These changes have followed key policy initia-tives. The removal of borders within the EU reduceduncertainty and the costs of international transport.At the same time, the EU acted to increase competi-tion in the industry through the adoption ofcabotage—a policy that enables carriers to operate

Percentage of revenue

0

10

5

15

Source: ELA/ATKearney.

1987 1993 1998 2003

Logistics cost in Europe

Administration and planning

Transport

Warehousing

1.5

5.8

74.6

4.1

1.4

1.2

3.4

3.1 3.1

2.6

1.1

domestically in all member countries, regardless oftheir country of registry.


information required by different bordercontrol institutions (customs, migratory; seeNofal 2004).

Regional cooperation reinforces transportand trade facilitation programsTrade liberalization—whether unilateral, mul-tilateral, or regional—may have a very mutedeconomic impact in the presence of very hightransport costs, weak logistical services, andlong delays to clear customs. Conversely, inthe presence of high trade barriers there maybe little reason for traders to lobby for im-provements in transport. And trade restric-tions that limit quantities may undermine theincentive to invest in improved transport andtrade facilitation services. Hence, actions toimprove transport and reduce trade barriersare often complementary.

Trade facilitation can be effective even inthe absence of a formal RTA. However, a for-mal agreement may help to entrench and en-hance facilitation initiatives beyond what ispossible through cooperation alone. In princi-ple, unilateral and multilateral liberalizationby itself should lead to larger trade volumes—and hence raise incentives to invest in tradefacilitation. By inducing a more thorough

dismantling of trade barriers among neighbor-ing countries, regional cooperation can createa broader constituency for facilitating tradeflows. Integrating fragmented markets canmake infrastructure projects more viable, andthus promote a virtuous cycle of integrationand growth.

Realizing the inherent potential of trans-port and trade facilitation requires both asound institutional environment and a con-ducive economic one. RTAs can help addressinstitutional gaps and reinforce the corridorapproaches that are common in agreementsbetween developing countries. At the sametime, RTAs can provide a forum for the dis-cussion and definition of norms and harmo-nized rules that are often necessary for effec-tive implementation. Involvement of theprivate sector in these discussions is often aprerequisite for effective action (box 4.5).Finally, negotiating transport issues in a re-gional forum can act to depoliticize issues(Schiff and Winters 2002).

Despite the enormous potential for gainsfrom regional initiatives to improve customsand transport services, progress in many caseshas been slow. Trade costs remain very high formany developing countries. Many initiatives


84

Box 4.5 The case of the Northern CorridorStakeholders Consultative Forum

• Development of a one-stop processing center;and

• Reduction of the number of required stamps togo through Mombasa port (from 21 to 11).

As a result of this forum, national transit and tradefacilitation committees are being established in the re-gion. Private sector participation has been extended toinclude insurance clearing agents, bank associations,shippers’ council, and the like. Public/private partner-ships to tackle trade and transport facilitation are alsobeing established in West Africa.


Since 1999, officials dealing with transport, transit,and private operators along the Northern Corri-

dor (including Ministry of Transport, Ministry ofTrade, customs agencies, exporters, and importersassociations, etc.) have been regularly meeting twicea year to discuss transit issues. This private-publicsector alliance has produced the following positivedevelopments:

• Elimination of charges on imports routedthrough the port of Mombasa (by KenyaBureau of Standards and the Kenya PlantHealth Inspectorate Service);

to facilitate trade have suffered from a lack ofeffective implementation. In several agree-ments, disputes over implementation can onlybe raised at the political level, which oftenmeans that small landlocked countries havegreat difficulty in securing the necessary com-pliance from larger neighbors. In these casesimplementation is very much a function of po-litical will.

The possibility of taking legal actionunder regional treaties can help drive imple-mentation of transport facilitation initiatives.The European Court of Justice has played animportant role in the implementation of acommon transport policy in Europe (Funck1998). A regional court of justice has re-cently been established in the Eurasian Eco-nomic community of the CIS; another re-gional court exists in UMEOA in West Africa.While a regional court of justice does notguarantee implementation, it does create po-tential for more efficient enforcement than isavailable through less formal dispute settle-ment channels.

Standards, ConformityAssessments, and RTAs

The construction and implementation ofsystems of standards, quality assurance,

accreditation, and metrology are crucial tocompetitiveness and sustained growth—andhence to development. Standards have becomekey elements for facilitating transactions andtrade both within countries and in interna-tional exchange between countries. Standardssupport markets and provide for efficienttransactions. Standards and technical regula-tions stipulate what can or cannot be ex-changed, and they define the procedures thatmust be followed for exchange to take place.10

The ability of would-be exporters to com-ply with mandatory health and safety stan-dards, as well as market-driven voluntarystandards in overseas markets, is a major fac-tor determining access to those markets.Divergent product standards and duplicativesystems for assessing conformity with those

standards can constitute substantial barriersto trade, but these may only become clearafter other barriers have been addressed. Re-ducing tariffs and improving customs andtransport can be likened to reducing the waterlevel in a swamp only to find a range of previ-ously covered “snags and stumps that need tobe cleared away” (Baldwin 1970).

When producers must alter their product tomeet divergent standards in foreign markets,they lose some of the benefits of larger scalesof production. When the foreign governmentdoes not recognize standards-compliance testsperformed in the exporter’s home market, orthe home country does not have the facilitiesto test the product, the exporter must foot thebill for additional tests in the foreign market.For example, in Moldova the certification oforganic nut production exported to Germanyhas to be renewed every 6 months, and eachvisit from an international certifying companycosts $5,000 plus $2,000 per productiontest—once before processing and once afterprocessing. This can amount to $18,000 peryear, which is a heavy burden for firms in aneconomy such as Moldova, an economy tryingto compete in international markets. Upgrad-ing testing facilities and measuring equipmentis essential for reducing the costs of confor-mity assessments.11

To reduce the damping effect of divergentstandards on international trade, WTO mem-bers have agreed to discipline the use ofmandatory standards by governments. Theseare relatively modest provisions—they dealwith transparency of standards regimes, equaltreatment, and the need to justify standardsthat differ from internationally agreed-onnorms. Efforts to reduce barriers to tradecaused by standards and conformity assess-ment have been more extensive in a smallnumber of RTAs, although empirical evidenceidentifying the benefits of these interventionsis scant at best. The issue for developing coun-tries is whether regional initiatives can providefor more efficient and effective standards andconformity assessment systems. These im-proved systems would allow governments to


85

meet domestic objectives to raise health andsafety levels, and at the same time, facilitatetrade.

Different paths to better standards systems Different approaches are available to raisestandards and to address technical barriers totrade. Countries can unilaterally upgradestandards by adopting international stan-dards. However, the technological content andthe health, security, and environment objec-tives of the international standard may not beappropriate for developing countries, becausethe international standards are strongly influ-enced by the OECD countries. Further, someof the returns to adopting the internationalstandard—in terms of greater market access—only materialize if the country’s trading part-ners also accept products produced to thatstandard.

A second approach requires cooperationbetween countries to upgrade standards.Countries agree that products satisfying par-ticular standards will be accepted in eachother’s markets. However, cooperation agree-ments do not discipline other market accessbarriers, so that returns from the upgrading ofstandards may be undermined if other barriersare raised to protect a particular sector oncestandards are harmonized. Typically, the dis-pute panels in cooperation agreements have amediation role, not an arbitration role. Forthese reasons, and unless all parties are com-mitted to the upgrading process, the process ofstandards upgrading could have importantobstacles.

The upgrading of standards within a RTAis characterized by more formal institutions, ahigher degree of enforcement, and greatertrust originating from the frequent interac-tions between members and the comprehen-sive nature of the agreement. Members cannotuse tariffs to prevent the entry of a productsatisfying the regional standard; this increasesthe certainty that a country’s upgrading effortswill be translated into greater market access.Within PTAs different approaches to stan-dards have been followed, which reflect the

different levels of development and institu-tional capacities.

We start by discussing the EU experience,where integration has proceeded the furthest.The basis for the free movement of goods inthe EU is the principle of mutual recognitionof the regulations of partners or the recogni-tion of equivalence. Although standards varyfrom one country to another, it is presumedthat they are designed to meet the same regu-latory objectives and to offer equivalent levelsof protection to the public. Thus productsproduced in partner countries can be acceptedwith the assumption that those products willnot undermine basic regulatory objectivesconcerning health, safety, and the environ-ment.12 Mutual recognition of regulations isthe simplest approach to differences in stan-dards: it is a powerful tool for removing bar-riers to trade in goods and services, and withthis approach the difficulties of detailed har-monization measures, which intrude on na-tional policy making, can be avoided.13 How-ever, mutual recognition of standards requiresa high degree of trust between regulatory au-thorities (essentially the responsibility for pro-tection of domestic consumers is, in part,transferred to the overseas partner). As such,mutual recognition can only work in regionscomprising countries of similar levels of in-come that have comparable standards.

Effective institutions are also important. Inthe EU, governments can defer from nondis-crimination and the free circulation of goodsfor reasons of “public policy or public security”and protection of health as long as such re-strictions are not a disguised restriction ontrade. To ensure the latter, the EU has devel-oped the following mechanisms for disciplin-ing national regulations and interventions intoproduct markets:14

• Infringement procedures, whereby theEuropean Commission acts to enforcecommunity law, although such proce-dures are very time consuming and costly,have an impact only after the event andare ad hoc.


86

• Notification procedures, whereby mem-ber states are required to notify all drafttechnical regulations for scrutiny by anEU Committee, whose objective is to pre-vent new regulatory barriers to trade. Inpractice, all new national regulations ofEU member states have to pass an EUtest regarding their impact on the freemovement of goods.

• Notification of derogation procedures,which require member states to notifyauthorities of cases in which they wish toprevent the sale of goods lawfully pro-duced or marketed in another memberstate, on the grounds of nonconformityand nonequivalence with domestic re-quirements. This ensures that any dero-gations from the principle of mutualrecognition are transparent and subjectto scrutiny.

While mutual recognition of regulationsunderpins the EU Single Market, it has beenapparent for a long time that for certain prod-ucts and for certain risks (when consumers aredirectly exposed to hazards), equivalence be-tween levels of regulatory protection embod-ied in national regulations cannot be assumed.In these cases the EU seeks agreement amongmembers on a common set of legally bindingrequirements. EU legislation harmonizingtechnical regulations has involved two distinctapproaches, the “old” and the “new.”

The old approach mainly applied to prod-ucts (chemicals, motor vehicles, pharmaceuti-cals, and foodstuffs), involved extensiveproduct-by-product or even component-by-component legislation, and was carried out bydetailed directives. Achieving this type of har-monization was slow for two reasons. First,the process of harmonization became highlytechnical, with attention given to very detailedproduct categories. Consultations were oftendrawn out. Second, the adoption of directivesrequired unanimity in the Council, whichmeant that they were slow to be adopted. Thelimitations of the old approach as a broad toolfor tackling technical barriers to trade become

apparent in the 1970s and early 1980s, whennew national regulations were proliferating ata much faster rate than the production ofEuropean directives harmonizing regulations(Pelkmans 1987).

It became clear that the degree of interven-tion by the public authorities before a productwas placed on the market needed to be re-duced, and that changing the decision-makingprocedure to allow the adoption of harmo-nization directives by a qualified majority inthe Council was needed. The “new approach”regulations indicate only “essential” healthand safety requirements, allowing greater free-dom to manufacturers to satisfy the essentialrequirements and to industries to flesh outproduct specifications in the form of volun-tary standards. The new approach makesgood use of established standardizationbodies—European Committee for Standard-ization (CEN), European Committee for Elec-trotechnical Standardization (CENELEC),and the national standards bodies. Standard-ization work is achieved in a more efficientway, is easier to update, and involves greaterparticipation from industry. Products thatconform to the standards promulgated by theEuropean standards agencies are presumed tocomply with the essential requirements of theregulations. However, these standards are vol-untary, and firms can produce to differentstandards if they can prove compliance withthe requirements of the regulation.

MERCOSUR has followed the oldapproach of the EU and focused its limitedresources on harmonizing national stan-dards at the regional level (see Nofal 2004).MERCOSUR has formulated 366 commontechnical regulations and some 300 voluntarystandards. The Andean Community has re-cently decided to focus regional harmoniza-tion on a targeted number of standards—thoseof the products most traded. Only 40 regionalstandards were created, although they cover60 percent of trade.

When harmonized regulations are pursued,it is important to avoid overly bureaucraticmechanisms. Harmonization through the use


87


88

of detailed regulations can lead to excessive in-tervention by public authorities before a prod-uct can be placed on the market and have achance to prove its viability. Regulators shouldconcentrate on defining essential health andsafety requirements while allowing firms theflexibility to meet those requirements and notstifle technological change and competitive-ness. The CIS, MERCOSUR, and AndeanCommunity countries still apply an approachbased on very detailed harmonized regula-tions. When a company wishes to introduce anew product, it is often necessary to changethe existing regulations or wait for a new tech-nical regulation to be promulgated, which cantake considerable time and be very costly. A re-cent Peruvian technical regulation for gas con-tainers specifies the minimum width of thewalls, stating the exact thickness, which effec-tively prevents the use of new materials thatmight be lighter but thicker.

ASEAN is also following a policy of tar-geting key sectors, but it is harmonizingaround international standards rather thanpromulgating its own standards. For 20 keyproduct groups, members should adopt, asnational regulations, the agreed-on interna-tional norms. Members that do not adopt anyof the identified international standards astheir national standards still need to acceptproducts from partners that comply withthese international standards—unless theycan demonstrate an inability to adopt the in-ternational standard due to “climatic condi-tions or infrastructural reasons.”

Here the contribution of the RTA has beento provide an enforcement mechanismthrough dispute settlement procedures, suchthat members who do not accept from part-ners products that satisfy an ASEAN standardare ultimately liable to fines for compensationor removal of concessions. Therefore, ASEANcountries can adopt the ASEAN standardswith confidence that incurring the associatedcosts will not be undermined by subsequentdenial of access to partners’ markets.

In agreements where regional institutionsare weak, especially free trade areas, barriers

to trade can be removed when standards fromdifferent members are shown to be compati-ble. This is the approach of NAFTA and alsotends to be applied in bilateral trade agree-ments that have a standards component, suchas those between Chile and the EU and Chileand the United States.15 The compatibility ap-proach is the converse of the mutual recogni-tion of regulations. Under the compatibilityapproach, the standards of a trading partnerare assumed to be insufficient in their abilityto satisfy the importer’s regulatory objectives,unless proven otherwise.

Recognizing the results of conformityassessment in partners Mutual recognition of conformity assessment[usually negotiated in the form of a mutualrecognition agreement (MRA)] is necessary ifnontariff barriers are to be fully removed.This ensures that the test results from labora-tories in the exporter’s home market are ac-cepted by the importer so that the costs ofduplicative testing can be avoided. This agree-ment does not require that both countrieshave the same standards nor that both coun-tries be members of a PTA. For example, theEU and the United States, for certain sectors,accept the results of product tests (for com-patibility with their own standards) that havebeen completed in the partner’s laboratories.

If conformity assessment institutions arerelatively weak, however, harmonization ofstandards may be a necessary step to facilitatemutual recognition of conformity assessment.This is the approach being followed inASEAN. The Andean Community establisheda regulation for compulsory mutual recogni-tion in 2003 for sectors covered by regionalstandards. MERCOSUR will proceed withmutual recognition of conformity assessmentprocedures in the near future. However,MERCOSUR is a perfect example of how theconformity assessment infrastructure is lack-ing: Policymakers prefer to harmonize stan-dards before moving to mutual recognition ofconformity assessment, but firms do not showmuch interest in standards because, in the


89

absence of mutual recognition of conformityassessment, the returns to investment in stan-dards are low. This suggests that improve-ments in the conformity assessment infra-structure are necessary.

Singapore has signed a bilateral tradeagreement whereby the United States recog-nizes certifications provided by Singapore tosome of its East Asian partners. This high-lights that rules of origin can be an importantelement in an MRA. If Singapore has a com-parative advantage in the region in testingand laboratory facilities and is well endowedwith professional staff in this activity, then theU.S. agreement with liberal rules of origin(whereby the United States accepts tests fromSingapore labs of products from other coun-tries), may help to establish or enhance theposition of Singapore as a regional hub fortesting and conformity assessment. Rules oforigin that restrict the testing and conformityactivities to products produced only inSingapore would tend to constrain such a de-velopment. EU MRAs tend to have theserestrictive rules of origin.16

Regional trade agreements can facilitatemutual recognition of standardsEffective solutions to problems arising fromdifferent standards require a high degree of di-alogue and trust among trading partners.RTAs, while not the only path to trust, tend topromote dialogue and communication, whichin turn build trust. This has been the case formember countries of MERCOSUR and theAndean Community, in which trust has grownas integration has deepened (Nofal 2004).Such trust needs to be nurtured through open-ness and transparency when new nationalregulations are being considered.

RTAs also can provide a favorable negoti-ating environment and so reduce politicizationin standard disputes among members, makingit easier to find common solutions for the re-moval of non-tariff barriers. The interactionsthat take place in an RTA often improve insti-tutional relationships between the differentstandards bodies of member countries—and

sometimes even between the institutions ofindividual countries. These close relationshipsallow obstacles to be overcome in informalways, circumventing cumbersome formal in-terventions. MERCOSUR has yet to adoptmutual recognition of standards, but manyconflicts over standards have already beensolved by telephone between relevant officialsin member countries.

A degree of trust between public institu-tions and the private sector is also importantif more flexible approaches, such as mutualrecognition of conformity assessment and/orof regulations, are to succeed. Strong, central-ized regulatory cultures tend to produce tech-nical regulations that are too detailed and dif-ficult to change. It is important to ensure theeffective participation of the private sectorand consumers so that the setting of stan-dards and their enforcement reflect broadrather than narrow interests.

Successful cooperation in harmonizingstandards depends on simple principles To date, RTAs in the developing world havenot realized their full potential for overcom-ing standards-related obstacles to regional orglobal trade, although some slow progress isevident, such as in MERCOSUR. That islikely to change as the WTO agreements onTechnical Barriers to Trade (TBT) and Sani-tary and Phytosanitary Measures (SPS) comeinto full practical application, and as the im-portance of reforming standards systems indeveloping countries gains prominence. In themeantime, several principles can contribute tosuccessful cooperation in standards and con-formity assessment procedures.

A first step for developing countries is toidentify priority sectors for reform to keepcosts low and gather momentum for furtherreform. The sectors to prioritize are thosewhere trade costs resulting from differences instandards and conformity assessment proce-dures are higher and where trade betweenmembers is large.

Second, if international standards exist forthese products and they are appropriate for

product costs (World Bank 2003). This hurtsMENA exporters and prevents MENA coun-tries from joining global production chainsdestined for EU markets.

Many developing countries are too small toefficiently offer a full range of these confor-mity assessment services. Often there are toomany laboratories offering poor qualityservices.

RTAs can contribute to better conformityassessment first, and most simply, by facilitat-ing dialogue and the sharing of technicalknowledge. More ambitious initiatives canbuild regional accreditation bodies to increaseefficiency and enhance the reputation of localcertification bodies in the global market. Anopen regional market for laboratory servicescan lead to cheaper yet higher quality testingon the basis of specialization and economies ofscale. However, it must be stressed that whilethe potential gains are large, there are very fewsuccessful initiatives that can provide usefulguidelines. ASEAN provides an example wheremembers are pushing forward with a numberof initiatives, including cooperation on legalmetrology, and efforts to enhance conformityassessment bodies to facilitate mutual recogni-tion of test reports and certifications.

RTAs can also provide a framework forcollaboration that increases the effective par-ticipation of developing countries in interna-tional standards organizations and at theWTO. This is important if international stan-dards and conformity assessment measuresare to reflect the interests of developingcountries—and therefore make the TBT andSPS agreements relevant for the majority ofWTO members. For this participation to besuccessful, the structure of international stan-dards institutions needs to be modified toreduce the costs of representation of develop-ing countries. The International StandardsOrganization (ISO) has moved one step in thisdirection by being the first standard institu-tion to allow electronic voting. This could beextended to other institutions. Another possi-bility would be to allow RTAs to representtheir members in standard-setting committees,


90

all members given their level of development,then the simplest approach is to harmonizearound these international standards. Thiswill only be relevant if all members have thecapacity to implement them. It is important toclearly define objectives before harmonizingstandards so as to avoid overregulation. Forexample, requiring information on labels thatthe consumer is unlikely to understand willincrease costs and contribute little to theobjective of making information available toconsumers.17

An open and transparent system, withstandards published in an accessible officialbulletin before being implemented, is essentialif regional initiatives are to facilitate tradebetween members and preclude the difficul-ties facing exporters from outside of the re-gion. Ensuring flexibility is also important;thus, regulations that set minimum standardsrather than detailed requirements are less re-strictive for firms. When countries face inter-nal resistance to modernizing and adoptingnew standards, the compliance with stan-dards can be offered on a voluntary basis.This allows those firms that are able and will-ing to satisfy the new standards to progress.Daskalov and Hadjikolonov (2002) showhow such an approach made it possible formore advanced and competitive local pro-ducers in Bulgaria to quickly adopt Europeanstandards before they were formally intro-duced as mandatory Bulgarian standards.

In the short run, the greatest gain for manydeveloping countries is likely to come fromimprovements in the testing, certification, andaccreditation institutions to underpin greaterenforcement capacity. Initiatives that improvethese institutions are likely to have large pay-offs by allowing governments to achieve moreeffectively their existing objectives concerninghealth and safety and facilitating greater ex-ports on both a regional and a global basis.For example, most MENA countries requirethat testing be done at their national labora-tories, which are usually less sophisticatedthan European testing centers and saddledwith cumbersome procedures, pushing up


91

which would reduce members’ costs ofrepresentation.

Attempts to remove barriers caused by dif-ferences in standards and conformity assess-ment requirements will be more effective in aclimate of trust and mutual understanding.Such a climate requires a genuine willingness toliberalize and is unlikely to result in agreementswith many exceptions, frequent recourse tosafeguard measures, and high barriers at theborder that can be due to customs delays andinefficient port and transport services.

The lack of relevant examples of successfulintervention at the regional level to deal withstandards issues makes it difficult to deriveclear proposals based on best practices. In thislight the best recommendation is for develop-ing countries to proceed with caution andconcentrate on targeted coordinated actionfor which the institutional requirements arenot extensive and the gains are clear.

Trade-Related RegionalCooperation Agreements

Countries can benefit from other forms ofcooperation that are linked to trade di-

rectly through RTA arrangements or indirectlywhen they influence trade-related inputs oroutputs.18 Such trade-related cooperation candeal with shared resources, such as water, fish-ing areas, power, railroads, or the environ-ment. Schiff and Winters (2002) make the casethat in the presence of economies of scale orinter-country externalities, market solutions toproblems are not necessarily the best, andregional cooperation can often pay largedividends.

When regional cooperation arrangementsare embedded in RTAs, it may be easier toconclude and implement these arrangements.Increasing trade raises the level of salience ofall aspects of regional cooperation and mayfoster greater high level attention to the re-gional arrangement and allow for more effec-tive and informal dispute resolution. More-over, agreements that cover more policydomains—for example, trade, transport,

power, and the like—allow countries to tradeoff gains in one area against losses in another,reducing or even eliminating the explicit com-pensatory schemes that would otherwise beneeded (Schiff and Winters 2002).

Consider some examples. The SouthernAfrican Development Community (SADC)provided the coordination point for regionalintegration in a regional power cooperationagreement. The Southern African Power Pool(SAPP), launched in 1995, was designed totake advantage of power resources in the re-gion and was the first formal internationalpower pool outside of North America. The12-country region has abundant hydropowerresources, especially the Inga Reservoir, largereserves of cheap coal in South Africa, and theKarriba Dam on the Zambia/Zimbabwe bor-der. The pool covers 6 million square milesand serves 200 million people. Utilities in theregion had been trading electricity for decadesthrough bilateral contracts, but these werecumbersome to administer. The objective forshifting to the pool was to create a more effi-cient regional market. The SAPP is modeledon the “loose” pools in Western Europe andthe United States, which emphasize constantexchange of information to maximize the costand reliability benefits from trading and sys-tem autonomy. Rather than relying on centraldispatch, loose pools rely on long-term bilat-eral contracts drawn up with common designsand security standards plus some central ser-vices. Unlike in the developed world, SAPPmembership is limited to national utilities.Each member must meet its AccreditedCapacity Obligation, a requirement that eachutility have capacity to cover its forecastmonthly peak. Each member is also obliged tocover emergency energy up to six hours, toprovide automatic generation control andother facilities in its control area, and to allowwheeling through its system. SAPP includesmost Southern African Development Commu-nity (SADC) members and is predicated onthe latter’s institutions, including the SADCTreaty, the SADC Dispute ResolutionTribunal, the SADC energy ministers, and the

Technical and Administrative Unit. The en-ergy ministers are responsible for resolvingmajor policy issues.

Though still in its early stages, the pool’s po-tential benefits include reducing or postponingnew requirements for generating capacity andreserves, lower fuel costs, and more efficientuse of hydroelectricity. A SADC electric powerstudy conducted in 1990–92 estimated a sav-ings of 20 percent ($785 million) in costs over1995–2010.

Three factors were critical to the develop-ment of the regional agreement: The availabil-ity of complementary power sources, an activeregional organization for economic coopera-tion, and the political will to support in-creased regional energy trade. SADC and itspredecessor, the Southern African Develop-ment Coordination Conference, served asfocal points for promoting regional integra-tion and facilitating investments in the neededinterconnection projects.

NAFTA offers another example. NAFTAhas also fostered regional cooperation for theenvironment by tying essentially extraneousenvironmental issues to the trade and invest-ment deal. This link helped to create the nec-essary political support for NAFTA in theUnited States, and it encouraged Mexico toaccelerate their environment program in orderto close the deal. The North AmericanAgreement on Environmental Cooperation(NAAEC) was signed as one of the side agree-ments appended to NAFTA at the last mo-ment. It created the Commission for Environ-mental Cooperation (CEC) in Montreal inearly 1994 to carry out the provisions of theagreement. The CEC has a young but growingconservation portfolio, focused mainly on pro-tecting habitats and species. A broad programof cooperation to protect North Americanbirds is in place, aimed at identifying impor-tant bird areas across the three member coun-tries and tying them into a protected network.A Biodiversity Information Network is undercreation, and strategies are being developedfor cooperation to protect marine and coastalecosystems. The CEC has also coordinated

measures to protect the monarch butterfly.Currently there is an active task force workingto stop the smuggling of endangered species.Under this program, U.S. Fish and Wildlifeofficers are training Mexican officers.

While the trade agreements underpinningthese regional cooperation initiatives were notessential to the actual activities, it is clear thatthey have provided useful political and institu-tional synergies.

Conclusions

One advantage of regional preferentialtrade arrangements is that they create

opportunities to lower trade costs in areasother than tariffs and non-tariff barriers totrade. This review of trade facilitation, stan-dards administration, and regional coopera-tion agreements points to several conclusions.

The potential to expand trade by loweringtrade costs other than policy border barriers isgreat—and it may have a higher payoff to co-operative governmental efforts than reciprocalreductions in border barriers. This is becausethe costs of institutional obstacles, informalbarriers, and sub-optimal regulatory scales areoften higher than the costs associated withpolicy border barriers. Further, many of thesebarriers do not generate revenues but simplywaste economic resources and directly con-strain productivity. These issues are alsoimportant elements defining the investmentclimate.

RTAs can precipitate cooperation to lower-ing trading costs in these areas because RTAsraise the level of policy salience, spread infor-mation about members and about interna-tional markets, improve the institutional effi-ciency of countries (better coordinationbetween the different institutions within acountry and between countries), provide “in-stitutional homes” for joint initiatives, andmay facilitate dispute resolution acrossmultiple areas.

Countries need not act in concert to reapthe benefits of unilateral reforms; the chancesof unilateral success are much improved,however, when policymakers are well


92

informed about international standards andthe trade-facilitation activities of other coun-tries. In the absence of such information, andof the capacity to act on it, it is unlikely that acountry, acting alone, will be able to matchthe benefits from participating in an RTA.

Finally, it seems clear that many RTAs arenot realizing their potential as a forum forreducing trade costs. North-South agreementsappear to have had somewhat greater success,perhaps because of the institutional interestsand strength of the more advanced partner.

Notes 1. Lawrence (1997).2. For example, McCallum (1995) reports results

suggesting that Canadian provinces are more than 20times more likely to trade among themselves than theyare to trade with U.S. states after controlling for the maineconomic determinants of trade. Subsequently, Nitsch(2000) found evidence of substantial border effects inEurope, with internal trade being, on average, larger bya factor of 10 than trade with EU partners, although themagnitude of this effect did decline during the 1980s.

3. See, for example, Hummels and Skiba (2002).4. GATT Article V mandates freedom of transit and

national treatment of products in transit. However, thisprovision has never been invoked. The WTO frame-work provides little leverage for poor, landlocked coun-tries to improve transit conditions.

5. Trade facilitation is of particular importanceto landlocked countries, whose products must passthrough numerous border crossings and checkpoints.Of the 50 least developed countries, 16 are landlocked:Afghanistan, Bhutan, Lao People’s Democratic Repub-lic, Nepal, Burkina Faso, Burundi, Central African Re-public, Chad, Ethiopia, Lesotho, Malawi, Mali, Niger,Rwanda, Uganda, and Zambia. Even coastal develop-ing countries may be effectively landlocked if they arenot on major shipping routes and are served by ineffi-cient and high cost coastal feeder services to mainports. Being landlocked has a significant and depress-ing effect on trade. For these countries, a regional ap-proach may be the only way to improve access toglobal markets, since there seems to be little scope atpresent for solving transit issues within the WTO. Cor-ridor solutions are efficient responses to the transportproblems of landlocked economies with deficient infra-structure. By definition they require bilateral or re-gional intervention.

6. RTAs with TTF approaches include theEuropean Union, MERCOSUR, Andean Community,SADC, COMESA, EAC, UMEOA, SAFTA, Eurasec,

and ASEAN. Two RTAs have no associated TTF:NAFTA and GCC. Three examples of TTFs existingindependently of RTAs are ECO, ECOWAS, and theNorthern Corridor Transit and Transport Agreement.

7. In treaties, competition is literally ensuredthrough “equal treatment of carriers” or non-discrimination regarding carrier’s nationality, knownas respect of the third party rule.

8. According to UMEOA Commission (2000),73 percent of the legal rules and customs governingtransport and transit regimes were derived from bilat-eral treaties (34 percent), national legislation (24 per-cent), and customs (15 percent); 27 percent were de-rived from regional treaties.

9. Collusion between transport operators leads toagreement on price setting. National associations playa role in determining which goods a company willtransport.

10. Standards can be mandatory as defined by gov-ernments (through technical regulations) so as to meettheir objectives regarding health, safety, and environ-mental issues; as well as voluntary, reflecting the de-mands and tastes of consumers or the technological re-quirements of industrial purchasers. In addition to thewriting of standards, an essential element of the systemof standardization is conformity assessment, the tech-nical procedures such as testing, verification, inspec-tion, and certification, which confirm that productsfulfill the requirements laid down in regulations andstandards.

11. From The Republic of Moldova Trade Diag-nostic Study, World Bank, 2004.

12. The principle of mutual recognition was devel-oped on the basis of European Court of Justice caselaw, specifically, the Cassis de Dijon and Dassonvillejudgements. In the former case, cassis from France wasprevented from being sold in Germany because it didnot contain enough alcohol!

13. Mutual recognition of regulations has alsobeen used within federal countries to remove barriersto inter-state trade. For example, Australia formallyadopted mutual recognition in 1993 to remove regula-tory barriers to the free flow of goods and labor be-tween Australian states and territories. As in the EU,some harmonized regulations are promulgated at thefederal level.

14. See Pelkmans and others 2000.15. Bilateral trade agreements with the EU tend

to contain support for developing capacity in thedeveloping country, whereas those with the UnitedStates provide little such support.

16. Chen and Mattoo (2004) show evidence thatMRAs promote trade, but that restrictive rules of ori-gin lead to trade diversion, especially against develop-ing countries.


93

17. Large companies in Brazil used their influenceto their advantage in setting a voluntary labeling stan-dard that requires water bottlers to include the resultsof numerous tests that are not understood by, or par-ticularly relevant to, the consumer, but that constitutean effective barrier to entry to small companies thatcannot afford the battery of tests required to providethe information. SEBRAI, a private institution thatprovides support services for small enterprises inBrazil, is working to improve access to certificationand metrology for its client firms. Through “solidaritycertification,” for example, SEBRAI helps groups ofsmall enterprises become certified at subsidized grouprates. SEBRAI also provides bonds that subsidize smallenterprises’ expenditures for metrology services.

18. This section draws heavily from Schiff andWinters (2002).

ReferencesAnderson, J., and E. van Wincoop. 2004. Trade Costs.

NBER Working Paper 10480, National Bureau ofEconomic Research, Cambridge, MA.

Arvis, J. F. 2004. Transit and the Special Case of Land-locked Countries. In Handbook of Customs Mod-ernization, eds. L. de Wulf and J. Sokol. Wash-ington, DC: World Bank.

Banomyong, R. 2000. Multimodal Transport in SouthEast Asia: A Case Study Approach. PhD thesis,University of Cardiff, www.bus.tu.ac.th/usr/ruth.

Baldwin, R. 1970. Non-Tariff Distortions of Interna-tional Trade. Washington, DC: BrookingsInstitution.

CEC (Commission of the European Communities).1998. Technical Barriers to Trade, Volume 1.Subseries III: Dismantling of Barriers of The Sin-gle Market Review. Luxembourg: CEC.

Cudmore, E., and J. Whalley. 2003. Border Delays andTrade Liberalization. NBER Working Paper9485, National Bureau of Economic Research,Cambridge, MA.

Carruthers, R., J. Bajpai, and D. Hummels. 2003.Trade and Logistics: an East Asian Perspective. InEast Asia Integrates: A Trade Policy Agenda forShared Growth, eds. K. Krumm, and H. Kharas.Washington, DC: World Bank.

Chen, M., and A. Mattoo. 2004. Regionalism in Stan-dards: Good or Bad for Trade? World Bank.Processed.

Daskalov, S., and D. Hadjinikolo. 2002. The Impact ofTechnical Barriers to Trade on Bulgaria’s Exportsto the EU and to the CEFTA Countries. In En-largement, Trade and Investment: the Impact of

Barriers to Trade in Europe, eds. P. Brenton, andS. Manzocchi. Cheltenham, UK: Edward Elgar.

Engel, C., and J. Rogers. 1996. How wide is the bor-der? American Economic Review 86: 1112–1125.

Evans, C. 2001. Home Bias in Trade: Location orForeign-ness? Staff Report 128, Federal ReserveBank of New York.

Faini, R. 2004. Trade Liberalization in a GlobalizingWorld. Paper presented at Annual World BankConference on Development Economics,Washington, DC, May 2004.

Funck, R. 1998. Integration of Transport Systems- theEuropean Experience. Report for the EU-SADCConference on Transport, Maputo, October.

Hoekman, B., and Konan, D. 1999. Deep Integration,Nondiscrimination, and Euro Mediterranean FreeTrade. Discussion Paper 2095, CEPR, London.

Hummels, D., and A. Skiba. 2002. A Virtuous Circle?Regional Tariff Liberalisation and ScaleEconomies in Transport. In Integrating theAmericas: FTAA and Beyond, eds. A. Estevade-ordal, D. Rodrik, A. M. Taylor, and A. Velaso.MA: Harvard University Press.

Hummels, D. L. 2000. Time as a Trade Barrier. GTAPWorking Paper 18, Center for Global TradeAnalysis, Department of Agricultural Economics,Purdue University.

Lawrence, R. 1997. Preferential Trading Agreements:The Traditional and the New. In Regional Part-ners in Global Markets: Limits and Possibilitiesof the Euro-Med Agreement, eds. A. Galal, andB. Hoekman. London: CEPR.

McCallum, J. 1995. National borders matter: Canada-U.S. Regional Trade Patterns. American Eco-nomic Review 85: 615–23.

Mentzoni, J. T. 2003. Trade Logistics—A Study ofLogistics Costs in Norwegian Wholesaler Indus-try. Processed. www.transportbrukerne.no.

Nofal, Beatriz. 2004. Constructing a deeper integra-tion in MERCOSUR. Background paper for theGlobal Economic Prospects 2005. Washington,DC: World Bank.

Nitsch, V. 2000. National Borders and InternationalTrade: Evidence from the European Union. Cana-dian Journal of Economics 33: 1091–105.

Pardo, M. 2001. Pasos Fronterizos en la ComunidadAndina. IADB. Processed.

Pelkmans, J. 1987. The New Approach to TechnicalHarmonisation and Standardisation. Journal ofCommon Market Studies 25.

Pelkmans, J., E. Vos, and L. Di Mauro. 2000. Reform-ing prodict regulation in the EU; a PainstakingTwo-level Game. In Regulatory Reform and Com-petitiveness in Europe, Volume I, eds. G. Galli andJ. Pelkmans. Cheltenham, UK: Edward Elgar.


94

Röschlau, F. 2003. Double Standards Cause RegionalTransport Nightmares, e-corridor, Walvis BayCorridor Group, number 3.

Ruijgrok, C. 2001. European Transport: Insights andChallenges. In Handbook of Logistics and Supply-Chain Management, eds. A. Brewer, K. Button, andD. Hensher, chapter 3. New York: Pergamon Press.

Schiff, M., and A. Winters. 2002. Regional Coopera-tion, and the Role of International Organizationsand Regional Integration. Policy Research Work-ing Paper 2872, World Bank, Washington, DC.

Trefler, D. 1995. The Case of the Missing Trade andOther Mysteries. American Economic Review 85:1029–46.

UEMOA (West African Economic and MonetaryUnion) Commission. 2000. Projet de Rapport deSynthèse Préparatoire à la Table Ronde des

Bailleurs de Fonds sur les Infrastructures et leTransport Routier des Etats Membres del’UMEOA. Ouagadougou.

UNESCAP (United Nations Economic and Social Com-mission for Asia and the Pacific). 2003. TransitTransport Issues in Landlocked and Transit De-veloping Countries. New York, ST/ESCAP/2270.

Wilson, J., C. Mann, and T. Otsuki. 2003. TradeFacilitation and Economic Development; Measur-ing the Impact. Policy Research Paper 2988,World Bank, Washington, DC.

World Bank. 2003. Trade, Investment, and Develop-ment in the Middle East and North Africa:Engaging with the World. Washington, DC:World Bank.

World Bank. 2004. The Republic of Moldova TradeDiagnostic Study. World Bank, Washington, DC.


95

As barriers to merchandise trade have comedown and trade has expanded, policymakersand trade negotiators have turned their atten-tion to services and trade-related regulatoryissues. Of these, services, investment, intellec-tual property, and temporary movement oflabor arguably have the greatest potential foraffecting incomes and trade in developingcountries. Agreements on these four issues arenow becoming common in bilateral and somepreferential regional trade agreements (RTAs).

North-South agreements, notably thebilateral free trade agreements of the UnitedStates and of the European Union (EU), havebeen the important drivers for services, in-vestment, and intellectual property rights(IPRs). In broad terms, the United States, forexample, offers access to its large market forgoods in exchange for access to services mar-kets in developing countries and their accep-tance of rules governing investment andintellectual property rights. The EU marketaccess agreements also cover many of thesetopics, if less specifically. Labor services—that is, the temporary movement ofworkers—are largely confined to profes-sional and skilled workers, often intra-corporate transfers. South-South agreementstend to feature services liberalization lessprominently, and their rules governinginvestment, intellectual property, and eventhe temporary movement of workers, arecommonly weak or absent altogether.

From a development perspective, the mostpotentially beneficial components of this set ofissues are provisions that open services mar-kets to additional potential suppliers throughforeign subsidiaries (in GATS terminology,Mode 3) and the temporary movement ofworkers (Mode 4).1 Services liberalization inpreferential arrangements can enlarge thenumber of competitors and carries fewer risksof income losses than preferential merchan-dise trade because lifting most common re-strictions does not cost the government rev-enue. Though multilateral liberalization isusually preferable even in services,2 RTAs inservices can be predicted, in general, to in-crease welfare. Similarly, preferential agree-ments that widen the scope for the temporarymovement of workers have the potential toraise incomes.

In both services and labor mobility, how-ever, agreements have yet to fulfill their devel-opment potential. Many of the North-Southagreements are between countries with unusu-ally open service sectors, so the additionality tothe various parties is limited to a handful ofrelatively small sectors and to the credibilityeffects of locking in openness via treaties and“seals of approval” that investors might take asa sign of lower risk. Meanwhile, in many of theSouth-South agreements, where the potentialscope for liberalizing measures is often fargreater, RTA-driven additional liberalizationhas been sporadic. For labor services, both the

97

Beyond Merchandise Trade:Services, Investment, IntellectualProperty, and Labor Mobility

5

North-South and South-South agreements areconfined to intra-firm movement of profession-als, and neither agreement has substantiallywidened market access for the temporarymovement of labor.

By contrast, North-South agreements re-garding investment and IPRs have succeededin promulgating comprehensive new rules thatgo beyond multilateral rules in the agreementon Trade Related Intellectual Property Rights(TRIPs). The United States and EU bilateralagreements have enhanced market accessthrough negative-list and positive-list (respec-tively) pre-establishment rights, and theUnited States has implemented investor-statedispute settlement mechanisms that empowerforeign investors to seek arbitration awards incases of uncompensated expropriation orother violations of treaties.

Ironically, of the four areas, investment andIPRs are the two where the development po-tential is largely unproven. The investmentprovisions that enhance investor’s rights havenot been shown to increase the flow of invest-ment to developing countries. Nor havestronger IPRs embedded in the TRIPS-Plusagreements been shown to accelerate techno-logical flows to low-income countries—though it may do so for middle-income coun-tries. On the other hand, because free tradeareas that result in larger markets do attractadditional investment flows, it may be that incombination with large, preferential tradeareas, enhanced investor protections and IPRsdo have a positive impact—but agnosticismseems warranted.

This chapter begins with a synoptic com-parison of agreements and a focus on theregulation-intensive bilateral U.S. and EUfree trade agreements (FTAs). Understandingthe diversity and reach of these agreementspermits us, in a subsequent section, toreview the economic consequences of provi-sions that deal with services, investment,and intellectual property. A final sectionexamines the treatment of movement oftemporary labor.

Services, Investment, and IPRsin Regional Agreements

North-South agreements differ sharply intheir coverage of services, investment,

and intellectual property. At one end of thespectrum, U.S. FTAs usually involve the mostexplicit negotiations for market access in ser-vices and U.S.-style rules for investment andintellectual property. The EU market accessagreements similarly contain market accessprovision in services, but tend to reinforce pre-vailing international rules for intellectual prop-erty; its Economic Partnership Agreements inAfrica use development assistance in combina-tion with trade preferences to promote rulesbeyond international agreements, includingEU-style concerns for competition policy andgeographical indications. At the other end ofthe spectrum, most South-South agreementsare focused primarily on merchandise trade,and tend to treat services, investment, andIPRs unevenly, if at all. These distinctionsshould become clearer when we consider theU.S., EU, and South-South approaches in turn.

U.S. FTAs are rule intensiveKey features of the U.S. FTAs that cover ser-vices, investment, and intellectual propertyrights include:

• Opening services markets to competitionfrom foreign suppliers or locking inprior autonomous liberalization, exceptin those sectors excluded (i.e., on a nega-tive list). Because most of the countrieswith which the United States has con-cluded bilateral FTAs are already open inmost sectors, the agreements generallylock in prevailing openness and affectchanges in only a few still-restrictedactivities. Significant market openingstook place in the Costa Rican telecom-munications and insurance sectors andless dramatic market openings occurredin the banking sector in Bahrain. Provi-sions range from inclusion of insurance,


98

99

Tab

le 5

.1S

ervi

ces,

inve

stm

ent,

and

inte

llect

ual

pro

per

ty:

A c

om

par

iso

n

Inte

llect

ual

Serv

ices

Inve

stm

ent

Prop

erty

Pre-

esta

blis

hmen

tR

ight

to

Nat

iona

lN

atio

nal a

nd&

Lim

itat

ions

Prov

ide

Tre

atm

ent/

MF

Pre-

Ban

on

Inve

stor

-Sta

teM

FN/T

reat

men

tR

ule

of O

rigi

nM

arke

t A

cces

sSe

rvic

es w

/oR

atch

etPo

st-

Ow

ners

hip

esta

blis

hmen

tPe

rfor

man

ceD

ispu

teIn

telle

ctua

lA

gree

men

tsM

arke

t A

cces

sa(N

onre

stri

ctiv

e)b

Exc

epti

ons

esta

blis

hmen

tdM

echa

nism

ees

tabl

ishm

ent

Lim

itat

ions

fL

imit

atio

nsR

equi

rem

ents

Sett

lem

ent

Prop

erty

U.S

.U

.S.-

Jord

anY

esY

esN

egat

ive-

list

No

No

Yes

Neg

ativ

e-lis

tN

egat

ive-

list

TR

IMS+

Yes

Yes

h

U.S

.-C

hile

Yes

Yes

Neg

ativ

e-lis

tY

esY

esY

esN

egat

ive-

list

Neg

ativ

e-lis

tT

RIM

S+Y

esT

RIP

S+U

.S.-

Sing

apor

eY

esY

esN

egat

ive-

list

Yes

Yes

Yes

Neg

ativ

e-lis

tN

egat

ive-

list

TR

IMS+

Yes

TR

IPS+

U.S

.-A

ustr

alia

Yes

Yes

Neg

ativ

e-lis

tY

esY

esY

esN

egat

ive-

list

Neg

ativ

e-lis

tT

RIM

S+N

oT

RIP

S+U

.S.-

CA

FTA

Yes

Yes

Neg

ativ

e-lis

tY

esY

esY

esN

egat

ive-

list

Neg

ativ

e-lis

tT

RIM

S+Y

esT

RIP

S+U

.S.-

Mor

occo

Yes

Yes

Neg

ativ

e-lis

tY

esY

esY

esN

egat

ive-

list

Neg

ativ

e-lis

tT

RIM

S+Y

esT

RIP

S+N

AFT

AY

esY

esN

egat

ive-

list

Yes

Yes

Yes

Neg

ativ

e-lis

tN

egat

ive-

list

TR

IMS+

Yes

TR

IPS+

EU E

U-S

outh

Afr

ica

No

No

No

No

No

No

No

No

No

Non

Yes

i

EU

-Mex

ico

Yes

Yes

Stan

dsti

llcN

oN

oN

oN

oN

oN

oN

onY

esi

EU

-Chi

leY

esY

esPo

siti

ve-l

ist

No

No

Yes

No

Posi

tive

-lis

tN

oN

onY

esi

Sout

h-So

uth

ME

RC

OSU

RY

esY

esPo

siti

ve-l

ist

No

No

Yes

No

Neg

ativ

e-lis

tT

RIM

S+Y

esN

oj

And

ean

Com

mun

ity

No

Yes

Posi

tive

-lis

tN

oN

o–

No

Posi

tive

-lis

tT

RIM

S+N

oN

ok

CA

RIC

OM

Not

spe

cifi

edY

esN

egat

ive-

list

No

No

No

No

Posi

tive

-lis

tN

oY

esN

oA

SEA

NY

esY

esPo

siti

ve-l

ist

No

No

Yes

Yes

Posi

tive

-lis

tN

oN

oN

o1

SAD

CN

oN

oN

oN

oN

one

No

No

TR

IPS

CO

ME

SAY

esN

oPo

siti

ve-l

ist

Nog

No

Posi

tive

-lis

tN

oN

oN

om

Oth

erJa

pan-

Sing

apor

eN

oY

esPo

siti

ve-l

ist

No

No

Yes

Yes

Can

ada-

Chi

leY

esY

esN

egat

ive-

list

Yes

Yes

Yes

Chi

le-M

exic

oY

esY

esN

egat

ive-

list

Yes

Yes

No

a. I

nclu

des

fair

and

equ

itab

le t

reat

men

t.b.

Den

ial b

enef

its

only

to

juri

dica

l per

son

that

do

not

cond

uct

“sub

stan

tial

bus

ines

s” in

one

of

the

mem

ber

coun

trie

s.c.

Pro

vide

s fo

r fu

ture

neg

otia

tion

of

com

mit

men

ts à

la G

AT

S.d.

Rig

ht o

f no

n-es

tabl

ishm

ent,

tha

t is

no

esta

blis

hmen

t re

quir

ed t

o su

pply

a s

ervi

ce.

e. A

uton

omou

s lib

eral

izat

ion

is a

utom

atic

ally

inco

rpor

ated

into

the

agr

eem

ent.

f. L

imit

s on

equ

ity

shar

ehol

ding

s fo

r co

mpa

nies

in s

ecto

rs o

ther

tha

n th

ose

excl

uded

fro

m p

re-e

stab

lishm

ent

limit

atio

ns.

g. C

OM

ESA

doe

s gr

ant

fair

and

equ

itab

le t

reat

men

t to

mem

bers

, but

not

to

non-

mem

bers

.h.

The

IP

prov

isio

ns a

re c

onsi

dere

d T

RIP

s Pl

us. H

owev

er t

he c

hapt

er c

over

age

is le

ss s

peci

fic

and

com

preh

ensi

ve t

han

othe

r su

bseq

uent

U.S

. fre

e tr

ade

agre

emen

ts.

i. R

equi

res

only

adh

eren

ce t

o in

tern

atio

nal c

onve

ntio

ns.

j. T

he M

ER

CO

SUR

agr

eem

ent

does

not

incl

ude

IP, b

ut p

rovi

des

for

inte

rpar

liam

enta

ry c

omm

itte

es t

o be

gin

wor

k on

har

mon

izat

ion

of I

P la

ws.

k. A

ndea

n co

mm

unit

y re

gula

tes

all p

aten

ts.

l. A

SEA

Nha

s a

fram

ewor

k ag

reem

ent.

m. A

ct 1

28(e

) ca

lls f

or a

dopt

ion

of n

ew p

aten

t la

ws.

n. E

U b

ilate

ral i

nves

tmen

t tr

eati

es p

rovi

de f

or in

vest

or-s

tate

dis

pute

res

olut

ion.

Sour

ces:

Leg

al t

reat

ies;

Mat

too

and

Sauv

e 20

04; t

e V

elde

and

Fah

nbul

leh

2003

; Man

n an

d C

osbe

y 20

04; S

zepe

si 2

004a

, 200

4b; A

bbot

t 20

04a

and

2004

b; O

EC

D 2

003;

info

rmat

ion

prov

ided

by

gove

rnm

ents

.

financial advisory services, and selectedtelecommunications services to arguablyrelatively minor changes to the alreadyopen regimes, such as the commitment ofSingapore to cease cross-subsidies in ex-press mail delivery or the commitment ofChile to open selected insurance services(table 5.2).

• Ratchet provisions and negative-list ex-clusions. The ratchet clauses mean thatnew autonomous liberalization will au-tomatically be subsumed under the termsof the agreements. Negative lists ensurethat yet-to-be-invented new service areasare guaranteed to be covered by thetreaty. Notable for their absence is theexclusion of labor services, except provi-sional visas for professionals associatedwith investing firms (discussed in thepenultimate section of this chapter).

• Investment rights. Investment rights, withprovisions for national treatment,nondiscrimination, pre-establishmentprovisions for companies based in eachothers markets, bans on trade-relatedinvestment measures (TRIM), andinvestor-state arbitration of dispute lim-ited only by a negative list of exclusions.

• TRIPs-Plus provisions that providestronger protections for IPRs than underthe TRIPS agreement, with investor-statearbitration dispute settlement permittedin the event of disputes (subject tocertain limitations).

Other noteworthy provisions (not the sub-ject of this chapter) include labor protectionsand environment issues that figured promi-nently in the CAFTA, Chile, and Singaporeagreements, among others. Signatory coun-tries committed to enforcing their own laborlaws in five areas: right of association, theright to organize and bargain collectively, pro-hibitions on forced labor, a minimum age foremployment of children, and acceptableworking conditions. Complaints can be filed,and if the agreed-on procedures to mediate thedispute fail, a panel of experts would reviewthe case and, if warranted, impose a fine to beused for the enforcement of labor rights; thatis to say, trade sanctions are not an agreed-onremedy (Weintraub 2004).

The FTAs involve innovations in trade lawin two important areas: investment and IPRs:

Investment Access and Protections. TheFTAs have incorporated the provisions ofbilateral investment treaties (BITs), and insome cases, provided new measures cover-ing investment (table 5.1). Agreements, espe-cially post-NAFTA ones, include broaddefinitions of investment, comprising notonly foreign direct investment (FDI), but alsoportfolio flows, private debt, and even sover-eign debt issues as well as intellectual prop-erty (Mann and Cosbey 2004; Vivas-Eugui2003). The inclusion of short-term debt, to-gether with pre-establishment rights, led theU.S. Treasury to demand that Chile modifyits controls on capital inflows that were


100

Table 5.2 Additional services liberalization in U.S. FTAs

Chile Australia Bahrain CAFTA Morocco Singapore

Banking � � �

Insurance � � � �

Telecommunication � � � � �

Broadcasting & Audiovisual � �

Financial Advisory Services and Data �

Retail/Wholesale Distribution �

Restrictions on Foreign Directors & Managers � �

Express Mail Delivery �

Real Estate �

Legal Services �

Source: Legal treaties.

designed to curtail destabilizing hot moneyinflows.3 Such broad definitions exposecountries to dispute settlements across a rangeof assets that go far beyond multilateralcommitments.

All agreements provide for treatment offoreign investors on the same basis as domes-tic investors (national treatment) and haveprovisos banning discrimination among in-vestors from member countries (MFN, ornondiscriminatory treatment). For many ofthe initial FTA countries, these stipulationshad been included in national legislationsand/or had been incorporated into bilateralinvestment treaties, mainly on a post-establishment basis.

What is new is the extension of the pre-establishment right to invest in businesses andactivities in all sectors, except where expresslyprohibited via a negative list.4 These pre-establishment rights lock in the right ofMexican and Canadian investors underNAFTA to invest in all activities in the UnitedStates. Exceptions for the United States in-clude foreign investment with NAFTA guaran-tees in selected areas of communication,media, transportation, and social services.Pre-establishment rights mark a broad expan-sion of market access by foreclosing futuregovernment policies that would raise barriersto foreign investment. The rationale for ac-cepting such disciplines is that it provides cer-tainty on the rules of the game, whichwill in turn translate into increased investmentinflows.

Another discipline more expansive thanmultilateral accords is in trade-related invest-ment measures (TRIMs). The WTO TRIMsagreement of 1995 attempted to clarify disci-plines on government policies that requireforeign companies to establish joint ventures,export in a certain portion of its sales or bal-ance trade, use local inputs to achieve value-added objectives, or hire local staff. However,the agreement failed to provide adequate def-initions of disciplines, and it presented poorlyformulated implementation periods andinadequate notification and monitoring

procedures; the operation of the agreementwas to be reviewed by January 1, 2000, butso far the review has not occurred (Bora2003). All of the bilateral FTAs ban, in someform, trade-related investment requirements,such as by local content rules, value-added re-quirements, and restrictions on management.The U.S. bilateral agreements have, in effect,established a “TRIMs-Plus” set of obligationsthat includes outright bans on certain perfor-mance requirements, including exports, mini-mum domestic content, domestic sourcing,trade balancing, and technology transfer.In general, government procurement,environmental standards, some health mea-sures, and requirements for local research anddevelopment (R&D) are all exempt (Te Veldeand Fahnbulleh 2003).

Freedom to make transfers is a nontrivialinvestment right granted under the investmentagreements. This assures investors thatthey will be able to transfer profits, makeinvestments, or lend without governmentinterference.

Finally, all U.S. agreements except theAustralian FTA create an investor-statedispute resolution provision that permits in-vestors to take foreign governments to disputeresolution for violation of the treaty’s nationaltreatment, nondiscrimination, or expropria-tion provisions, among others. NAFTA’sChapter 11 and Chile’s Chapter 10 are themost widely known mechanisms, but thesemechanisms are contained in the other bilat-eral agreements as well.

Intellectual Property Rights. The IPR pro-visions embedded in all recent U.S. FTAs gobeyond the multilateral IPR standards estab-lished in the WTO’s TRIPS Agreement.“TRIPS-Plus” elements found in many—butnot all—of the IPR chapters include:5

• Extension of the patent term for delayscaused by regulatory approval processes;extension of the term of copyright pro-tection to life of author plus 70 years(compared to life of author plus 50 yearsin TRIPS).

B E Y O N D M E R C H A N D I S E T R A D E

101

• A requirement to provide patent protec-tion for plants and animals.

• A limitation on the use of compulsorylicenses for national emergencies and asantitrust remedies, and for public non-commercial use.

In the area of pharmaceuticals:

• An obligation to prohibit the marketingapproval of generic drugs during theterm of the drug patent.

• A five-year period of marketing exclusiv-ity following the submission of safetyand efficacy data to drug regulatoryauthorities (so-called data exclusivity).In addition, marketing exclusivity effec-tively applies across borders, so that mar-keting approval in one market—say, theUnited States—impedes registration ofcompeting products in another market.

• An additional three-year period of mar-keting exclusivity based on the submis-sion of new clinical data with respect tonew uses of previously approved drugs.Exclusivity would also apply to drugs forwhich the patents have expired (althoughgeneric competition for previously ap-proved uses would remain unaffected).

• Imposition of restraints on parallel im-portation, impeding the possibility thatparties to the agreements open theirmarkets to the import of products thathave already been sold—possibly morecheaply—in foreign markets.

In the area of digital works:

• An obligation against circumventingso-called technological protectionmeasures—devices and software devel-oped to prevent unauthorized copying ofdigital content. Rules on the liability ofInternet service providers (ISPs) whencopyright infringing content is distrib-uted through their servers and networks.These provisions are based on standardsfound in the U.S. Digital MillenniumCopyright Act of 1998.

The inclusion of these services, investment,and IPR issues was a contributing factor to thebreakdown in negotiations in the FreeTrade Area of the Americas (Nogues 2004).We return to these issues below when consid-ering the economic consequences of thesearrangements.

EU FTAs take a different approach In addition to market access in merchandise,the EU has focused heavily on services in itsbilateral FTAs. The earliest (and least specific)is the South African agreement (1999) thatcontains only the promise of potential liberal-ization after discussions transpire in 2004 and2005. In the EU-Mexico FTA, several generalprovisions were included (many ratifyingGATS arrangements), as well as specific liber-alization commitments in the financial sector.The EU-Chile agreement went further than theother two and included liberalization oftelecommunications and maritime services(Ullrich 2004).

The EU agreements with Mexico and Chilediffer from the U.S. agreements in importantrespects. First, the trade provisions arephrased on the basis of a positive list andimplicitly exclude new products. Second, thetreatment of intellectual property effectivelyreaffirms a multilateral approach to IPRs,because the agreements provide only the listof conventions that signatory countries havealready ratified, those it intends to ratify, andthose that it will consider ratifying in thefuture.6 This approach differs from that takentoward the EU-accession countries, in whichnew entrants were required to apply the rigor-ous EU standards on data protection and mar-keting exclusivity; these have a major impacton generic producers.

The treatment of investment and capitalflows in both agreements does not appear tobe extensive. For example, the EU-Mexicoagreement simply states that the existing re-strictions on investment will be progressivelyeliminated and no new restrictions adopted;the agreement did not specify particular sec-tors or set a timeline for liberalization. The


102

language in the EU-Chile agreement calls forthe “free movement of capital relating to di-rect investments made in accordance with thelaws of the host country.” In both instances,the agreements allow for the use of safeguardsin the event of monetary or exchange rate dif-ficulties, and although the time limit is setat 6 months for Mexico and 12 months forChile, it would allow for continuation of thesafeguard after the time limit through its for-mal reintroduction. That said, many of thesame investor protections found in U.S. FTAsare also found in EU bilateral investmenttreaties with developing countries.

The treatment of dispute settlement is sim-ilar in both agreements. In general, the EU hasno special provisions pertaining to investment,but these are covered under the general dis-pute settlement provisions for all matters inthe agreements (Szepesi 2004a and b). Disputesettlement is covered on a state-to-state leveland is first attempted through consultationswith a Joint Committee (Association Com-mittee in the case of Chile) within 30 days ofa party’s request. If this step of “dispute avoid-ance” proves unsuccessful, the concernedparty can forward its request to an arbitrationpanel comprised of representatives of bothparties. The arbitration panel’s decisions arebinding, and the panel can also rule on theconformity of any measures undertaken as aresult of its decision with the original ruling.Both agreements provide extensive detail onthe process of appointing members to thearbitration panel, timelines for the panel’sruling, and compliance with the panel’sdecisions.

South-South agreements focus on expanding tradeVirtually all of the other major agreementscontain references to services liberalization.Most agreements allow for national treat-ment, post-establishment nondiscriminatoryprovisions (table 5.1). At the other extreme,there are more limited agreements like Associ-ation of Southeast Asian Nations (ASEAN)and Southern Common Market Agreement

(MERCOSUR) that have delivered servicesliberalization additional to levels negotiatedmultilaterally or determined unilaterally. Re-gional agreements in services have competedto create complex structures of rules and com-mitments. But in many cases, the sound andfury of the negotiations has signified limitedliberalization. Many agreements do not pro-vide new market access beyond what coun-tries have already scheduled with the GATS.In telecommunications and financial services,the GATS has in fact achieved a higher level ofbound liberalization than that offered in mostRTAs.

All South-South agreements have a rela-tively nonrestrictive definition of preferentialaccess; by allowing firms from nonmembercountries that have “substantial business” inmember countries to invest through sub-sidiaries based in member countries, the num-ber of potential competitions in the market isenlarged.

Agreements with negative lists have severaladvantages in terms of market access: theypermit automatic liberalization of new serviceindustries; they establish a stronger floor forliberalization by locking in the status quo;they are more transparent; and they may leadto a more productive internal dialogue withsectoral private interests (Mattoo and Sauve2004). Ratchet mechanisms that allow newautonomous liberalization to be incorporatedautomatically into treaties are most likely toco-exist with negative list provisions.However, most of the South-South agreementshave not liberalized many sectors, and some,like MERCOSUR, have not implementedaccords in the way that was anticipated atsigning (Nofal 2004).

Investment provisions have differed aswell. South-South RTAs generally have beenless ambitious with respect to investor protec-tions. This is true for the right to provide ser-vices without establishing local affiliates. It isalso true for investor-state dispute settlement.For the most part, only the United States andEU bilaterals have established sophisticatedmechanisms to deal with disputes on


103

investment. Some agreements provide forinvestor-state dispute resolution, though theseprotections are less strong than in the North-South Agreements.

Intellectual property rights, while men-tioned in South-South agreements, rarely gobeyond disciplines negotiated at the multilat-eral level, and they do not have the tightly for-mulated provisions that characterize theNorth-South agreements, notably those withthe United States. MERCOSUR, for example,has agreed to establish a commission to exam-ine areas of intellectual property harmoniza-tion, while ASEAN has a framework agree-ment. The Andean community has moredetailed restrictions, but these are written lesswith the view of protecting intellectual prop-erty than of eliminating territorial restrictionson the use of patented technology; the res-trictions were designed to end the restraintsmultinational companies put in their technol-ogy contracts with their foreign affiliates,which explicitly prevented them from usingthe patents in export production.

Economic Consequences ofServices, Investment, and IPRProvisions in RTAsNew market access in services couldpromote growthServices liberalization with proper regulationcan be a powerful driver of economic growthand poverty reduction. At the sectoral level,removing barriers to competition can lowerprices, improve quality, and add variety. Be-cause of the linkage effects—the fact that pro-ducers require telephones, use finance, needadequate transportation services, and benefitfrom business services—improving service sec-tor performance can generate huge economicgains. Mattoo, and others (2001) show thatcountries with fully liberalized financial andtelecommunications sectors grew annually onaverage about 1.5 percentage points fasterthan other countries, controlling for other fac-tors. These gains are not automatic—theyrequire adequate regulation and a supportive

investment climate—but the potential gainsare large (World Bank 2001).

Realizing these gains requires allowing for-eign investors greater market access, and thisis the most important provision in a preferen-tial arrangement. Countries can open previ-ously closed sectors to RTA partners as part ofan agreement. Since today most countries ac-cept, indeed clamor for, foreign investment inmanufacturing and natural resources, RTA-driven reductions in entry barriers affectmainly services. Moreover, services now play alarger role in investment flows, and for somecountries, such as Mexico, they have dwarfedinvestments in manufacturing. The great bulkof services investment are market-seeking,horizontal investments. These cover a vastrange of large multinationals: Deutsche Bank,WalMart, Starbucks, Microsoft, and so on.These “mode 3” services require the commer-cial presence of affiliates, branches, or fran-chises to deliver the service. To be sure, somecountries (such as India) have experiencedsubstantial flows associated with call centersand data processing, and this new investmentaccompanies these cross-border supply(“mode 1”) activities, though these activitiesremain small in comparison to trade throughcommercial presence.

Because preferential arrangements permitmore suppliers to compete in the market, acountry is almost certain to gain from prefer-ential liberalization of the services trade, irre-spective of the supplier. This is in sharp con-trast to merchandise trade, where the incomeloss associated with trade diversion can occurwith the loss in tariff revenue. In services, bar-riers to entry usually take nonmonetary formssuch as regulatory restrictions on entry, for-eign equity limitations, quotas on outputs andforeign service workers, and requirements onlegal form of establishment. None of thesegenerate revenue for the government, so re-moving these restrictions is less likely to pro-duce income losses (with merchandise trade,income losses associated with trade diversionoccur because the government loses the tariffrevenue as trade is diverted to higher-cost


104

sources of imports). Moreover, the scope forincreased competition and exploitation ofscale economies, as well as the possibility ofinducing knowledge spillovers, strengthens thepresumption that a country would gain from apreferential agreement in services.

Multilateral, nonpreferential liberalizationis likely to produce even larger gains thanpreferential regional agreements. This is be-cause multilateral liberalization opens themarket to the largest number of competitorsand permits consumers maximum choice; itallows imports from the most competitivesource. It also leads to a less complex policyregime than a preferential arrangement, andtherefore implies lower administration costsfor government agencies and lower transac-tion costs for the private sector. Finally, it ispossible that preferences could lead to ahigher-cost firm gaining a competitive advan-tage relative to investors from outside the re-gion. And first mover advantages and barriersto entry can make it difficult for lower-costsuppliers from third countries to enter themarket. Inefficient suppliers from membercountries might establish positions behindmarket barriers sufficiently high that new andeven more efficient potential competitorswould not choose to pay the cost of entry.

Rules of origin for services, as with mer-chandise trade, can play a significant role in de-termining the degree to which regional tradingarrangements discriminate against nonmembercountries, and hence the degree of competitionin services associated with an RTA. For exam-ple, if one participant has a fully liberalizedmarket, the adoption of a nonrestrictive rule oforigin by the other participants can be likenedto MFN liberalization. Service suppliers canenter the liberal jurisdiction and from theremove to the other partner countries. Many gov-ernments take the liberal rules of origin one stepfurther and extend regional preferences on anMFN basis under the GATS. This widens thenumber of competitors in the market and offersgreater opportunities for securing access to themost efficient suppliers—particularly of infrast-ructural services likely to exert significant

effects on economy-wide performance. Becauseof the strong potential links to growth (WorldBank 2002), the additional market access pro-vided through RTAs could be important. Un-fortunately, the actual additional liberalizationhas not yet matched this promise.

Nonetheless, restrictive rules of origin canlimit the potential benefit to liberalization.Participants who seek to benefit from prefer-ential access to a protected market and denybenefits to third country competitors are likelyto argue for the adoption of restrictive rules oforigin, based on criteria such as ownership orcontrol considerations. This could be the atti-tude of regionally dominant but globally non-competitive service providers toward third-country competition within a regionally inte-grating area.

Examples of restrictive rules of origin forservices and investment can be found inMERCOSUR and the Andean Pact, both ofwhich limit benefits to juridical persons thatare owned and controlled by natural personsof a member country. The Hong Kong-ChinaFree Trade Agreement, for example, features adetailed annex spelling out the set of criteriaby which Hong Kong service suppliers maybenefit from the terms of the agreement.

Do RTAs attract more investment?RTAs can, in theory, promote more invest-ment through new trade rules that create alarger market, new investment rules thatpermit market access by relaxing restrictionson market entry (such as discussed above forservices), and new investor protections.

New trade rules that eliminate internal bar-riers create a larger internal market, which canraise the return to investment and create anincentive to invest for members and for thirdcountries. Firms investing in the RTA coun-tries can achieve economies of scale and scopein serving a larger market of potential buyers,may experience reductions in transactionscosts, and if services are included, benefit frommore efficient financial, telecommunications,and other services (Schiff and Winters 2003).Trade rules can induce greater efficiency in


105

transactions with the global economy.Markusen (2004), for example, notes thatinward investment to reach the local marketmay also include tapping into lower cost pro-duction sites within the new RTA to serve thewealthier parts. Japanese multinational com-panies might well locate in Mexico to reachthe U.S. market, though the net effects mightbe diluted as U.S. firms also set up in Mexico.Frischtak (2004) found that MNCs in autos,textiles, and electronics reallocated their pro-duction to Mexico to serve the U.S. market.

The larger market can also increase pro-ductivity in other ways. Aside from economiesof scale, a larger market can increase competi-tion among a potentially larger set of suppli-ers, and take advantage of differing regionalfactor prices to drive productivity increasesand hence more rapid growth; and the morerapid growth provides a dynamic attractionto intra-bloc and extra-bloc investment. If theRTA reduces border protection on investmentgoods and allows domestic producers to sourcecheaper and higher technology capital goods,members may benefit.

However, efficient results are not automatic.Even though investment may be destined for alarger market, border barriers may create incen-tives to invest in high-cost import-substituting

activities that are not internationallycompetitive. Latin America’s early experimentswith admitting FDI behind high border barri-ers produced inefficient investment and a pro-tectionist political economy that took decadesto unwind (box 5.1). The formation of RTAsmay increase both internal and external in-vestment, but the resulting market size maynot be sufficient to realize modern scales, andthe high external tariffs drive up the costsof imported inputs. Indeed, the first AndeanPact in the early 1970s and the Central Amer-ican Common Market in the 1960s failedto generate investment-related productivitygains.

RTAs may include new investment rules tofacilitate market access. As with services, thedecision to lift an administrative barrier im-peding manufacturing investment or an in-vestment in natural resources can create anopportunity for investors and hence promptnew investment. Most remaining restrictionstoday—equity ownerships limitation and banson foreign investment in particular activities—are not restrictions on manufacturing, butrather on services (e.g., broadcasting,telephony, and airlines in the United States,among other countries) and natural resources(e.g., oil in Mexico). RTAs that reduce market


106

Trade barriers in preferential arrangements candivert trade to higher-cost sources. No less im-

portant is the investment undertaken to produce thenew trade. If trade policies provide high protection,they will limit competition, allow for shared monop-oly pricing, and incur the inefficiencies of price dis-tortions. This pattern of investment was common—and costly—in the period before high protection wasbrought down. Lall and Streeten (1977) who studiedsome 90 foreign investments using a cost-benefitmethodology, found that more than 33 percentreduced national income; this was mainly from ex-cessive tariff protection that allowed high cost firmsto produce for the local market at very high

Box 5.1 Not all investment is good investmentprices, even though they could have imported muchmore cheaply. (It turns out domestic firms performedeven more poorly.) Encarnation and Wells (1986)found that between 25–45 percent of 50 projectsstudied (depending on analytical assumptions) re-duced national income; again the main culprit washigh protection. As average tariff levels have comedown, these low-quality type of investments havefaded in importance. Trade and tax policy often in-teract in ways that magnify their competition-re-stricting effects.

Sources: Lall and Streeten 1997; Encarnation and Wells 1986;Newfarmer 2001.

access barriers associated with restrictions onforeign entry are likely to have greater im-pact through this channel than throughmanufacturing.

Granting new investment rights may alsoattract additional investment, though here thecase is more contentious. In general, thestrongest investor protections entail nondis-crimination among all investors, provisionsagainst expropriation, dispute settlement witheligibility for investor-state suits, and indepen-dent arbitration. The legal power granted toinvestors to sue governments under terms ofthe bilateral or regional agreements is ar-guably the strongest new protection in thetrade agreements. These provisions differ indetail, but they closely mirror the bilateral in-vestment treaties, even though they are an-chored to the trade agreement.

Despite the proliferation of new protectionsto foreign businesses, the positive economicconsequences have yet to be demonstrated.Theory would suggest that sound propertyrights are a foundation of any country’s invest-

ment climate, and, other things being equal,stronger rights would lower risk and enticemore investment at the margin. Since investorsput money at risk against the promise of re-turns in subsequent periods, predictable regu-lation and protection of property rights are in-tegral to the investment decision. However, theevidence for many of the same protections con-tained in the bilateral investment treaties is thatthese additional protections have no significanteffects on inflows of FDI (box 5.2). To be sure,signing an FTA with new investor protectionsmay enhance the credibility of a reform pro-gram, but evidence that these have observableconsequences is scarce.

While the benefits of these protections inthe form of new FDI inflow are open to ques-tion, the costs in the form of investor suitsare nontrivial and growing. In NAFTA, forexample, as of July 2004, there were 31cases brought under Chapter 11 (including 14against Mexico, 9 against Canada, and 8against the United States). Six cases have beendecided in favor of the investor, but the


107

Does increasing investor protections produce thebenefit of high investment? One test of this

proposition was Hallward-Dreimeier’s (2003) studyof the enhanced investor protections through bilat-eral investment treaties for flows of FDI among sig-natory countries.6 Analyzing bilateral flows ofOECD members to 31 developing countries over twodecades, she found that, controlling for a time trendand other factors, BITs had virtually no independenteffect in increasing FDI to a signatory country froma home country. Said differently, countries signing aBIT were no more likely to receive additional FDIthan countries without such a pact. Even comparingflows in the 3 years after a BIT was signed to the3 years prior, there was no significant increase inFDI. This agrees with the findings of UNCTAD(1998) that the number of BITs signed by the hostwas uncorrelated with the amount of FDI it received.

Box 5.2 Do more investor protections mean moreinvestment? Lessons from bilateral investment treaties

�3 �2 �1

Source: World Bank Global Economic Prospects 2003.

Share of annual FDI flow

1 2YearsignedYears before signing Years after signing

3

BITS do not add much

0

0.1

0.2

0.3

Sources: Hallward-Driemeier 2003; UNCTAD 1998.

amount awarded has been small compared toinitial—and inflated—claims. Tribunal awardshave totaled $35 million, compared to claimsof $1015 billion. Under the similar BITs, 48alleged BIT violations are under review arbi-tration at the International Center for DisputeResolution. Cases have arisen out of theArgentine devaluation, the changes in tax pol-icy perceived as adverse by investors, expro-priations following conflict or coups, irregu-larities in bidding processes, and others(Peterson 2003a). In perhaps the most signifi-cant case to date, a tribunal in Stockholm or-dered the government of the Czech Republicto pay one company, Central European Media(CME), $350 million for violation of a bilat-eral investment treaty that deprived CMEfrom a stake in an English language TVstation in Prague.

This amount was 10 times higher than pre-viously known awards under arbitration casesand about equal to the entire public sectordeficit of the Czech Republic (Peterson2003b).

The legal and macroeconomic conse-quences of investment rights in treaties arelargely unknown. They have not been thor-oughly analyzed and tested in arbitrationcases, and are without precedent. One couldcertainly speculate about adverse outcomes.For example, new rules for Chile could com-plicate management of short-term capitalflows; in fact, the IMF expressed reservationsto the U.S. government that the limitationsthat the U.S.-Chile bilateral FTA imposed onthe Chilean government regarding short-termcapital inflows reduced the government’s abil-ity to manage a macroeconomic crisis. Simi-larly, the breadth of definition of investmentcoverage opens the government to investor-state arbitration in event of default on debt orsuspension of payments in emergencies—which may ultimately be unenforceable. Forinstance, Argentina’s default has led investorsto file nearly 30 arbitration cases; none ofthese appear to have been associated withnonpayment of debt. However, these debts arealso subject to ongoing discussions between

the government and creditors. By definingintellectual property as an “investment,” aforeign investor who claims his intellectualproperty rights have been abrogated has re-course beyond the national court system to in-ternational arbitration proceedings under theinvestment provisions of the bilateral agree-ments.7 These provisions have unquantifiabledevelopment benefits—and bring risk, whichincurs uncertain costs.

The combination of changes in RTAs mayhave an effect on investmentEven if protections by themselves contributelittle additional inflows, evidence is mountingthat RTAs—that is, the combination of appro-priate trade rules, liberalized market access,and investor protections—can have positiveeffects on inflows of foreign investment, pro-vided that the investment climate is supportiveand the size of the newly created market isattractive.

Indeed, Lederman, and others (2004)found that RTAs that formed large markets at-tracted FDI, (controlling for other factors thatinfluence location), but that small marketshad no effect. They also found positive effectsfor NAFTA, although the flow of FDI, evencontrolling for privatizations, appears to havesurged in the first years but has not been sus-tained. Waldkirch’s (2001) study, with lesscomplete annual data, found that NAFTA in-creased FDI substantially, mostly from theUnited States and from Canada. Chudnovskyand Lopez (2001) found that FDI increased inthe MERCOSUR, largely from outsidesources, but that it often entered via acquisi-tion, displaced domestic investment, and wastariff-hopping to produce for the localmarket—so it probably contributed to less togrowth less than it otherwise would have.They also found that FDI inflows tended tolocate in the larger countries, underscoring theneed for stronger institutions and policies inthe smaller countries.

Levy Yayati, Stein, and Daude (2004) useda gravity model to analyze the effects of RTAson FDI inflows in 13 major agreements, and


108

then applied their findings to a simulation forthe FTAA. They found that RTAs have astrong positive impact on inflows, and that ifthese average magnitudes hold after the sign-ing of an FTAA, the results would be substan-tial increases in flows to FTAA countries.However, the distribution is uneven, andcountries with larger post-RTA market size,low inflation rates, strong domestic institu-tions, and open economics are likely to be thebiggest beneficiaries.

To investigate further whether RTA forma-tion can affect FDI flows in a consistent fash-ion, we examine the effects of RTA member-ship and other variables on FDI inflows for apanel of 152 countries over the 1980–2002period. The sample takes into account 238RTAs (both regional and bilateral), many ofwhich overlap, that encompass the vast ma-jority of sample countries.8 In general, coun-tries that are more open (measured as the sumof exports and imports over GDP), growingmore rapidly, and are more stable (capturedin less volatile inflation rates), attract greaterquantities of FDI, controlling for growth ratesof FDI to all countries and the world growthrate.9 RTAs that result in larger markets do at-tract greater FDI. The interaction of an RTAsigning and additional market size associatedwith the integrated markets is significant andpositively related to FDI. On average, a 10percent increase in market size associated withan RTA produces an increase of 5 percent.10

This has important policy implementations: Ifa country seeks to use an RTA to attract in-vestment, it should seek to amalgamate withthe largest possible markets; RTAs amongsmall market countries have little effect.

Two important caveats to this conclusionare worth underscoring: First, a preferentialtrading arrangement cannot compensate foran inadequate investment climate. Stein andDaude (2001) have shown that institutionalvariables that make up the whole of a coun-try’s investment climate—including politicalstability, government effectiveness, rule oflaw, and lower risks of expropriation—are allsignificantly associated with increases in

investment flows, controlling for other deter-minants of FDI. These wash out the otherwisepositive effects of RTAs. If the economy suf-fers from poor macroeconomic management,high levels of corruption, and poor infrastruc-ture, an RTA by itself will not offset the dis-advantages. To be sure, an RTA may help gov-ernments through their collective action toimprove the investment climate and bring inmore investment; but an RTA is no substitutefor an adequate investment climate (WorldBank 2004). Second, creation of an RTA willnot have much effect on investment inflowsfrom outside the region if restrictions on mar-ket access are severe and remain unchanged.

How do IPRs affect the priceof technology? Creation and enforcement of IPRs have an im-portant role to play in development, but nei-ther theory nor available studies provide muchguidance on the likely outcomes of imple-menting in trade agreements the strongest ofthe IPRs or none at all. On the one hand,stronger IPR enforcement in general is likelyto enhance the overall investment climate,especially for high technology firms. On theother, recognizing full patent protection forfirms may require poor countries to payhigher prices, with little additional incentiveeither to innovate or to make investments inthe local market.11 Full enforcement ofpatents12 could produce substantial financialflows, estimated roughly at $19 billion tothe United States and $7 billion to Germany(World Bank 2001). Moreover, the adminis-trative costs of upgrading IPR systems are nottrivial (Finger and Schuler 2004).

It was the prospect that developing coun-tries would have to pay higher prices forpatented drugs that motivated the interna-tional community to agree to clarify flexibili-ties embedded in the TRIPS Agreement at theWTO Ministerial Meeting in 2001. Theresulting Doha Declaration on TRIPS andPublic Health reaffirmed the right of WTOmembers to use the flexibilities of TRIPS inthe areas of compulsory licensing and parallel


109

importation to “. . . promote access to medi-cines for all.”13 In August 2003, WTO mem-bers created a special mechanism under theTRIPS Agreement that allows countries withinsufficient manufacturing capacity to effec-tively use compulsory licenses by importinggeneric drugs.

At first blush, the TRIPS-Plus portions of theU.S. FTAs seem to circumscribe the policy spaceprovided in the Doha Declaration. In particu-lar, provisions on the link between patent sta-tus, marketing approval, and data exclusivityappear to put limits on the spirit of the DohaDeclaration, because countries may be pre-vented from effectively employing compulsorylicenses to introduce competition from genericdrug producers.14 To address these concerns,the U.S. bilateral agreements with Bahrain,CAFTA-DR, and Morocco contain side lettersthat share the understanding that the intellec-tual property chapters do not affect the abilityof governments to “. . . take necessary mea-sures to protect public health by promotingmedicines for all [. . .].”15 In other words,government can be justified as protectingpublic health, as permitted under the threeFTAs. The United States Trade Representativeoffice recently clarified: “. . . if . . . a drug isproduced under a compulsory license, and it isnecessary to approve that drug to protect pub-lic health . . . the data protection provision inthe FTA would not stand in the way.”16

Notwithstanding the potential flexibilitiesprovided by these side letters, they raise sev-eral questions. How widely will the parties tothe three agreements define the “protection ofpublic health”—or, what definitions would anarbitration panel use? Uncertainty in this re-spect may become itself a barrier to makinguse of the flexibilities and may open the doorfor restrictive interpretations by vested inter-ests. Also, several of the other U.S. FTAs donot contain comparable side letters, raisingquestions about conflicts between intellectualproperty obligations and public health objec-tives in at least some of the affected countries.

The welfare effects of stronger and newcopyright protection standards are ambiguous.

On the one hand, most countries haveindustries that rely on copyright protection andthat may benefit from strengthened protection.And new technologies that greatly facilitate thecopying of digital works pose challenges thatpolicymakers need to address. On the otherhand, copyright laws have historically soughtto strike a balance between the interests ofcopyright producers and the interests of thegeneral public. So-called fair use exemptionsallow the copying of protected works for edu-cational or research purposes. There are ques-tions that new rules on the technological pro-tection measures and the liability of Internetservice providers could diminish the rights ofconsumers and the general public (CIPR 2002).Ensuring fair use of copyrighted material isparticularly important for educational mater-ial. The opportunities and gains from the use ofdigital libraries, Internet-based distance learn-ing programs, or online databases would belimited if access to such tools became unaf-fordable or otherwise restricted by copy-right law.

Evidence is inconclusive about the respon-siveness of FDI to intellectual propertyregimes. Although surveys of foreign investorstypically indicate concerns for IPRs, this isoften of secondary priority (Mansfield 1995).Maskus (2000) concludes that countries (espe-cially low-income countries) should focus ontheir overall investment climate to attractmore and high technology investment, ratherthan to fine-tune their IPRs. Nonetheless,some multinational companies (MNCs), whenselecting an investment location amongmiddle-income countries, clearly take into ac-count the laws governing intellectual property.Other studies have found that weak laws andweak enforcement deter investment in middle-income countries,17 but the results for low-income countries are inconclusive. Finger andNogues (2002), in fact, argue that the intro-duction of patent protection for drugs in Chilemade several multinational pharmaceuticalcompanies stop production and investmentand source this market from other locations.In summary, Fink and Maskus (2004) in their


110

review of the evidence conclude: “. . . coun-tries that strengthen their IPR regimes are un-likely to experience a sudden boost in inflowsof foreign direct investment,” but that IPRscan stimulate formal technology transferthrough FDI and licensing.

All in all, the general conclusion is that coun-tries have to develop an IPR strategy appropri-ate to their level of development, and then ana-lyze carefully which if any IPR provisions oughtto be contained in trade treaties or RTAs.

RTAs and Provisions forMovement of LaborUsing RTAs to promote movementof unskilled workersWalmsley and Winters (2003) estimated that ifa temporary visa system were introduced inthose developed countries that permitted themovement of up to three percent of their laborforce, world incomes would rise some $160billion. Some 70 percent of the global welfaregains from increased migration would comefrom the movement of unskilled workers. Todate, progress under the GATS Mode 4 nego-tiations in easing restrictions on the temporaryentry of workers has been limited; the agree-ment has generally been used for skilledworkers, not unskilled.

Regional agreements might offer a morepromising venue for realizing the gains from

temporary movement of workers—that is, bygoing beyond the relatively limited scope ofthe GATS Mode 4 provisions. Regional agree-ments are often between countries with histor-ical ties, and former migrants (or their chil-dren and grandchildren) tend to supportincreased opportunities for citizens of theirhome countries. Working on bilateral orregional levels may provide receiving coun-tries with greater control over the numbersand nationalities of temporary workers. Thatis, receiving countries that wish to ensuremaximum control in immigration decisionsmay prefer to design their policies withouthaving to negotiate the terms with countriesoutside the region. Given both labor marketand security concerns, this control may makeit easier for countries to implement temporaryprograms for unskilled workers. Close neigh-bors tend to have a high proportion of immi-grants in receiving countries. To what extenthave RTAs facilitated labor movement?

Regional agreements treat labor movementin varying ways, ranging from full labor mo-bility to no provisions at all. As shown intable 5.3, RTAs treat labor in one of four dif-ferent ways:18

• free labor mobility, with limited excep-tions;

• temporary market access for certain(usually skilled) groups of workers;


111

Table 5.3 Summary of agreements by degree of labor mobility

Degree of labor mobility under agreement Agreements

Full labor mobility European Union, Agreement on the European Economic Area, European FreeTrade Association, Australia-New Zealand Closer Economic Relations

Market access for certain groups Caribbean Community, North American Free Trade Agreement, EuropeAgreements, Group of Three, and Canada-Chile, U.S.-Singapore, U.S.-Chile,Japan-Singapore Free Trade Agreements

Based on GATS Mode 4, with ASEAN Free Trade Area, Euro-Med Association Agreements, New Zealand-additional provisions or limitations Singapore Closer Economic Partnership, Southern Common Market

Agreement, and EU-Mexico, EU-Chile, MERCOSUR, and US-Jordan Free Trade Agreements

No effective provisions for labor mobility Asia Pacific Economic Co-Operation Forum, South Asian Association forRegional Cooperation, Central European Free Trade Agreement, andCommon Market for Eastern and Southern Africa

Source: World Bank staff; Nofal 2004.

• temporary movement based on theGATS Mode 4 model, often with addi-tional provisions or limitations;19 and

• no provision in place for market access(beyond facilitating entry visas), or plansdesignated to be realized only in thefuture.

Most agreements do not override migrationlegislation, and parties retain broad discretionto grant, refuse, and administer residence per-mits and visas. It should be noted also that theright of labor mobility does not automaticallyentail the right to practice a certain profession;national regulations regarding licensing andrecognition of qualifications are stillapplied.20

RTAs with full labor mobility The EU, the European Economic Area andEuropean Free Trade Association, and theAustralia-New Zealand Closer EconomicRelations Trade Agreement (ANZCERTA)allow for free labor mobility, with very limitedexceptions.21 The EU also allows the right toreside (with family), although residence per-mits are not required for stays of less than

three months. The European agreements in-clude exceptions for public services, public se-curity, and/or public health works.ANZCERTA provides for both full marketaccess and national treatment for all servicesuppliers, excluding a few sectors.

The migration provisions of these agree-ments facilitated regional integration by mak-ing it easier for firms from one country totransfer personnel to their operations in othercountries in the region. The potential forworkers to move to higher-paying jobs inother countries may have improved disciplinein some labor markets. However, the EUagreements had almost no impact on stocks ofpermanent migration from other EU countries(figure 5.1), in part because several EU coun-tries had already provided for free labor mo-bility. In addition, most of these countries areof similar income levels, so the incentive todisrupt family and personal relationships bymoving for higher incomes was limited. Theentry of less wealthy countries into the EU inthe 1980s also did not result in greatly in-creased migration.

Several factors may explain this: amongthese, the free movement of workers from the


112

Source: OECD.

Figure 5.1 Share of EU nationals in total population, for selected EU countries and Norway,1985–2000

Denm

ark

Sweden

Irelan

d

United

King

dom

Austri

a

Nethe

rland

s

Finlan

d

Germ

any

Belgium

Fran

ceIta

lySpa

in

Portu

gal

Norway

Percent

0

1

2

3

4

5

6

1985

1990 1995

2000

Highest income Lowest income

new member states were subject to transi-tional periods; the EU’s regional funds aimedat developing less prosperous regions com-bined with the prospect of a positive economicdevelopment in the new countries are likely tohave thwarted a significant migration of labor.Greece, Portugal, and Spain underwent6 years of transition, which limited the freemovement of people after their entry to theEU. Even the richest EU countries (Denmarkand Sweden) experienced only a slight rise inEU nationals as a share of population after thetransition periods of the poorest EU countries(Spain and Portugal) expired.22

It remains to be seen whether the recent ac-cession of Central and Eastern European coun-tries to the EU will result in substantial migra-tion. While this agreement will ultimatelyprovide for full labor mobility, most of theoriginal EU countries have taken advantage

of provisions that delay this migration for alimited period (renewable up to 7 years). Moststudies find that the accession agreements areunlikely to greatly boost migration to WesternEurope. Forecasts of the additional migrationdue to expansion of the EU, which relies onboth econometric models and opinion polls,generally find that migration will be limited toabout 3–4 percent of residents of the 10 first-round East European nations within a decadeof freedom of movement; this amounts toabout 4 million people or 1 percent of the cur-rent EU population. Additionally, about half ofthese laborers are likely to return home withinthe 10-year period, meaning a net migration of2 million persons (Martin 2003). However,some of the EU countries, particularly thosebordering the Eastern European countries likeGermany and Austria, could experience largerflows relative to population.


113

The number of migrant workers without properwork permits, so-called undocumented or illegal

workers, is increasing. The International Organiza-tion of Migration estimates that each year, some-where between 700,000 and 2 million people crossborders to take jobs without legal permission.*Of this number, up to 500,000 seek entry intoWestern Europe and 500,000 into the United States,Canada, Australia, and New Zealand. In the UnitedStates, the stock of undocumented workers increasedfrom an estimated 3.5 million in 1990 to 7 million in2000, of which 69 percent were Mexicans (USCIS2003). In Europe, illegal migrants numbered as manyas 3 million by the end of the 1990s.* Countries inAsia have also experienced an increased number of ir-regular foreign workers in their labor force. In Singa-pore, undocumented workers accounted for 4.5 per-cent of the total labor force in 2000, and in HongKong (China), it was 2.3 percent. In Hong Kong(China), Korea, and Taiwan, there is one undocu-mented foreign worker for every three legal foreign

Box 5.3 Illegal migration: A growing global phenomenon

workers. Thailand is one of the countries most af-fected by undocumented workers, with nearly 5.5undocumented workers for each registered foreignworker (IOM 2003). In 2000–01, some 73,000Chinese irregular migrants were assumed to be livingin South Korea, and, in turn, China hosted some50,000 irregular migrants.

Stricter enforcement of immigration rules may nothave been successful in cutting off illegal migration,but it may have contributed to the very high costsinvolved, particularly between countries that do nothave a common border. Today, irregular migrantsfrom China wanting to enter the United States payup to $35,000 to smugglers,† to Europe they pay inthe range of $10,000–$15,000, and those seekingentry into Japan up to $10,000. The “fee” for mov-ing from Lebanon to Germany varies between$5,000–$10,000; from India to the United States it isaround $25,000, and from North Africa to Spain itis between $2,000–$3,500.



114

The vast majority of undocumented workers areunskilled for several reasons. Policies in most receiv-ing countries offer greater opportunities for skilledworkers to migrate legally than for unskilled. Manyskilled workers, coming from middle- or upper-income households, may be less willing to experiencethe uncertainty inherent in breaking the immigrationlaw (including the prospect of being jailed temporar-ily), than persons from lower-income households.Also, employers are more likely to offer unskilledjobs to undocumented workers, whose tenure is rela-tively uncertain, as skilled jobs may require a greaterinvestment to apply technical backgrounds to thedemands of specific jobs.

Efforts to control illegal entry have varied ineffectiveness.‡ Receiving countries have addressed un-documented workers by offering amnesties that gen-erally involve the promise of regularization for un-documented workers who have been in the countryfor a period of time. For example, the U.S. Immigra-tion and Regularization Control Act (IRCA) of 1986regularized the status of undocumented workers whohad been living in the country since before 1982, aswell as undocumented agricultural workers. This pro-gram, which took effect mostly in 1989–91, granted aregular status to more than 2.5 million people, mostlyfrom Mexico and Central America. However, thismassive regularization program did not prevent Mex-icans or other undocumented immigrants from con-tinuing to cross the U.S. border. Similarly, in Europe,approximately 1.5 million undocumented migrantssaw their status regularized under amnesty programsimplemented by Belgium, France, Greece, Portugal,and Spain during the 1990s.§

Box 5.3 (continued)

Regularization programs for undocumented arealso present across developing countries. Forexample, in the 1990s, Argentina regularized undoc-umented workers from Bolivia, Paraguay, and Peruunder bilateral agreements. Thailand had a similarprogram. Other countries have opted for more dras-tic measures to address undocumented workers. Forexample, Malaysia in 2002 deported hundreds ofthousands of irregular migrants back to Indonesiaand the Philippines, in reaction to rising levelsof criminality; however, there was a slowdown inits economy given its dependency on foreignworkers.

*IOM (2003). Unless otherwise noted, the data in this boxare based on this document.

**The number may be even higher today. The United King-dom may have up to one million irregular migrants (UK Immi-gration Service), France 500,000, Belgium 90,000 (Belgium Anti-racist Centre), and Ireland 10,000 (Irish Police). In terms offlows, some 100,000 irregulars are smuggled into Germany eachyear (German Police Trade Union), and some 95,000 irregularsfrom Albania, Romania, and Iraq alone enter Greece each year.

†Annually, some 25,000 to 50,000 Chinese irregular mi-grants enter the United States.

‡Hanson and Spilimbergo (1999) find that additional re-sources devoted to border enforcement in the United States haveyielded only limited results, and Boeri and others (2002) find thattighter controls at a given entry point can be effective, but divertmigrants to other points of entry. Worksite inspections can havea greater impact, but in the United States such efforts are verylimited.

§Italy regularized 716,000 irregular migrants in three waves;Greece accepted 370,000 people (in 1997–98) mostly from theBalkans and Eastern Europe; Spain regularized 260,000 irregularimmigrants mostly from Africa and Latin America; and Portugalregularized 61,000.

Some countries have recognized the need tomanage both regular and irregular migrationon a regional basis. Regional consultativeprocesses, such as the Manila Process (1996),the Migration Dialogue for Southern Africa(MIDSA 2000), the Puebla Process (1996), andothers have emerged—and all of them includeconsultations to deter human trafficking and

the movement of undocumented workers.However, most of these groups are informalones that mostly share information onmigration-related issues and generally do notimpose requirements on the immigration poli-cies of the participating countries. It appearsthat neither unilateral policies nor consultativearrangements have had significant success in

Source: International Organization for Migration (IOM) 2003.

controlling the substantial pressures formigration to industrial countries.

Agreements that permit temporary accessfor certain groups Several agreements provide for market accessof certain groups of workers, usually thehighly skilled. The most recent agreementshave focused heavily on intra-corporate trans-ferees, including managers and skilled techni-cal staff.

The Central America and Caribbean Com-munity (CARICOM) allows university gradu-ates to move among member countrieswithout passport requirements and allowsuniversity graduates, professionals, skilledpersons, and workers from some selected oc-cupations to work without a permit.

NAFTA, along with the Canada-Chile,U.S.-Chile, and the U.S.-Singapore Free TradeAgreements, do not provide for permanentmigration, but allow for the temporary move-ment of business visitors, traders and in-vestors, intra-corporate transferees, and pro-fessionals.23 Visas are still required for all fourcategories, and work permits are required forall except business visitors.24 Numeric restric-tions on temporary entry are not allowed forthe first three categories, but were imposed bythe United States on professionals from Chile,Mexico, and Singapore, but not from Canada.U.S. professionals seeking to enter either mar-ket are not subject to numerical limits. In theCanada-Chile agreement there are no numeri-cal limits to any of the four categories. Fol-lowing the NAFTA agreement, the number ofprofessionals entering the United States fromCanada and Mexico increased substantially,and although the gap in favor of Canadiansremained high, it narrowed considerably by2002. Similarly, in the treaty traders and in-vestors category, the ratio of Mexican overCanadian admittances into the United Statesincreased from 0.4 in 1996 to 1.1 in 2002.25

The EU agreements with Central and East-ern European countries (prior to their 2004accession to the EU) allowed for temporaryentry of workers providing a service,

managers and/or highly qualified employees,and company representatives negotiating forthe sale or supply of services. Transition peri-ods as well as restrictions for public serviceworks and sectoral exclusions applied. Theyare still in force for the nonaccession coun-tries, Bulgaria and Romania. The Japan-Singapore FTA provides for the temporarymovement of natural persons for business pur-poses, including investors, subject to condi-tions for entry (such as pre-employment for atleast one year for intracorporate transferees)and time limits of stay. The Group of Three(Colombia, Mexico, and Republica Bolivari-ana de Venezuela) facilitates temporary entryfor business persons. As in other agreements,GATS Mode 4 restrictions regarding accessto the employment market or permanent em-ployment also apply. MERCOSUR also hasprovisions to facilitate the temporary entry ofbusiness persons and the exercise of temporaryprofessional practices, and it is working on theissues relative to a “MERCOSUR NationalResidency” and a more flexible migratory reg-ulation for MERCOSUR citizens (Nofal2004). Other agreements among Latin Ameri-can countries (such as the Mexico-Nicaragua,Mexico-Chile, Mexico-Bolivia, Mexico-CostaRica, and the agreement between CentralAmerica and Dominican Republic) containsimilar provisions.

Agreements based on GATS Mode 4 Several bilateral agreements, along with theASEAN Free Trade Agreement (AFTA) andMERCOSUR, base labor mobility provisionson the principles of Mode 4 of the GATS. How-ever, most of these agreements add provisionsthat allow for labor mobility in categories notcovered by the GATS. For example, the U.S.-Jordan agreement facilitates visa arrangementsfor independent traders and persons linked toinvestment. The EU-Mexico agreement pro-vides for a standstill and sets common regula-tions of work, labor conditions, and residencypermits for temporary workers in eachcountry. The AFTA since 1998 has covered allmodes of supply, including services sectors not


115

previously covered by the GATS. The Frame-work Agreement on the ASEAN InvestmentArea (1998) commits members to support freerflows of skilled labor, professionals, capitaland technology among ASEAN members. Bycontrast, the Euro-Med agreements withMorocco and Tunisia do not provide for pref-erential access beyond GATS (the case for otherMediterranean agreements as well). The NewZealand-Singapore agreement generally fol-lows the GATS model regarding labor mobilityfor service suppliers. MERCOSUR (and itsagreements with Bolivia and Chile) is limited tothe GATS provisions on the movement of

natural persons as services suppliers (Nofal2004).

Agreements without effective provisionsfor market access The Asia Pacific Economic CooperationForum (APEC), the South Asia Association forRegional Cooperation (SAARC), and Com-mon Market for Eastern and Southern Africa(COMESA) do not provide for labor mobility.However, the first two have taken measures tofacilitate business travel. Most of the membersof APEC (except the United States andCanada) participate in the Business Travel


116

The U.S. has several visa categories for the admis-sion of temporary workers. NAFTA provided

for the temporary admission of professionals underTN visas and made it easier for Mexicans to takeadvantage of already existing visa categories forskilled workers and professionals. However, NAFTAmade no provision for increasing admissions of un-skilled workers to the United States. In contrast, theUnited States has unilateral programs that do allowfor the temporary admission of unskilled workers,and these have grown from from about 14,000 in

Box 5.4 U.S. temporary admission programs underNAFTA and unilateral policies

1996 to about 66,000 in 2002. Moreover, the largesttemporary business visitor program (B1 visa) is lim-ited to six months (renewable, but total time cannotexceed one year), and generally is used for short-term trips to the United States. The other visas canbe granted for one year or more, often with exten-sions. Thus more long-term unskilled Mexican work-ers have been admitted to the United States underunilateral programs than under programs initiatedunder, or supported by, NAFTA.

Temporary entry of Mexican workers to the United States(thousands of persons)

1996 2002Program initiated under NAFTA

Professional (TN visa) 0.2 1.8Programs supported by NAFTAa

Intracompany transferees (L1 visas) 4.8 15.3Treaty traders and investors (E1, E2 visas) 1.0 4.0Business visitors (B1 visas) 309.0 475.0

Unilateral programsProfessionals with specialty occupations (H1B visas) 5.3 15.9Agricultural temporary workers (H2A visas) 8.8 12.8Non-agricultural temporary workers (H2B visas) 5.5 53.0

aThese visa categories existed before NAFTA and apply to many countries. However, provisions of NAFTA make it easier forMexicans to use them. Source: USCIS 2002, INS 1997.

Card Scheme, under which holders of the cardreceive expedited entry at the airport, and arenot required to submit separate applicationsfor business visas. Members of the SAARCadopted a Visa Waiver Scheme in 1992, whichexempts 21 categories of persons from visa re-quirements. Free labor movement is envisagedas a long-term objective of COMESA, to beaccomplished by 2025. However, progress hasbeen limited to date.26

To date: Limited labor integration Most regional agreements have had little im-pact on increasing migration. First, the agree-ments with full labor mobility have been be-tween countries of similar income levels, sothere has been little incentive for migration.Most agreements, and particularly thoseagreements that include both industrial anddeveloping country members, do not allow forpermanent migration.

Second, while these agreements often pro-vide for some temporary labor mobility, par-ticularly for the service sectors, the provisionsare generally restricted to higher-skilled work-ers. This is consistent with the trend in indus-trial countries of changing unilateral migra-tion policies to attract higher-skilledworkers—in principle on a temporary basis,but in practice allowing for the possibility tosettle after some period of time. For example,the Temporary Immigration Program of Aus-tralia allows for unlimited visa renewal forskilled workers. In the United Kingdom,highly skilled temporary workers in certainoccupations can settle after 4 years of contin-uous work. In Norway, temporary workerswith special skills can be issued a permanentwork permit after 3 years of stay. In Canada,the 2002 Immigration and Refugee ProtectionAct (IRPA) placed more emphasis on educa-tion, job experience, and language ability inallowing entry. This trend toward favoringhigher-skilled workers limits the potentialglobal gains from migration.

To the extent that developed countries havepermitted entry for unskilled workers fromRTA countries, it has occurred through

parallel programs, many of which predatedthe RTA, such as programs in the UnitedStates with Mexico and in the EU with non-members of Southern Europe. Where indus-trial countries have allowed for some admit-tance of unskilled workers, the rules differsignificantly from those governing admissionof skilled workers. Unskilled workers legallyentering host countries usually do so underseasonal work agreements, project (or guest)worker agreements, or specific provisionssuch as the working holiday maker programs(WHMP). Seasonal employment usually al-lows foreign workers to stay in the host coun-try for periods between three months and oneyear, with work permits provided to foreignworkers only in the event that no domesticlabor can do the job and only in some specificsectors (such as agriculture, forestry, andtourism). Project worker agreements allowforeign workers entry for specific projects andusually include limits on the maximum stayand quotas (often determined on the basis oflabor market conditions). WHMP allowsyoung people (roughly 18–30 years old) toholiday and work for short periods, providedthere is a prior bilateral working holidayagreement among the involved countries.

Developing countries’ regional agreementsalso tend to discriminate in favor of skilledworkers when providing for labor mobility. Asmentioned above, CARICOM allows for thefree movement of university graduates, otherprofessionals, skilled persons, and workersfrom selected occupations. Several LatinAmerican agreements (Group of Three, Cen-tral America-Dominican Republic, and agree-ments between Mexico on the one hand, andChile, Nicaragua, Bolivia, and Costa Rica onthe other) also provide for the movement ofsome skilled, not unskilled, workers.

Conclusions: Beyond MerchandiseTrade

Agreements differ markedly in their treat-ment—to say nothing of their implementa-

tion—of non-merchandise provisions for


117

services, investment, intellectual property, andtemporary movement of workers. Differencesbetween the United States and EU bilateral FTAsand other agreements are particularly vast.

In those areas where RTAs could seriouslypromote development—services liberalizationand temporary movement of workers—resultshave ranged from mixed to missed opportuni-ties. In general, the U.S. FTAs have promptedsome additional services market openings inBahrain, CAFTA, and Morocco, some changesin Singapore, but relatively few changes inAustralia and Chile. To be sure, even thoseagreements that required no additional mar-ket opening could benefit developing coun-tries, because investors may attach credibilityto the lock-in effects of the treaty.

In the South, agreements that have sub-stantially improved services access are rela-tively limited, and those with greater marketaccess often have the most restrictive rulesof origin for investor nationality. Morecommon is the lack of progress on servicesliberalization.

More disappointing from a developmentperspective is the minimal attention given tocreating opportunities for the temporarymovement of workers, particularly unskilledworkers. In neither North-South nor South-South agreements is there evidence of muchactivity. In the wake of the September 11,2001, attacks on the United States, concernsfor security have made cross-border move-ment of all persons subject to greater controlsand scrutiny. This atmosphere does not bodewell for expanding programs for temporaryworkers.

At the same time, in those regulatory areaswhere development benefits are largely un-proven, the North-South bilateral FTAs arestrengthening their rules. Strengthening therules governing investment and intellectualproperty may contribute to better institutionalenvironment, but the greatest gain is to befound in services. Enhancing protections of-fered to investors has not been shown to in-crease the flow of investment, and preventing

the erosion of monopolistic returns to theowners of technology through enhanced IPRsis of doubtful development benefit for the av-erage developing country.

Moreover, the downside risks of misjudg-ments in terms of adverse legal and economicramifications are nontrivial, especially forunsophisticated governments. Internationaltreaty law in these areas is evolving fast andis being set through case laws of arbitrationpanels, whose judgments at times conflict(Ewing-Chow 2001). Governments may findthemselves hauled before arbitration panelsand compelled to pay large amounts of com-pensation for enacting regulations they hadconsidered in their sovereign domain.

Broadly defining investment to include allcapital flows and assets, including intellec-tual property, carries risks. For example, onepotential risk, is the set of provisions associ-ated with the U.S.-Chile arrangement. Theseprovisions could limit a less sophisticatedgovernment’s ability to deal with a financialcrisis.

Realizing the promise of services,investment, IPRs, and labor mobilityUsing RTAs as a lever to liberalize services hasseveral advantages. The potential for improv-ing economy-wide performance is often great.Moreover, most services liberalization is in-herently multilateral. This is because membergovernments that open markets want to real-ize the immediate gain of full competition andso open markets to all comers, or becausemembers adopt lenient rules regarding domi-cile of investors, thereby permitting externalinvestors to invest in the region through sub-sidiaries located in member countries. Even ifrestrictions impede full MFN access, regionalagreements can be cost-free stepping stonesto open markets insofar as they do not carrythe trade-diversion costs of lost revenues asso-ciated with tariff reductions for goods trade.More could be done in RTAs, particularlyin South-South agreements, to realize thesebenefits.


118

Countries have to design strategies towardinvestment and intellectual property that areappropriate to their development prioritiesand then analyze carefully which elements, ifany, ought to be contained in trade treaties.For some countries, improving other propertyrights—for example, land rights or small busi-ness assets—may have a higher priority thanestablishing market access rights for investorsor rights for patent holders; for other coun-tries, especially middle-income countries,these may indeed be a priority. Once a strategyis in place, it is possible to ascertain which ofthe new rules customarily associated withregional trade agreements make sense fordevelopment.

Evaluating the development benefit of anygiven RTA cannot rest solely on one compo-nent because often rules are accepted in onearea to achieve market openings in partnercountries in another. For example, the net de-velopment benefits of the investment and IPRcomponents hinge critically on their appropri-ateness for development and the market accessgranted in product markets of partnercountries.

A corollary lesson is that multilateral liber-alization in goods markets is essential forreaping any gains from RTAs that contain newrules. Governments wishing to maximizegains from RTAs should ensure that new rulesare consistent with extant border protection,and if not, they should consider lowering thatprotection. If a country has high tariff barriersand forms an RTA with a partner requiring in-vestor protections and TRIPS-Plus, it may endup entrenching the businesses of that partnerbehind new barriers. It may confer first moveradvantages in services and IPR restrictions indrug or other high technology manufacture,with reinforced barriers to parallel imports ornew entry from third parties.

Finally, RTAs—and other regional cooper-ation agreements—do offer some importantopportunities for countries to collaborate, es-pecially in South-South agreements. To stimu-late investment, particularly in services, they

might adopt common standards that facilitatecross-border competition in services and in-vestment. Adopting common technical stan-dards for telecommunications, for example,has helped integrate markets and opens theway for competition. To adopt internationalnorms regulating technology, new agreementscould adopt their own IPR standards, with theadvantage of foreclosing additional restric-tions motivated by private interest groups inthe North. At the same time, some regionsmay find opportunities to agree on commonadministration of patent and copyright law.

Notes1. Agreements can also increase the number of

competitors by allowing cross-border provision ofservices, Mode 1, such as supplying back-office ser-vices. Restrictions on these tend to be less commonthan on supplying through commercial presence of aforeign subsidiary, Mode 3. One exception is supplyingvarious types of insurance and reinsurance, which canbe done cross-border, though a sales affiliate is nor-mally required.

2. RTAs that restrict service providers to membercountries by definition limit the number of potentialcompetitors relative to multilateral liberalization. Evenif later followed with multilateral opening, RTAs mayconfer on members first mover advantages, and, ifsome market barriers to entry remain, they will not bereadily competed away; the result is higher prices toconsumers.

3. According to the agreement, if Chile chooses toimpose restrictive measures on capital flows it consid-ers speculative, then special dispute settlement ruleswill apply.

4. For some FTA countries, pre-existing BITs hadvirtually the same rights; however, about half of thecountries had no BIT prior to the FTA, and where BITSdid exist, pre-establishments were often fewer and lessextensive.

5. See Fink and Reichenmiller (2004) for a moredetailed review.

6. These include, for example, TRIPS, the ParisConvention for the Protection of Industrial Property,the Berne Convention for the Protection of Literaryand Artistic Works, the Rome Convention for the Pro-tection of Performers, Producers of Phonograms andBroadcasting Organizations, and others also contains alist of new conventions that parties are expected to rat-ify within a specified timeline, such as the Budapest


119

Treaty on the International Recognition of the Depositof Micro-organisms for Mexico and the World Intel-lectual Property Organization (WIPO) CopyrightTreaty for Chile. Finally, the list includes several con-ventions that the member states are expected to ratify“as soon as possible” without mentioning a specificdeadline, such as the WIPO Copyright Treaty forMexico or the Madrid Agreement concerning the In-ternational Registration of Marks for Chile.

7. BITs customarily provide a definition of invest-ment coverage, provide investor protections such asagainst expropriation, require national treatment forpost-entry establishments, stipulate compensationfor the expropriation of their investments, and providefor a dispute resolution mechanism. The latter usuallypermit the investor to sue the state for breach of treatyunder binding arbitration. In some cases, treaties pro-scribe any government action that would reduce thevalue of the private investment, even if it were envi-ronmental, and establish grounds for compensation.Such compensation could either entail extensive liabil-ities for the host government or compel them to refrainfrom making certain policy choices.

8. The United States initiated the practice of defin-ing investment to broadly include intellectual propertyin the late 1990s in negotiations of its bilateral invest-ment treaties. See Vivas-Eugui (2003).

9. Some of the problems include the absence ofdata on implementation and the variable coverage ofFDI provisions across agreements, which makes it dif-ficult to distinguish the effects of investment rules fromtrade rules. Moreover, the absence on FDI data thatwould enable us to distinguish the effects of RTAs ondiffering type of investment—vertical or horizontal–limits the analysis. Nonetheless, the regressions arerobust to variations in specifications.

10. The regression with fixed effects estimation ofnet FDI inflows is:

lfdi | Coef. Std. Err. t P�|t|lgdp | .9404982 .2065772 4.55 0.000lgnppc | �.1228465 .2008249 �0.61 0.541open | .0051226 .0011387 4.50 0.000growth | .0198651 .0040816 4.87 0.000cpi | �.0196485 .0060906 �3.23 0.001lfdiwld | .4472645 .0719058 6.22 0.000growld | �.0611576 .0432152 �1.42 0.157lftagdp | .0518633 .0163279 3.18 0.002R-sq: within � 0.3973 corr(u_i, Xb) �-0.0410between � 0.7469 F(28,2003) � 47.16overall � 0.6690 Prob � F � 0.0000F test that all u_i � 0: F(143, 2003) � 12.79 Prob � F � 0.000011. As mentioned earlier, this variable contains the

sum of the host country’s RTA partners GDP, excluding

the host country itself. Thus if we consider Brazil as thehost country and MERCOSUR as the relevant RTA,the variable lftagdp would be the log of the sum ofGDP of Argentina, Paraguay, and Uruguay. This vari-able serves a twofold purpose in the estimation routine.First, since it is equal to zero prior to signing an RTAand carries a positive value afterwards, it measureswhether signing an agreement has an effect on FDI in-flows (i.e., including a dummy variable for RTA mem-bership would be counterproductive in the presence ofthis variable, since it will capture the “threshold effect”of signing an RTA). Furthermore, this variable alsocaptures the effects of participating in a larger marketfollowing the signing of an agreement. This is particu-larly important if a country is party to more than oneagreement—the variable will then be a sum of all of itspartners’ GDP, reflecting the fact that the country hasnow created a larger market. The fact that this variableis positive and significant shows not only that signingan RTA will generally bring benefits in terms of greaterFDI inflows, but also that larger market size of thecountry’s partners tends to generate more incomingFDI.

12. See Finger and Schuler (2004) for a richer dis-cussion of the the asymmetry in TRIPS; they argue thatit protects the knowledge that businesses and individu-als have in rich countries and that poor people buy, butnot the knowledge that poor people generate and sellto the world.

13. In this sense, full enforcement is equivalent to aTRIPs standard (Maskus 2000, 183–85).

14. See paragraph 4 of the Doha Declaration onTRIPS and Public Health, available at http://www.wto.org.

15. Technically, the Doha Declaration does notaddress questions of marketing approval during thepatent term and test data exclusivity. However, the pro-visions of the bilateral FTAs in these areas can still beseen as being at odds with the spirit of the Doha Dec-laration, to the extent that they preclude the effectiveuse of compulsory licenses.

16. See the letter from USTR General CounselJohn K. Veroncau to Congressman Levin datedJuly 19, 2004, available at Inside US Trade. Moreover,the side letters refer “in particular [to] cases such asHIV/AIDS, tuberculosis, malaria, and other epidemicsas well as circumstances of extreme urgency on na-tional emergency.” The language chosen mirrors word-ing in the Doha Declaration on TRIPS and PublicHealth, except that the latter employs the term “espe-cially” instead of “in particular” and does not mention“circumstances of extreme urgency or national emer-gency.” This difference in language may imply a nar-rower definition of public health and therefore flexibil-ity to issue a compulsory license, but this is subject to


120

legal interpretation that must await actual cases of dis-pute settlement.

17. See the letter from USTR General CounselJohn K. Veroneau to Congressman Levin datedJuly 19, 2004.

18. See the excellent studies in Fink and Maskus(2004) including particularly Fink (2004) for the U.S.and German MNCs; Smarzynska Javorcik (2004) forEastern Europe; and Maskus, Dougherty, and Mertha(2004) for China.

19. The description on agreements that follows isstrongly based on Nielson (2003) and a complement,in some cases, to the text of the agreements.

20. GATS (under “Mode 4”) covers the movementof some temporary foreign workers among WTOmembers, specifically individual service suppliers. Al-though in theory GATS covers services suppliers at allskill levels, in practice WTO members’ commitmentshave been limited to the higher skilled.

21. Some agreements include provisions facilitat-ing mutual recognition (e.g., EFTA), and others havecomplementary arrangements (e.g., the ANZCERTAServices Protocol, the Trans-Tasman Travel Arrange-ment, and the Trans Tasman Mutual RecognitionArrangement together provide that persons registeredto practice an occupation in one country can practicean equivalent profession in the other country).

22. ANZCERTA does not cover labor mobilityother than for services suppliers, but under the Trans-Tasman Travel Arrangement (signed in 1973), nation-als of each country have the right to visit, reside, andwork in each other’s country without time restrictions.

23. Between 1987–97, there was a total increase of102,000 Greeks in the rest of the (11) European coun-tries, which is an annual average of 10,000 only. Inthe case of Portugal, the annual average increase ofPortuguese immigrants in the rest of the EU countriesduring 1986–97 was only 7,700.

24. In NAFTA, definitions for business personsand, in particular, a list of professionals are provided,as well as the minimum academic conditions that thelatter must satisfy (in general, at least a baccalaureatedegree, and sometimes complemented with someyears of experience). The other agreements providefor a more flexible requirement, in that instead oflisting the professions, they specify the academicand/or experience requirements that professionalsmust satisfy.

25. Business visitors need to demonstrate thoughthat the proposed activity is international in scope andthat they are not seeking to enter the local labor mar-ket. They need also to demonstrate that the primarysource of remuneration, their principal place of busi-ness, and actual place of accrual of profits are fromoutside the territory they are seeking to enter.

26. Mexicans entering through this category intothe United States increased to 4,000 business personsin 2002, from 980 in 1996.

27. Nonetheless, migration flows within COMESAhave been substantial. The annual inflow of workers toSouth Africa from other African countries is estimatedto have increased from roughly 500,000 in 1990 tomore than 3.5 millions in 1995; most were temporaryworkers in mining and farming.

ReferencesAbbott, Frederick M. 2004a. The Doha Declaration on

the TRIPS Agreement and Public Health and theContradictory Trend in Bilateral and RegionalFree Trade Agreements. Quaker United NationsOffice Occasional Paper 14.

. 2004b. Towards a New Era of ObjectiveAssessment in the Field of TRIPS and VariableGeometry for the Preservation of Multilateralism.Paper for The World Trade Forum 2004, Bern,Switzerland, June 4–5, 2004. (also forthcoming inJournal of International Economic Law No. 12005 Oxford.)

Bora, Bijit. 2003. The Agreement on Trade RelatedInvestment Measures, 1995–2002. UNCTADDiscussion Paper Series No. 15, Geneva.

Boeri, Tito, Gordon Hanson, and Barry McCormick,eds. 2002. Immigration policy and the welfaresystem. New York: Oxford University Press.

Chudnovsky, Daniel, and Andres Lopez. 2001. LaInversión Extranjera Directa en el MERCOSUR:Un Análisis Comparativo. In El Boom de Inver-sión Extranjera Directa en el MERCOSUR, SigloXXI, Buenos Aires 2001.

Commission on Intellectual Property Rights (CIPR).2002. Integrating Intellectual Property Rights andDevelopment Policy. Report of the Commissionon Intellectual Property Rights established bythe UK Secretary of State for InternationalDevelopment.

Encarnation, Dennis, and Louis T. Wells. 1986. Evalu-ating Foreign Investment. In Investing in Devel-opment: New Roles for Private Capital? ed.Theodore Moran. New Brunswick, NJ: Transac-tion Books.

Ewing-Chow, Michael. 2001. Investor Protection inFree Trade Agreements: Lessons from NorthAmerica. Singapore Journal of International andComparative Law 748.

Finger, Michael, and Julio Nogues. 2002. The Unbal-anced Uruguay Round Outcome: The New Areasin Future WTO Negotiations. The WorldEconomy 25.


121

Finger, Michael J., and Philip Schuler. 2004. Poor Peo-ple’s Knowledge: Promoting Intellectual PropertyRights in Developing Countries. Washington,DC: World Bank.

Fink, Carsten. 2004. Intellectual Property Rights andU.S. and German International Transactions inManufacturing Industries. In Intellectual Propertyand Development: Lessons from Recent Research,eds. Fink, Carsten and Keith Maskus. Washington,DC: World Bank, forthcoming.

Fink, Carsten, and Keith Maskus, eds. 2004. Intellec-tual Property and Development: Lessons fromRecent Research. Washington, DC: World Bank,forthcoming.

Fink, Carsten, and Patrick Reichenmiller. 2004.Tightening TRIPS: the Intellectual Property Provi-sions of Recent US Free Trade Agreements.Report. World Bank.

Fink, Carsten, and Aaditya Mattoo. 2002. RegionalAgreements and Trade in Services. Policy Re-search Working Paper 2852, World Bank,Washington, DC.

Frischtak, Claudio. 2004. Multinational Firms’Responses to Integration of Latin AmericanMarkets. Business and Politics 6:1.

Hallward-Driemeier, Mary. 2003. Do Bilateral Invest-ment Treaties Attract Foreign Direct Investment?Only a Bit . . . and They Could Bite. PolicyResearch Working Paper 3121, World Bank,Washington, DC.

Hanson, Gordon H., and Antonio Spilimbergo. 1999.Illegal Immigration, Border Enforcement andRelative Wages: Evidence for Apprehensions atthe U.S.-Mexican Border. American EconomicReview 89 (5): 1337–57.

Immigration and Naturalization Service (INS). 1997.Statistical Yearbook of the Immigration andNaturalization Service. Table 40. http://uscis.gov/graphics/shared/aboutus/statistics/Nim97list.htm.

International Organization for Migration (IOM).2003. World Migration 2003: Managing Migra-tion Challenges and Responses for People on theMove. Geneva.

Lall, Sanjaya, and Paul Streeten. 1977. Foreign Invest-ment, Transnationals, and Developing Countries.Boulder, Colorado: Westview Press.

Lederman, Daniel, William Maloney, and Luis Serven.2004. Lessons from NAFTA for Latin Americanand Caribbean Countries: A Summary of Re-search Findings. World Bank, Washington, DC.

Levy Yeyati, Eduardo, Ernesto Stein, and ChristianDaude. 2004. The FTAA and the Location of FDI.Paper for Central Bank of Chile and World BankConference on The Future of Trade Liberalizationin the Americas, Santiago, Chile, March 22–23.

Mann, Howard, and Aaron Cosbey. 2004. InternationalInvestment Agreements: Trends and Impacts forDeveloping Countries Background paper forGlobal Economic Prospects 2005, World Bank,Washington, DC.

Mansfield, Edwin. 1995. Intellectual Property Protec-tion, Direct Investment and Technology Transfer.IFC Discussion Paper No. 27, World Bank,Washington, DC.

Markusen, James. 2004. Regional Integration andThird-Country Inward Investment. Business andPolitics 6 (1).

Martin, Philip. 2003. Economic Integration andMigration: The Mexico-US Case. Wider Discus-sion Paper No. 2003/35. United Nations Univer-sity World Institute for Development EconomicsResearch. Helsinki.

Maskus, Keith. 2000. Intellectual Property Rights inthe Global Economy. Washington, DC: Institutefor International Economics.

Maskus Keith, Sean Dougherty, and Andrew Mertha.2004. Intellectual Property Rights and EconomicDevelopment in China. In Intellectual Propertyand Development: Lessons from RecentResearch, eds. Fink, Carsten, and Keith Maskus.Washington, DC: World Bank, forthcoming.

Mattoo, Aadittya, R. Rathindran, and A. Subramanian.2001. Measuring Services Trade Liberalization andits Impact on Economic Growth: An Illustration.Policy Research Working Paper 2655, WorldBank, Washington, DC.

Mattoo, Aadittya, and Pierre Sauve. 2004. Regional-ism and Trade in Services in the Western Hemi-sphere: A Policy Agenda. In Integrating theAmericas, eds. A. Estevadeodal, Dani Rodrik,Alan Taylor, and Andres Velasco. Cambridge,MA: Harvard University Press.

Newfarmer, Richard. 2001. Foreign Direct Investment:Policies and Institutions for Growth. In NewHorizons for Foreign Direct Investment. Paris:OECD.

Nielson, Julia. 2003. Labor Mobility in Regional TradeAgreement. In Moving People to Deliver Services,eds. Mattoo, Aaditya, and Antonia Carzaniga.Washington, DC: World Bank.

Nofal, Beatriz. 2004. Construyendo una intergracionmas profunda en Mercosur. Background paperfor the Global Economic Prospects 2005, WorldBank, Washington, DC.

Nogues, Julio. 2004. The Roles of Multilateralism andRegionalism for Latin America. Paper preparedfor the Economic Commission for Latin America(ECLA). Santiago, Chile.

OECD. Trends in International Migration, variousyears.


122

OECD. 2003. Regionalism and the MultilateralTrading System. Paris: OECD.

Ozden, Caglar, and Francisco J. Parodi. 2004. CustomUnions and Foreign Investment: Theory and Evi-dence from Mercosur’s Auto Industry. Paper forCentral Bank of Chile and World BankConference on The Future of Trade Liberalizationin the Americas, Santiago, Chile, March 22–23.

Peterson, Luke. 2003a. Emerging Bilateral InvestmentTreaty Arbitration and Sustainable DevelopmentResearch Note, Invest-SD News Bulletin, Interna-tional Institute for Sustainable Development(IISD). April. http://www.iisd.org/pdf/2003/trade_bits_disputes.pdf.

. 2003b. Czech Republic Hit With MassiveCompensation Bill in Investment Treaty Dispute.Invest-SD News Bulletin, International Institutefor Sustainable Development (IISD). March 21.

Schiff, Maurice, and Alan Winters. 2003. RegionalIntegration and Development. Washington, DC:World Bank.

Smarzynska Javorcik, Beata. 2004. The Compositionof Foreign Direct Investment and Protection ofIntellectual Property Rights: Evidence fromTransition Economies. In Intellectual Propertyand Development: Lessons from Recent Re-search, eds. Fink, Carsten, and Keith Maskus.Washington, DC: World Bank, forthcoming.

Stein, Ernesto, and Christian Daude. 2001. Institu-tions, Integration, and the Location of ForeignDirect Investment. In New Horizons for ForeignDirect Investment. Paris: OECD.

Szepesi, Stefan. 2004a. Coercion or Engagement?Economics and Institutions in ACP-EU TradeNegotiations. European Center for DevelopmentPolicy Management (ECDPM) DiscussionPaper 56. Maastricht, Netherlands.

Szepesi, Stefan. 2004b. How Did David Prepare to Talkto Goliath? South Africa’s experience of tradenegotiations with the EU. European Center forDevelopment Policy Management (ECDPM)Discussion Paper 53. Maastricht, Netherlands.

te Velde, Dirk Willem, and Miatta Fahnbulleh. 2003.Investment Related Provisions in Regional TradeAgreements. Paper prepared for the project on Re-gional Integration and Poverty. UK Departmentfor International Development (DFID). London,October.

Ullrich, Heidi. 2004. Comparing EU Free Trade Agree-ments: Services. ECDPM In Brief.

UNCTAD (United Nations Conference for Trade andDevelopment). 1998. Bilateral Investment Treatiesin the Mid-1990s. New York: United Nations.

US Citizens and Immigration Services (USCIS, formerlythe INS). 2002. Yearbook of Immigration Statis-tics (formerly Statistical Yearbook of the Immi-gration and Naturalization Service, INS). Table 27.http://uscis.gov/graphics/shared/aboutus/statistics/TEMP02yrbk/Temp2002list.htm.

US Citizens and Immigration Services (USCIS, formerlythe INS). 2003. Estimates of the UnauthorizedImmigrant Population Residing in the UnitedStates: 1990 to 2000. Office of Policy and Plan-ning. Table B. http://uscis.gov/graphics/shared/aboutus/statistics/Illegals.htm.

USTR (United States Trade Representative). 2004.USA—Chile Free Trade Agreement.

Vivas-Eugui, David 2003. Regional and bilateral agree-ments and a TRIPS-plus world: the Free TradeArea of the Americas (FTAA). TRIPS IssuePapers 1. Geneva: Quaker United Nations Office.

Waldkirch, A. 2001. The ‘New Regionalism’ andForeign Direct Investment: The Case of Mexico.Oregon State University Working Paper.

Walmsley, Terri Louise, and L. Alan Winters. 2003. Re-laxing the Restrictions on the Temporary Move-ment of Natural Persons: A Simulation Analysis.Centre for Economic Policy Research DiscussionPaper Series No. 3719. January.

Weintraub, Sidney. 2004. Lessons from the Chile andSingapore Free Trade Agreements. In Free TradeAgreements: US Strategies and Priorities, ed. J. J.Schott. Washington, DC: Institute for Interna-tional Economics.

World Bank. 2001. Global Economic Prospects andthe Developing Countries 2002: Making TradeWork for the World’s Poor. Washington, DC:World Bank.

. 2002. Global Economic Prospects and theDeveloping Countries 2003: Investing to UnlockGlobal Opportunities. Washington, DC: WorldBank.

. 2004. World Development Report 2005:A Better Investment Climate for Everyone.Washington, DC: World Bank.


123

125

6

The emergence of a more proactive stance inmultilateral negotiations by developing coun-tries in parallel with an explosion of preferentialregional trade agreements (RTAs) around theworld raises an important question: Are thesetrends complementary, representing differentpaths to the same desired outcome of fastergrowth, development, and poverty reduction,or are they competing and incompatible? Inprinciple, the best outcome for all countrieswould be a nondiscriminatory trading system;developing countries, in particular, would ben-efit from a nondiscriminatory trading systembecause most poor people and many poor coun-tries might find themselves excluded from pref-erential deals. If the explosion in RTAs impliesa higher probability that the majority ofdeveloping countries would face greater dis-crimination than under an MFN (or nondis-criminatory) regime, the world as a whole willbe worse off, and individual developing coun-tries may lose substantially.1

RTAs can be a complement to multilateralreform, but they are not a substitute. Considera scenario, modeled in the first section of thischapter, in which each developing countrysigned an agreement with the United States,the European Union (EU), Canada, andJapan—the Quad. Each developing countrycould raise its real incomes by individuallysigning bilateral agreements with the Quadcountries; and in some cases they would raisetheir real incomes more than they wouldthrough multilateral accords. But these gains

disappear if all developing countries were tosign such agreements. In fact, all developingcountries would lose relative to a multilateralagreement and even relative to the baseline.This scenario underscores the fact that, whilestalled collective action through multilateralchannels creates a strong incentive for eachdeveloping country to sign a regional agree-ment, if every country does so, they all lose.

RTAs do alter the incentives for countriesto participate in multilateral liberalization.RTAs can be a stumbling block to multilateralarrangements by creating incentives to resistthe preference erosion that can occur throughnew multilateral liberalization. However,because the gains are often substantially largerin multilateral agreements, concerns over pref-erence erosion may be limited to a few smallcountries that could conceivably block a mul-tilateral agreement. Those recalcitrant coun-tries resisting reforms are likely the beneficia-ries of preferences associated with distortedagreements encompassing agriculture or theclothing and textile trade, not RTAs. Large de-veloped countries may gain more from signingindividual bilateral agreements than theywould from a multilateral accord, becausethey can use the carrot of preferential access toextract concessions in nontrade areas fromdeveloping country partners that would be re-sisted in the WTO negotiating framework. Butwe see little evidence that the high-incomecountries have reduced their effort to bring thecurrent multilateral negotiations to fruition.

Making RegionalismComplementary to Multilateralism

From a development perspective, the WTOremains the best-available forum to disciplinethe use of trade-distorting policies. RTAs cancomplement the WTO efforts by cooperatingon behind-the-border policies, especially onregulation-intensive issues such as services,trade facilitation, and the investment climate.Governments pursuing this agenda throughRTAs must adopt rules that are appropriate totheir own level of development. To be effec-tive, the rules must target a priority concern,must be enforceable, and must avoid becom-ing a tax on scarce resources that would bebetter used elsewhere. Getting the rules“right” so that they promote developmenthas implications for negotiations and enforce-ment of the resulting disciplines. The potentialfor inappropriate outcomes is higher inNorth–South RTAs because the asymmetry innegotiating power can overtake real develop-ment priorities.

Reinforcing the complementarity betweenregionalism and multilateralism and minimiz-ing the latent tensions must begin with thecompletion of the Doha Development Agenda.If the Doha Development Agenda is completedin a way that actually promotes development,it would bring down tariffs, enhance the gainsfrom open regionalism, and discipline any ex-clusionary effects of regional accords.

All countries could take steps to promoteopen regionalism—the developing countries,high-income countries, and the internationalcommunity working together through theWTO. Developing countries are likely to havethe greatest success in harnessing trade forgrowth and poverty reduction if they adopta three-pronged strategy that involvesautonomous liberalization, active multilateral-ism, and open regionalism.

High-income countries could promoteopen regionalism by including agriculture inRTAs. They could adopt more common andnonrestrictive rules of origin across agree-ments; and, to the extent that these rules setpatterns common to most agreements, theburden on customs administration would bereduced. They could work with prospective

partners to ensure that new regulations re-garding investment and intellectual propertyare appropriate to the level of development,which would reduce risks of undue enforce-ment costs. Finally, they could provide trade-related technical assistance not only in the im-plementation phase but also in the negotiatingphase, which could promote greater liberaliza-tion and supply response to new marketopportunities in regional or global markets.

The international community, workingthrough the WTO, can reduce discrimination inthe system. The procedures associated withRTA disciplines as currently configured are ill-suited to limit either their proliferation or tocontrol their discriminatory provisions. WTOmembers should establish stronger multilateralsurveillance mechanisms to document, analyze,and monitor the effects of RTAs on nonmem-bers. Expanding the information on the impactof RTAs to stakeholders—firms, consumers,taxpayers—would also help ensure that the po-tential benefits of liberalization are both real-ized and distributed more equitably. Medium-term efforts should focus on implementingWTO disciplines on regional agreements.

Preferential Agreements within the Global Context

To place preferential trade arrangements ina multilateral context, we evaluate how

different collections of trade agreements com-pare with multilateral alternatives. For thisevaluation, we utilize the World Bank’s globaltrade model known as LINKAGE, which hasbeen used in previous Global EconomicProspects. The model is built around theGTAP database that is widely used to assessthe global, regional, and country implicationsof alternative trade liberalization scenarios.However, the results described in this year’sreport reflect an update of the model’s data-base, which has two notable differences. First,it has a new base year, 2001 (the old base yearwas 1997). Second, it has a new protectiondatabase that takes better account of preferen-tial trade access. Box 6.1 provides a brief


126

M A K I N G R E G I O N A L I S M C O M P L E M E N T A R Y T O M U L T I L A T E R A L I S M

127

The simulation results from the World Bank’sLINKAGE model were presented in the past three

issues of Global Economic Prospects (2002, 2003,and 2004) and were based on the use of variousversions of Release 5 of the GTAP dataset.* An up-dated version of the data has become available; pre-release 6.04 was released in September 2004. Themain innovations of the new dataset include a morerecent base year (2001 instead of 1997), revised na-tional input-output tables, and a new database forestimating the levels of trade protection. This lastinnovation is likely to have a large impact on tradescenarios. First, a brief digression on a technicalissue regarding the new base year: The global data-base, to a large extent, is the outcome of mergingnational data. The national input-output tables havedifferent base years and are virtually always innational currencies. Thus they must be updated tothe given base year—2001 in the case of the newrelease—and converted to the database’s commoncurrency (i.e., the U.S. dollar). This has a practicalimplication because of movements in the value of thedollar. In 1997, the U.S. dollar was relatively strong(for example, $1.13 per euro), and it was muchweaker in 2001 ($0.90 per euro). This means thatthe relative weight of countries will change betweenthe two base years (holding growth constant). Infact, the global U.S. economy as a share of globaloutput was around 32 percent in 1997 and only27 percent in 2001. Thus the change in the value ofthe dollar will have some impact on the reportedgains from trade reform, both globally and by region.

The more important change in the database relatesto the change in protection. The new protection datarelies on the MAcMaps dataset—a collaborative ef-fort of CEPII (Paris) and the International Trade Cen-tre (ITC, Geneva). Among the prominent features ofthe MAcMaps dataset is the incorporation of prefer-ential tariff regimes and the conversion of specifictariffs to ad valorem equivalents; it thus represents amore realistic picture of the bilateral levels of protec-tion. In summary, the new database will capture the

Box 6.1 Impacts of the new GTAP databaseconsiderable reform between 1997 and 2001 (e.g.,continued implementation of the Uruguay Round andChina’s progress towards WTO accession), and animproved treatment of preferential trade agreements.

The overall impact of these changes is that theWorld Bank’s estimate of the global gains fromglobal merchandise trade reform is now around$260 billion** (in 2015, relative to the baselinescenario), compared with around $380 billion usingthe 1997-based results from previous GEPs.*** Thelower number reflects, to a large extent, the impactsof trade reforms achieved between 1997 and 2001and the incorporation of preferences.**** Theallocation of the gains across regions and sectors isbroadly consistent with the previous results. Thus41 percent (instead of 45 percent) of the gains fromglobal reform accrue to developing countries, andagricultural reform generates some 47 percent(instead of 58 percent) of the global gains.

*The GTAP dataset is a product of the Global Trade Analy-sis Project (GTAP), based at Purdue University, with fundingfrom an international consortium of international and nationalagencies (including the World Bank), universities, and researchcenters. Since its initial development in the early 1990s, theGTAP dataset has become the premiere dataset for undertakingglobal trade analysis. The current version includes a full socialaccounting matrix for 87 regions (of which 69 are individualcountries), 57 economic sectors, and fully consistent bilateraltrade flows.

**These results should still be viewed as preliminary as theGTAP consortium is preparing for the final release sometime be-fore the end of 2004. The final release may result in somechanges at the micro level, but will probably only have a rela-tively minor impact at the aggregate level.

***These gains refer to the so-called “static” effects, i.e., nottaking into account dynamic effects such as improvements inproductivity.

****There are other technical differences, including amongothers, the relative change of the value of the U.S. dollar as men-tioned above. The LINKAGE model, apart from the change in thedatabase, is identical to that used in the last Global EconomicProspects.

summary of these changes and the impacts ofusing the new database relative to the resultsoutlined in previous Global EconomicProspects.

To assess the relative impacts of variousRTAs, it is useful to develop a benchmarksimulation (apart from the baseline). Thebenchmark simulation is a global reform

scenario in which all merchandise trade dis-tortions are eliminated (services reform is leftout for lack of sufficient data), domestic dis-tortions in agriculture are removed (input andoutput subsidies, direct payments, and exportsubsidies), and import quotas in the textileand clothing sectors are removed. This sce-nario would be the ultimate long-run outcomeof successful multilateralism. Under this re-form scenario, the global gains in 2015amount to $263 billion, or an increase of 0.8percent in baseline income (table 6.1).2

How much do regional trade agreementsbenefit developing countries? To examine how bilateral agreements affectdeveloping countries, we look at three simula-tions: One in which all developing countriessign a bilateral agreement with Quad-pluscountries (United States, EU, Japan, Canada,plus Australia and New Zealand);3 a secondsimulation—similar to the first simulation—inwhich the large countries, such as Brazil,China, and India are excluded, which is per-haps a more plausible scenario; and a third inwhich each developing country/region signsan individual bilateral agreement with theQuad-plus countries, assuming other develop-ing countries do not sign agreements.4 Notethat these scenarios overstate bilateral andmultilateral effects because they assume thatno sectors are exempt, and rules of origin arenot restrictive. In reality, the United States andEU bilateral agreements usually exclude im-portant sectors, such as sensitive agriculturalproducts, or attach extended phase-in periods

beyond even our 2015 time horizon, and rulesof origin tests often limit preferential marketaccess.

The first simulation, in which all develop-ing countries sign a bilateral agreement withthe Quad-plus countries (columns 2 and 5 intable 6.1), shows that, as a group, developingcountries are substantially worse off than witha multilateral agreement. Instead of gaining$109 billion from global reform, they lose$22 billion relative to a baseline scenario, withno change in protection. If one looks at indi-vidual countries (table 6.2), the effect is nearlyuniversal; only a handful of developing coun-tries—for example, Brazil and China—wouldgain from a full hub-and-spoke system of bi-lateral agreements, and all developing coun-tries would lose compared to full multilateraltrade.

It is interesting that some of the Quadcountries would benefit from this strategy. Al-though the high-income countries would gen-erally lose from this set of bilateral agreementscompared to global reform, the impact is notuniform.5 Both the United States and the EU(the most aggressive advocates of bilateraldeals) would appear to benefit more from pur-suing bilateral agreements with all develop-ing countries than from global reform(table 6.2); the United States would gain anadditional $7 billion (0.1 percent of GDP),while the EU would gain $27 billion (0.4 per-cent of GDP). Although they would have toopen up their agricultural markets to some ex-tent (assuming exemptions are disallowed),they would not have to dismantle domestic


128

Table 6.1 Comparison of bilateral agreements to global trade reform(change in real income in 2015 compared to baseline)

Bilateral with Bilateral minus Bilateral with Bilateral minusGlobal Quad large countries Global Quad large countries

(1) (2) (3) (4) (5) (6)

$ billion Percent

High-income countries 154.4 133.6 46.9 0.6 0.5 0.2Low-income countries 16.6 �19.0 �1.9 0.9 �1.0 �0.1Middle-income countries 92.2 �2.6 �4.7 1.2 0.0 �0.1All developing countries 108.8 �21.5 �6.6 1.2 �0.2 �0.1World Total 263.2 112.0 40.3 0.8 0.3 0.1


129

Tab

le 6

.2C

om

par

iso

n o

f b

ilate

ral a

gre

emen

ts w

ith

glo

bal

tra

de

refo

rm(c

han

ge

in r

eal i

nco

me

in 2

015

com

par

ed t

o b

asel

ine)

Bila

tera

l agr

ee-

Bila

tera

l agr

eem

ents

B

ilate

ral a

gree

- B

ilate

ral a

gree

men

tsB

ilate

ral a

gree

-m

ents

bet

wee

n be

twee

n Q

uad

and

Bila

tera

l agr

eem

ents

men

ts b

etw

een

betw

een

Qua

d an

d m

ents

bet

wee

nG

loba

l mer

-Q

uad

and

all

deve

lopi

ng c

ount

ries

be

twee

n Q

uad

and

Glo

bal

Qua

d an

d al

lde

velo

ping

cou

ntri

es

Qua

d an

d ea

chch

andi

se t

rade

de

velo

ping

ex

clud

ing

larg

e ea

ch d

evel

opin

g m

erch

andi

sede

velo

ping

excl

udin

g la

rge

deve

lopi

ng

refo

rmco

untr

ies

coun

trie

sco

untr

ytr

ade

refo

rmco

untr

ies

coun

trie

sco

untr

y(1

)(2

)(3

)(4

)(5

)(6

)(7

)(8

)

$ bi

llion

Perc

ent

Aus

tral

ia, C

anad

a, a

nd

8.4

6.0

0.5

0.7

0.5

0.0

New

Zea

land

Uni

ted

Stat

es24

.932

.310

.70.

20.

30.

1E

urop

ean

Uni

on w

ith

EFT

A55

.082

.433

.60.

71.

10.

4Ja

pan

29.7

25.0

4.8

0.9

0.8

0.1

Rep

ublic

of

Kor

ea a

nd

26.4

�9.

8�

2.3

2.6

�1.

0�

0.2

Taiw

an (

Chi

na)

Hon

g K

ong

(Chi

na)

and

9.8

�2.

4�

0.3

2.8

�0.

7�

0.1

Sing

apor

eB

razi

l8.

01.

5�

1.7

7.3

1.4

0.3

�0.

31.

3C

hina

14.1

9.7

�7.

221

.80.

60.

4�

0.3

1.0

Indi

a4.

3�

10.0

�3.

12.

10.

5�

1.2

�0.

40.

2In

done

sia

3.6

�2.

33.

05.

11.

4�

0.9

1.2

2.1

Mex

ico

0.3

�1.

5�

1.3

2.6

0.0

�0.

2�

0.2

0.3

Rus

sia

2.9

�1.

7�

1.3

0.8

0.8

�0.

5�

0.3

0.2

SAC

U2.

5�

0.3

0.8

3.7

1.8

�0.

20.

52.

6V

ietn

am2.

4�

0.2

0.6

0.9

5.0

�0.

51.

31.

9R

est

of E

ast

Asi

a19

.6�

5.0

�2.

87.

44.

7�

1.2

�0.

71.

8R

est

of S

outh

Asi

a0.

4�

3.2

�1.

11.

20.

2�

1.3

�0.

40.

5E

U a

cces

sion

cou

ntri

es0.

8�

2.0

�0.

50.

90.

2�

0.4

�0.

10.

2R

est

of E

CA

2.3

�3.

3�

1.3

0.4

0.5

�0.

7�

0.3

0.1

Mid

dle

Eas

t6.

1�

2.7

�0.

11.

30.

9�

0.4

0.0

0.2

Nor

th A

fric

a19

.11.

94.

35.

76.

70.

61.

52.

0R

est

of S

ub-S

ahar

an A

fric

a2.

9�

3.0

�2.

5–0

.21.

1�

1.2

�1.

0�

0.1

Res

t of

LA

C16

.30.

96.

49.

61.

60.

10.

60.

9R

est

of t

he W

orld

3.0

�0.

31.

24.

01.

3�

0.1

0.6

1.8

Hig

h-in

com

e co

untr

ies

154.

413

3.6

46.9

0.6

0.5

0.2

Low

-inc

ome

coun

trie

s16

.6�

19.0

�1.

90.

9�

1.0

�0.

1M

iddl

e-in

com

e co

untr

ies

92.2

�2.

6�

4.7

1.2

0.0

�0.

1A

ll de

velo

ping

cou

ntri

es10

8.8

�21

.5�

6.6

1.2

�0.

2�

0.1

Wor

ld T

otal

263.

211

2.0

40.3

0.8

0.3

0.1

Sour

ce:

Wor

ld B

ank

sim

ulat

ions

.

support programs. Hence the EU and theUnited States improve market access in highlyprotected markets in developing countries, butthey do not face the full force of competitionbetween themselves, particularly in agricul-ture, nor the full brunt of competition in de-veloping countries. For example, when Indiaopens up to Quad country imports, Quadcountry exporters do not simultaneously faceincreased competition from developing coun-try exporters as they would in a multilateralagreement. In agriculture, Quad producerswill face greater competition from relativelyefficient developing country exporters, butsome of the fiercest competition will be amongthemselves. And the terms of trade losses thatthey would suffer from removing agriculturalprotection between Quad countries is mutedwhen signing the bilateral agreements. Notethat this is not the case with Japan, whoseagriculture is relatively more threatened bydeveloping country market access; it wouldgain more from a multilateral agreement, al-though it nonetheless gains significantly frombilateralism. It is the Quad-plus agriculturalexporters—Australia, Canada, and NewZealand—that would prefer multilateralism,because the gains from access to European,Japanese, and the American markets, and thedismantling of distortionary agricultural sup-port programs would be highly beneficial fortheir farmers.

Were the large developing countries6 to beexcluded from the hub-and-spoke bilateralarrangements (perhaps a more plausible sce-nario), the broad conclusions still hold, butare muted (columns 3 and 6). First, many de-veloping regions still lose in absolute termscompared with the baseline scenario. Second,all lose relative to the gains from a global re-form scenario, though for some, the hub-and-spoke gains approach those in the globalreform scenario (e.g., Indonesia, and to alesser extent, the rest of LAC and rest of theworld regions). The gains for the high-incomecountries, on the other hand, are significantlylower when the large developing countries are

excluded—not surprising given their weight inglobal trade with the Quad countries. Finally,the impacts on the excluded countries aremixed: Brazil and China, which would gain ina full hub-and-spoke system, lose when ex-cluded. The other excluded regions—India,Mexico, Russia, rest of East Asia, and rest ofSouth Asia—would see a dampening of theirlosses.

This adverse outcome from bilateral agree-ments with the Quad raises the question ofwhy developing countries are so anxious topursue them. There are a number of possiblereasons, not necessarily mutually exclusive.First, countries may hope to maximize theirbenefits through first-mover advantages. Sec-ond, countries aim to guarantee market accesson a permanent basis. Third, there might bea desire to pre-empt other countries. Fourth, abilateral agreement may be used as leverageto facilitate domestic reforms. And fifth, othercomponents of an agreement (for example,services, trade facilitation, and so on) mayhave significant benefits in addition to simplyimproving market access for merchandisegoods. Focusing on the first of these possiblereasons, we use the model to simulate the im-pact of each developing country signing anagreement with the Quad countries, but withno other country doing so. The results(table 6.2 columns 4 and 8) provide some jus-tification for developing countries’ pursuit ofRTAs with the Quad if they believe they can doso exclusively, or at least capture a “firstmover” advantage by getting there first. Aboutone-half of the developing regions would bebetter off with a bilateral agreement than witha global agreement; the winners (relative toglobal liberalization) include China, Indonesia,Mexico, Southern African Customs Union(SACU), rest of South Asia, and EU accessioncountries. Losers include Brazil (slightly),India, Russia, Vietnam, rest of East Asia, rest ofECA, Middle East, North Africa, rest of Sub-Saharan Africa, and rest of LAC.

A few cases deserve special mention. Therest of the Sub-Saharan Africa region could


130

suffer losses from a bilateral agreement withthe Quad. Because this region already has rel-atively free access to the Quad markets, open-ing up to permit greater imports from theQuad worsens their terms of trade and negatesany gains from the bilateral agreement. Russiaand the Middle East are dependent on energyexports and these face low tariffs in industrialcountries (even if energy is heavily taxed), sothese regions have little to gain from addi-tional market access. These cases highlightone of the key findings from recent GlobalEconomic Prospects, that developing coun-tries have much to gain from greater marketaccess to other developing countries. First, be-cause protection is, on average, higher in de-veloping countries, and second, because of thehigh growth potential of developing countriesover the next decade compared with the in-dustrial countries.

The idea that a single developing countryor region would be able to sign bilateral agree-ments with the Quad countries without otherdeveloping countries doing the same over thenext decade is unrealistic. Indeed, the increasein the number of agreements over the lastdecade means that a portion of any first-mover advantage has been eroded already. Butas a conceptual exercise, it does help illustratewhat may have motivated some developingcountries to aggressively pursue deals withone or more Quad members.

To summarize, developing countries couldgain an (unweighted) average of 1.7 percent inreal income from a global agreement (fig-ure 6.1). But if all developing countries signbilateral agreements with the Quad, creating acomplex hub-and-spoke system, developingcountries actually suffer losses averaging0.4 percent (1.0 percent for the low-incomecountries alone). While some individual devel-oping countries might have gained from enter-ing exclusive agreements with Quad countries,RTA proliferation has already eliminatedthat first-mover advantage. Moreover, RTA-inducedstructural changes couldproducedisin-centives for achieving broad-based multilateral

reforms (box 6.2). The implications are clear:The most development friendly outcome is as-sociated with global reform, and a full set ofbilateral agreements would leave virtually alldeveloping countries worse off than atpresent.

Countries are also pursuing other RTAs,both North-South and South-South. Theseagreements are linked to the ongoing talks onthe Free Trade Areas of the Americas (FTAA),the ASEAN�3 negotiations, and the EU’s vari-ous agreements/negotiations with the EU-accession countries toward the east, partnersaround the Mediterranean, and in Sub-Saharan Africa toward the south. Chapter 2 ofthis report documents the large number of ex-isting or prospective agreements among devel-oping countries. Figures 6.2 and 6.3 summa-rize the overall impacts from simulatingselected North-South and South-South agree-ments.7 Some of these proffer relatively signifi-cant gains (for example, a broad free trade re-gion in East Asia), but are nonetheless clearlyinferior to the gains from global merchandise


131

�1.5 �1.0 �0.5 0 0.5

Percent

1.0 1.5

Note: Global refers to the global merchandise trade reformscenario, JBIL corresponds to the simulation where alldeveloping countries sign bilateral agreements with theQuad-plus, and BILAT corresponds to the simulation wherethe bilateral agreements are signed individually. Resultsreflect un-weighted regional averages.


2.0

Figure 6.1 Global outcomes dominatealternatives

Quad�

Low-income

Middle-income

Developing

Change in real income in 2015 compared to baseline

Global

BILATJBIL

trade reform. Two additional comments re-garding the results of these regional simulationsare suggested by the figures.8 First, whenNorth-South agreements are implemented si-multaneously, the gains are dampened relativeto when they are implemented in isolation—areflection of the impacts of trade diversion. In

the case of the South-South agreements, theweak gains reflect, in part, the preferences al-ready granted in many of these regions as wellas a lack of distinct comparative advantage.9

This again emphasizes that broad South-Southand North-South trade reform is needed to reapsignificant gains.


132

While impact assessment of RTAs and globalagreements often focus narrowly on the real

income impacts, from a political economy point ofview, the main drivers of these agreements typicallywill be at the micro or institutional level, where it iseasier to identify the potential winners and losers.Moreover, the macro analysis typically ignores thetransitional costs. A final issue deals with the com-patibility of partial reforms (as represented by prefer-ential trade agreements) with multilateral reforms,which is arguably where the world economy is head-ing. In other words, how compatible are the struc-tural changes induced by a partial reform with thestructural changes one would anticipate from a mul-tilateral reform? For example, what if a country suchas Vietnam has a regional comparative advantage inagriculture, but a global comparative advantage inapparel? Would a regional agreement then make itmore difficult to achieve a multilateral accord?

The figure above provides a summary indicatorof the congruence or compatibility of the bilateralagreements—both the individual (BIL) and joint(JBIL). The indicator measures the average structuralchange of moving from the baseline to the partialagreement, and then moving from the partial agree-ment to global free trade, relative to the structuralchange induced by the global agreement. If the par-tial agreement is compatible (or congruent) with theglobal agreement, this measure is 1; that is, the two-step structural change is identical to the one-stepstructural change from implementing directly aglobal free trade agreement. For example, a globalagreement of a 50 percent tariff cut is largely con-gruent with a 100 percent tariff cut and would most

Box 6.2 Regional trade agreements, structural change,and congruence

likely lead to a 50 percent change in our structuraladjustment measure.

For most of the developing regions, the joint bilat-eral agreements are broadly consistent with the globalagreement, with the structural index varying fromaround 1 (or near perfect congruence) to a high ofaround 2—for example, for Middle East and NorthAfrica. On the other hand, the individual bilateralagreements are clearly not congruent with a globalagreement. The pattern of structural change inducedby the individual bilateral agreements is markedly dif-ferent from what one would anticipate with a globalagreement. Developing countries therefore face atradeoff. They can get short-term benefits from signinga bilateral agreement with the Quad—assuming othercountries do not—but the longer-term gains from aglobal agreement may be more difficult to attain be-cause the patterns of capital and labor allocationwould be misaligned by the partial agreement.

Index (1 —> fully congruent)

BraChn In

d Idn

Mex Rus Sac

Vnm Xea Xsa Cec XecM

de Naf Xss Rlc

Joint bilateral vs. bilaterals

0

1

2

6

5

4

JBIL

BILAT

3


Building Blocks versus StumblingBlocks

Whether regional agreements are buildingblocks or stumbling blocks to open

global markets—the terms Jagdish Bhagwati(1991) used—remains a central question.10

Proponents of the stumbling block theory em-phasize that: (1) RTAs may promote costlytrade diversion rather than efficient trade cre-ation, especially when sizable MFN tariffsremain—these tariffs create vested interests tomaintain preferential margins in “their” mar-kets; (2) proliferating regional agreements ab-sorb scarce negotiating resources (especially inpoorer WTO members) and crowd out policy-makers attention; (3) competing RTAs (espe-cially different North-South combinations)may lock in incompatible regulatory structuresand standards, and may result in inappropriatenorms for developing country partners; and(4) by creating alternative legal frameworksand dispute settlement mechanisms, RTAsmay weaken the discipline and efficiency asso-ciated with a broadly recognized multilateralframework of rules.

Building block proponents stress that mov-ing forward in smaller steps is often easier toaccomplish, and it creates a certain reformmomentum: (1) regional/bilateral agreementscan help sensitize domestic constituencies toliberalization and keep the stakes lower toallow for incremental progress on trade; (2)expanding the number and coverage of RTAscan erode vested opposition to multilateralliberalization because each successive RTA re-duces the value of the margin of preference,thereby reducing the discriminatory impact;(3) RTAs are often more about building strate-gic and/or political alliances or locking in do-mestic reforms than about actual trade liberal-ization, and so are not necessarily competitivewith multilateral efforts; (4) regional arrange-ments can provide an incubator for develop-ing country firms/producers to learn to tradewith RTA partners without facing full globalcompetition; and (5) for some issues, such asregulatory cooperation, RTAs may be a viable


133

Percent

Note: Global refers to the global merchandise trade reformscenario, separate is when the North/South regional blocksare formed individually and block is when the North/Southregional blocks are implemented simultaneously.


Change in real income in 2015 relative to the baseline

East Asia Europe The Americas

Figure 6.2 Global reform dominatesNorth–South agreements

0.0

1.0

0.2

0.4

0.6

0.8

1.2

1.4

1.6

Global

Separate

Block

Percent

Note: Global refers to the global merchandise trade reformscenario, and separate is when the South/South regionalblocks are formed in isolation.


Change in real income in 2015 relative to the baseline

LAFTA AFTA SSA MNA SAFTACIS

Figure 6.3 Global reform dominatesSouth–South agreements

�0.5

2.0

0.0

0.5

1.0

1.5

2.5

3.0

Global

Separate

and more manageable alternative to theWTO, where “lowest common denominator”outcomes tend to prevail.

One strand of analysis in the literature ex-plores theoretical bargaining models to shedlight on these divergent perspectives. For ex-ample, Saggi (2004) considers how RTA for-mation affects incentives to participate in mul-tilateral trade liberalization. Using a stylizedthree-country oligopoly model of intra-industry trade, he reasons that while forma-tion of RTAs can reduce tariffs in the shortrun, in the long run it hinders multilateraltrade liberalization because it lowers the in-centives for nonmembers to pursue multilat-eral cooperation. In a three-country model,where country size and costs differ, Saggi andYildiz (2004) look at tariff outcomes undertwo different sets of rules: One where coun-tries can form RTAs, and another where theycannot. They find that free trade is less likelyto occur as an outcome when countries canjoin RTAs as well. Moreover, the outcome de-pends on size and cost features: When coun-tries are relatively similar, the RTA optiontends to lower world welfare; whereas withlarger differences among countries, welfare in-creases. In the latter case, gains are somewhatlarger for the small country getting access tothe larger country (because it gains more innew export earnings and loses less in domesticsurplus) than for the country with lower costs(because gains from expanded market accessincrease when their own costs are lower andpartner costs are higher). Aghion and others(2004) construct a dynamic bargaining modelto assess the relative effects of simultaneousmultilateral liberalization or sequential bilat-eral liberalization, and they identify situationsin which the latter can lead to worldwide freetrade. Their model concludes that, as long aslarge coalition payoffs are higher than if allcountries were combined in alternative coali-tion structures, the outcome is likely to end upin global free trade, regardless of sequencing.

The Saggi findings support the empiricalevidence that North-South RTAs with smalldeveloping countries yield the biggest benefits

for developing country participants. This intu-ition comes from the fact that most empiricalmodels derive their impact from the effect ofRTAs on prices and quantities of tradedgoods. But when one partner is much smallerthan the other, its participation in an RTA hasvirtually no impact on prices, and any incre-mental exports it sells are a tiny share of thelarge country trade flows—so the resultinggains or losses for the larger partner are alsosmall. Support for this conclusion comes fromthe computable general equilibrium model(CGE) simulations on the impact of RTAs; inmost cases, the net impact on the Northernpartner has been exceedingly small, whethermeasured in terms of trade flows, terms oftrade, or welfare changes.

This result in turn has implications for howRTA proliferation affects incentives to pursuemultilateral liberalization. If the impact of ad-ditional North-South RTAs on the industrialpartner is indeed quite marginal, then it seemslikely that the economic consequences of theseagreements would not dampen their willing-ness to pursue multilateral deals. As we seebelow, neither the EU nor the United Statesseem to be less disposed toward multilateralnegotiations because of their RTAs or theRTAs of each other.

Other studies show less sanguine results.Limão (2003) notes that North-South RTAscan involve more than just the mutual loweringof tariffs; industrial countries often lower theirtariffs and expand market access in exchangefor cooperation in a variety of nontrade areas,such as labor standards, intellectual property,migration, security, and so forth. Limão mod-els the interaction between such RTAs andmultilateral liberalization, and concludes thatpursuit of RTAs creates a strategic incentive forindustrial countries to maintain their multilat-eral tariffs at a higher level than they otherwisewould—to hold back tariff concessions in themultilateral arena in order to have bargainingroom for negotiating RTAs. In the case of theUnited States, multilateral trade liberalizationcommitments are less deep for products thatare produced by regional firms.11


134

A more compelling argument is that devel-oping countries with preferential access mayhave a vested interest in perpetuating the tar-iff walls that screen out competing countriesfrom preferred markets. In fact, Barbados,Jamaica, and Mauritius have all been outspo-ken opponents of the agricultural liberaliza-tion that would erode their preferences in EUsugar markets, and Bangladesh has advocatedmeasures that would delay the phaseout of thetextile and apparel quotas required under theWTO Agreement on Textiles and Clothing.Nonetheless, this seems confined to a handfulof relatively small countries heavily dependenton a narrow range of preference-benefitingproducts.

An illuminating example of how prefer-ences can be a stumbling block to multilateralliberalization is provided by rum. Low-valuedbottled and bulk rum is one of the largest ex-ports from several Caribbean countries to theUnited States. It enters the United States duty-free under the Caribbean Basin Initiative. In1994 there was a single tariff line for rum in theU.S. schedule (HS 22084000) with a tariff of37 cents/liter. As result of WTO discussions,the United States and EU in 1996 agreed tophase out all tariffs on rum and other “whitespirits” by 2000. Caribbean governments wereconcerned that this would be costly toCaribbean exporters because they would haveto compete with the rest of the world inthe U.S. market. In response to this concern,the United States agreed to substantially liber-alize the duties on expensive rum but to main-tain duties on low-valued rum. As of 2003, theU.S. schedule now has four tariff lines for rum,two for high-valued rum with no MFN tariffcharged, and two for low-valued rum with anMFN tariff of 23.7 cents/liter.12 Since then, theimports of rum from the affected Caribbeancountries have fallen sharply, apparentlyreplaced by imports from Mexico—another bi-lateral RTA partner. So efforts to prevent ero-sion of preferential access may have only lim-ited the impact temporarily.

An increase in RTA activity may be associ-ated with multilateral trade negotiations.

Mansfield and Reinhardt (2003) argue thatmultilateral trade negotiations motivate coun-tries to conclude RTAs; the motivation is forincreased negotiating power and the desire toobtain or maintain preferential access to mar-kets. They also note that as WTO member-ship expands, larger membership reducesindividual countries’ ability to influence thecontent and pace of MFN liberalization andmakes it more difficult to formulate coordi-nated positions. Finally, they find that coun-tries use RTAs to increase leverage againstthird parties with which they are embroiled ina WTO dispute. In their empirical analysisthey conclude that countries are more likelyto form RTAs when: (1) GATT/WTO mem-bership rises; (2) a multilateral round is tak-ing place; and (3) parties have recently beeninvolved in a GATT/WTO dispute in whichthey lost.

Others have argued that RTAs are a mech-anism to enhance the pressure to move on themultilateral front, and they act as laboratoriesfor international cooperation on behind-the-border policy issues. This line of reasoning hasa long history. Winham (1986) and Lawrence(1991) have both argued that the creation andsubsequent expansion of the European Eco-nomic Community (EEC) motivated earlierGeneral Agreement on Tariffs and Trade(GATT) rounds—the objective being to reduceEuropean external protection. More recently,Schott (2004) noted that the United States haspursued bilateral trade agreements over thelast two decades in part to complement andcajole progress at the multilateral level. Heargues that tensions from the GATT meetingsin 1982 were the impetus to pursue bilateraldeals with Canada and Israel; he also claimsthat the start of the NAFTA negotiations in1991 reflected some frustration over failureto conclude the Uruguay Round negotiationson time in late 1990. In the early 1990s, theUnited States also began to pursue broaderumbrella regional initiatives, including AsiaPacific Economic Cooperation (APEC)—whose members have never formally commit-ted to a binding RTA—and the FTAA.


135

In short, there is only limited evidencefrom the theoretical literature and a handfulof empirical studies that the proliferation ofRTAs affects multilateral liberalization eitherway. On balance, the evidence does point totactical behavior in trade negotiations, whichseems to provide mild additional incentivesfor greater liberalization. The exception isfor small countries that suffer from preferenceerosion—a nontrivial obstacle to futureliberalization.

The Competitive LiberalizationGame: The Case of Doha

The rapid proliferation of RTAs may haveaffected the negotiating dynamics of the

Doha Round, and it is unclear whether the in-crease in bilateral deals is in response to slowprogress in the Doha talks. Concerns over RTAproliferation played a role in the launch of theRound and came back full force after the fail-ure of the Cancun ministerial. Consider theactions of the major players in the global andregional game.

The U.S. pursuit of like-minded partnersWith the change in U.S. administration in2001 and the subsequent congressional ap-proval of the Trade Promotion Authority in2002, the pace of bilateral, regional, and mul-tilateral efforts increased. Then, in the absenceof progress on the Doha agenda in Cancun inSeptember 2003, the U.S. Trade Representa-tive (USTR) indicated that the United Stateswould pursue deals with “like-minded” part-ners, and the result has been an unprecedentedspurt of U.S. negotiating activity. The UnitedStates had negotiated RTAs with sixcountries/groups by 2004 (Australia, Bahrain,Central America and Dominican Republic,Chile, Morocco, Singapore), and anotherdozen are under negotiation, including SACU(Botswana, Lesotho, Namibia, South Africa,and Swaziland), Colombia, Ecuador, Panama,Peru, and Thailand. As discussed in chapters 2and 5, the United States is pursuing new rulesin key areas (investment, intellectual property,

and services liberalization)—areas wheremultilateral efforts have gone slowly or haltedaltogether—while developing countries seekmarket access.

Eligibility is not guided by a single criterionor ranking system (box 6.3). Political criteriainclude national security-cum-foreign policyconcerns (a major factor in the recent agree-ments and ongoing discussions with Arabcountries), while others reflect a mixture ofclassic market access goals and a desire to ex-port U.S. regulatory standards (for instance, inthe area of investment and IPRs—seechapter 5). The U.S. decision to pursue bilat-eral arrangements in Latin America also hasthe effect of putting pressure on Brazil andother countries, which are seen to be impedingprogress on the now-stalled negotiations onthe FTAA. Reaching these agreements doesnot appear to have reduced the United States’participation in the WTO negotiations, nor tohave had much effect on the content of itsnegotiating position.

Recent EU regional initiativesHaving been a leading (and early) player in theRTA game, the EU had established manyRTAs prior to the launch of the Doha Round(see chapter 2). Since then, it has pursued mar-ket access agreements with a few individualpartners (notably Chile and Mexico, withSouth Africa predating 2001) as well asMERCOSUR. The EU has preferred to negoti-ate with blocs of countries and has encouragedthe Mediterranean countries to sign agree-ments with one another; the EU made the RTAwith Gulf Cooperation Council (GCC) coun-tries conditional on the adoption of a commonexternal tariff by the GCC and supported thedevelopment of a web of bilateral RTAs be-tween Southeastern European economies. Therecent Economic Partnership Agreements(EPAs)13 were launched to replace the Cotonouagreement with the African, Caribbean, andPacific (ACP) group in a WTO-compatible setof agreements, and the EPAs are seen by the EUas a development-promoting vehicle ratherthan as a way to gain additional market access.


136

These agreements do not appear to have al-tered the EU’s approach to the Doha Round.After the conclusion of the Mexican andChilean bilateral deals, the EU had declared itwould negotiate no new trade agreements,save for the EPAs that were necessary to re-place the Cotonou agreement set to expire in2008. However, this stance may be short-lived. In September 2004, the incoming EUTrade Commissioner announced that hewould review this policy and consider launch-ing new negotiations.

Japan is a latecomer but moving quicklyUntil recently, Japan remained disengagedfrom the RTA trend and instead pursued avoluntary approach that centered on APEC.A distinguishing feature of APEC is that itfocuses primarily on exchanging informationand identifying good practices in the tradepolicy (and other) areas. While it also sets spe-cific targets for trade policy (e.g., free intra-APEC trade by 2020), implementation of the

good practices and targets are left to individ-ual members. The primary enforcement de-vices are mutual surveillance and peer review.More recently, it appears that competitivepressures that emerged from the more inten-sive activity by the United States—also anAPEC member—and the concern that even inthe Asia region there was a risk of being leftout of the new generation of RTAs, haveprompted new Japanese activity. Japan com-pleted the negotiation of its first RTA withSingapore in 2002. Negotiations with Mexicoare advanced. The competitive disadvantagein Mexican markets relative to the UnitedStates and the EU is a strong inducement, withJapanese products facing an average customsduty of 16 percent (Tojo 2004). Negotiationsare beginning with Republic of Korea, whichare considered important because Japan-Korea could form the basis of a future EastAsian economic zone. Talks have also begunwith three individual members of ASEAN(Thailand, Malaysia, and the Philippines), and


137

The U.S. choices of developing country RTA part-ners reflect a range of different objectives, which

makes categorization difficult. In a report preparedfor the U.S. Congress, U.S. General Accounting Of-fice (GAO) (2004) reports that early RTA proposalswere (according to the USTR) based on evaluation of13 different factors (without any formal weightingscheme) that covered five themes: Congressionalguidance, business interests, political will of partners,security and foreign policy concerns, and regionalparity. Following consultations with relevant agen-cies, there is now a shortened list of six factors toguide the choice of future U.S. FTA partners: countryreadiness (political, trade, and legal); economic/com-mercial benefit; benefits to broader trade liberaliza-tion strategy (including success in meeting WTOobligations and support of key U.S. positions inFTAA and WTO negotiations); compatibility withU.S. interests (including foreign policy positions);

Box 6.3 Choosing partners: Selection criteria for U.S. RTAs

congressional/private sector support; and U.S. gov-ernment resource constraints. The report emphasizesthat the selections are not mechanical and arguesthat trade strategy and foreign policy factors domi-nate the selection criteria.

Schott (2004) identifies the same broad criteria,but also notes the role of partner choice in selectionof U.S. FTA partners. Current U.S. law requires thatpotential partners must request negotiations with theUnited States, rather than the other way around. Henotes that most of these requests reflect concernsover discriminatory treatment (from other agree-ments, especially NAFTA) and serve to demonstratecommitment to domestic audiences who may need tobe convinced of the benefits of reform trade anddomestic policies.

Source: Schott 2004.

consultations have been initiated over how tomove forward with a proposed Comprehen-sive Economic Partnership between Japan andthe full ASEAN group. Finally, there are pre-liminary discussions and analysis of a possibleASEAN�3 RTA (including China, Republicof Korea, and Japan) in the context of effortsto strengthen economic relations among thisgroup.

Developing country objectives andnegotiating strategiesDeveloping countries have three general ap-proaches to RTAs. Some have adopted an ag-gressive approach and pursued a serial RTAstrategy; that is, they negotiate a string ofagreements and use the sequence to demon-strate their commitment to trade reforms bylocking these in and increasing the incentivefor excluded countries to negotiate. Harrison,Rutherford, and Tarr (2002) have labeled thisnegotiating dynamic “additive regionalism.”Its most prolific proponents include Chile,Mexico, and Singapore, which have pursuedRTAs with most of their geographic neigh-bors, as well as with many of the other majorplayers (Schott 2004). The idea is that negoti-ating additional RTAs will progressively lowerthe effective average tariff (reducing potentialtrade diversion costs) and assure stability ofmarket access for partner countries. A second,much larger group of countries has pursued astrategy more explicitly regional in focus,which seeks to deepen ties with neighboringcountries; examples include ASEAN, theGCC, MERCOSUR, and SADC. A thirdgroup has focused on negotiating North-South RTAs, often in parallel with their re-gional integration efforts with neighbors;examples are the southern Mediterraneancountries with the EU and the US-CAFTAagreement. The ACP-EU Economic Partner-ship Agreements are another example.

Additive regionalism could create adverseglobal effects by reducing the incentives forcountries to participate in multilateral liberal-ization efforts. That has not been the case withChile, Mexico, or Singapore. The fact remains

that these countries still have incentives to seelower barriers in the many countries withwhich they do not have RTAs and to harnessthe multilateral process to achieve movementin areas outside of the RTAs, such as agricul-ture and anti-dumping. The WTO provides aforum to achieve these objectives at a lowercost than negotiating a plethora of bilateralRTAs. Countries pursuing such strategies arevery much a minority—most developing coun-tries are not in this game (box 6.4). Chile,Mexico, and Singapore are all economies thatmoved substantially toward free trade andthus have already captured most of the poten-tial gains from unilateral trade reforms. Froma global point of view, therefore, these are notcountries that have a vested interest in main-taining high MFN barriers. These countriesare examples of open regionalism.

The determinants of the multilateral nego-tiating stance of the broader group of devel-oping countries are more varied. For a hand-ful of countries, existing preferences underunilateral accords as well as multilateral andregional agreements are openly driving theiropposition to multilateral liberalization. Forsome other developing countries, it is possiblethat RTAs are undermining their interest inmultilateral negotiation, either because theymistakenly see RTAs as an alternative, becausethey feel subsumed in large coalitions withother countries, or because they do not havethe resources to negotiate with both the WTOand with potential regional partners.

This discussion suggests a few impression-istic conclusions. It is hard to argue that com-petitive liberalization through RTAs has muchinfluence on the behavior of the major playersin the WTO, either in their fundamental nego-tiating positions or their willingness to com-promise to achieve a result. As evident in theJuly WTO Framework Agreement, majorplayers are still working with commitmenttoward an agreement in Doha. To be sure, forsmaller countries with scarce negotiating ca-pacity, RTAs do absorb resources that couldbe devoted to multilateral negotiations; butthese countries tend to participate in the WTO


138

as members of coalitions, and it is not clearthat these constraints impede compromise. Ifthe effects on the multilateral round are negli-gible, it does seem that all players appear to bequickening their efforts to seek new preferen-tial agreements. The outcome of the game ofRTA-based competitive liberalization is likelymore RTAs.

One negative incentive effect created bypreferences—reciprocal RTAs or voluntaryprograms—is that members’ desire to safe-guard their preferential access to the regionalmarket may result in less support for MFN-based trade reforms. This incentive effect haslong been recognized; it was one of the argu-ments made against the Generalized System of


139

Debates over trade liberalization have often in-cluded extensive discussions of whether there is

a preferred (or even optimal) sequence for reforms.Theoretical guidance for policymakers in the designof RTAs speaks to two dimensions of the question—the choice (and sequence) of partners over timeand the substantive coverage of an RTA (and thesequencing of inclusion of different elements)—thathave implications for the debate. On the former, thesequencing of partners may affect incentives of mem-bers to pursue MFN liberalization. Thus as discussedelsewhere in this report, the formation of large blocscreates incentives to join for smaller countries thattrade heavily with members—what has been termed“domino regionalism” (Baldwin 1993). The EU is thebest example of this phenomenon: EFTA countriesthat were not in favor of the EU integration model ul-timately concluded that the costs of staying out weretoo high. Whether this process makes RTAs more orless receptive to MFN reforms depends in part on thepreferences and interests of those who join the blocover time—which in turn will affect the willingness ofincumbent countries to accept new members.

On the product/policy coverage issue, the conven-tional wisdom appears to be that agreements shouldfirst focus on trade liberalization and then move onto behind-the-border areas—that is, go from shallowto deeper integration. There is no theoretical justifi-cation for this, however, and there is a well-documented history that, in the case of the EEC,many policymakers were of the view that the twoneeded to be pursued in tandem. The rhetoric ofpolicymakers and their advisors often suggests thatdeeper integration is necessary to attain free trade.During the period leading to the creation of the EEC,

Box 6.4 Sequencing of RTAs:Is there a good practice?

Jelle Zijlstra, the Dutch Minister of Economic Affairsargued that credible tariff removal required commonpolicies on taxes, wages, prices, and employmentpolicy. Similarly, the Belgian government felt thatpolicy harmonization was required to equalize costs,and that without it, a customs union would not befeasible because countries would impose new formsof protectionist policies. French officials persistentlydemanded harmonization in social policies—equalpay for both sexes, a uniform work week—as a pre-condition for trade liberalization—French standardsin this area were higher than in other countries(Milward 1992).

Recent research on the effects of, and interplaybetween, efforts to liberalize trade and investment inservices (and FDI more generally) suggests that coun-tries may be better of pursuing both shallow anddeeper forms of liberalization in tandem. Hoekmanand Konan (2001) and Konan and Maskus (2003),for example, note that not only can this generatemuch greater welfare gains, it can also reduce aggre-gate adjustment costs over time—through avoidingoutcomes in which factors of production must moverepeatedly across sectors (as will, by definition, occurif goods are liberalized first and then services/invest-ment, or vice versa). They also note that becausemany services continue to be less tradable thangoods, there is greater scope for employment oppor-tunities to be created as a result of allowing greatercompetition in services markets, thus helping to ab-sorb labor from other sectors as prices change due totrade reform.


Preferences when they were initially proposedin the 1960s. Indeed, unilateral preferenceprograms, especially the more comprehensiveand meaningful schemes adopted in recentyears, such as the EU’s Everything But Arms(EBA) program and the United States’ AfricanGrowth and Opportunity Act (AGOA), po-tentially make matters worse, because theirvalue to recipients depends on the existenceof trade barriers against imports from othercountries—with the greatest rents generatedby programs that distort markets the most,such as in the sugar market. Maintaining pref-erence margins—whether for RTA partners orbeneficiaries under unilateral preferenceprograms—is not the answer. What is neededis a willingness on the part of developed coun-tries to move away from preferential access asan instrument to assist lower-income partnercountries and to move toward greater relianceon direct transfers of technical and financialassistance (Stoeckel and Borrell 2001). Thishas the advantage of allowing high-incomecountries to target their trade-related develop-ment assistance to those countries most inneed, something the system of preferencescannot do.

Multilateral Disciplines onRegional Arrangements

Efforts to deal with the implications ofRTAs within the multilateral trading sys-

tem are long standing. The primary disciplinesare laid out in Article XXIV of the GATT—others are discussed in box 6.5. They permitRTAs if: (1) external trade barriers do not rise(Article XXIV:5); (2) all tariffs and other regu-lations of commerce are removed on substan-tially all exchanges of goods between thepartner countries within a reasonable length oftime (Article XXIV:8); and (3) notification ismade to the WTO Council.

The first criterion is intended to limit thenegative impact of an RTA on nonmembers.14

The second condition might at first seemcounterintuitive—after all, the more extensiveis the liberalization with an RTA, the more

detrimental it is likely to be to nonmembers.But the rationale here was in fact different; asnoted by Finger (1993), the objective was toensure that participants in RTAs are serious.In other words, while more trade in an RTAhurts nonmembers, pursuit of numerous par-tial RTAs can severely undermine the incen-tives for multilateral trade negotiations andcreate opportunities for special interests(farmers, specific industries) to carve out spe-cial arrangements. As a counterexample, if arestrictive interpretation (say 99 percent of alltrade) of this criterion were agreed to and en-forced, it is likely that the appetite for RTAswould be substantially diminished.

Determining whether the GATT or GeneralAgreement on Trade in Services (GATS) testsare met is left for the Committee on RegionalTrade Agreements (CRTA). Before the cre-ation of the WTO, the GATT Council usuallycreated a working party to evaluate whetherits conditions were satisfied. Under the WTO,the CRTA was established for this task. TheGATT experience in testing reciprocal prefer-ential trade agreement (PTAs) against ArticleXXIV was very discouraging. Starting withthe examination of the treaty establishing theEEC in 1957, almost no examination of agree-ments that were notified under Article XXIVled to clear conclusions or specific endorse-ments that the GATT requirements had beenmet. Only one working party could agree thata regional agreement fully satisfied the re-quirements of Article XXIV (the Czech-SlovakCustoms Union).

There had been a conscious political deci-sion made by GATT contracting parties in thelate 1950s not to scrutinize the formation ofthe EEC because the EEC member states hadmade it clear that if the EEC treaty was foundto be inconsistent with Article XXIV, theywould withdraw from GATT (Snape 1993).Given that the EEC did not meet all the re-quirements of Article XXIV—agriculturaltrade was not liberalized—it created a prece-dent. It is also true that the criteria and lan-guage of Article XXIV are ambiguous. Legiti-mate differences of opinion are possible


140

regarding issues such as the definition of“substantially all” trade; how to determinewhether the external trade policy of a customsunion has become more restrictive on average;and what a reasonable time period is for thetransition toward full implementation of anRTA (Hoekman and Kostecki 2001).

In the Uruguay Round some of these issueswere clarified: Specific criteria were adoptedto assess whether a customs union’s externaltariff satisfies Article XXIV; a 10-year maxi-mum for the transition period for implemen-tation of an agreement was imposed; and, asmentioned, a standing committee was createdto oversee RTAs. None of these changes hadany effect on the ability of WTO members toagree on whether an RTA conformed to WTOrequirements. Most observers would agreethat existing WTO disciplines and enforce-ment mechanisms have no teeth, and are not

particularly effective at controlling, limiting,or shaping the growth and coverage of RTAs.At the Fourth Ministerial Conference in Doha,ministers agreed to launch negotiations toclarify and improve the disciplines and proce-dures under the existing WTO provisions thatapply to RTAs, taking into account the devel-opmental aspects of these agreements.

The Doha Agenda negotiations on rules forpreferential trade agreements have been con-ducted on two tracks, one focusing on trans-parency issues and the other addressing thesubstantive disciplines. Transparency is gener-ally less contentious, and includes:

• Administrative overload in reviewingproposals. Review of long, detailed, andoften unclear RTA documents imposesa substantial workload on committeemembers. One possible solution would


141

The fundamental building block of the multilat-eral system is the principle of nondiscrimina-

tion. This is enshrined in Article I of the originaltext of the GATT signed in 1947. But from thestart, there have been exceptions. One such provi-sion is Article XXIV of the GATT, which permitsWTO members to enter into preferential tradeagreements (RTA) that extend trade concessions toRTA participants not offered to nonparticipants, aslong as certain criteria are satisfied—in particular,regarding the scope of the RTA. Second, rules cov-ering preferential agreements that deal with tradein services are set out in Article V of the GATS.Finally, developing countries may, if they wish, invokethe provisions of the 1979 Decision on Differentialand More Favourable Treatment, Reciprocity andFuller Participation of Developing Countries (the so-called Enabling Clause) to exchange tariff and, to acertain extent, nontariff preferences among them.Unlike Article XXIV, the Enabling Clause does notrequire that internal barriers be removed on “sub-stantially all” trade among participants in thosearrangements. MERCOSUR was notified to GATTunder this provision, not under Article XXIV. The

Box 6.5 RTAs and WTO disciplinesEnabling Clause also legitimizes non-reciprocal pro-grams such as the Generalized System of Preferences(GSP).

The task of verifying the WTO compliance ofRTAs notified under GATT Article XXIV and GATSArticle V is entrusted to the CRTA. The CRTA wasestablished in 1996, in particular to (1) oversee,under a single framework, all regional trade agree-ments, and (2) consider the implications of suchagreements and regional initiatives for the multilat-eral trading system and the relationship betweenthem. RTAs among developing countries, when noti-fied under the Enabling Clause, are not, in principle,subject to review by the CRTA. As was the caseunder the GATT, however, the CRTA has been un-able to carry out effectively its functions of examin-ing the consistency of RTAs with the rules, and over-seeing their implementation. The reason for this isessentially a fear of setting a precedent and openingup agreements to dispute settlement proceedings. TheCRTA has thus far been unable to conclude the ex-amination of any of the 110 RTAs currently underscrutiny and has a backlog of about 35 RTA reportsto prepare.

be an expanded “first review” by WTOstaff based on clear and objective criteria;this would require a significant expansionof resources allocated by the Secretariat.

• Notification of RTAs to the WTO. TheWTO has been notified of around 300agreements (of which about 140 are cur-rently in force), and another 60 or so arein advanced stages of negotiation (seechapter 2). Many of these are for RTAsamong developing countries. Moreover,few RTAs have been designated as“interim arrangements,” even thoughmost RTAs have been implemented instages.

• Data and information requirements.There is no clarity on the type, quantity,and level of detail to be provided whenthe WTO is notified of an agreement.Developing countries often lack thecapacity to undertake this task. Whilesome members want more detail, it is un-derstood that unduly straining the capac-ity in developing countries should beavoided. Some suggest that WTO assis-tance could be provided for the notifica-tion process if necessary.

• Review of South-South agreements. TheEnabling Clause does not provide forany consistency examination. Currentpractice is for RTAs formed between de-veloping countries (and notified underthe Enabling Clause) to be reported tothe Committee on Trade and Develop-ment (some developing countries havechosen to notify the WTO of agreementsunder Article XXIV). Countries haveraised the issue of reporting EnablingClause Agreements to a single body, theCRTA.

• Services. Some of the requirements of theprovisions of Article V of GATS are un-clear (e.g., in terms of the departure fromMFN obligations in key areas such astransparency, emergency safeguards, andadministration of domestic regulations).

• Process. Some delegations want notifica-tion of RTAs to occur before they are

implemented. Others support issuing afinal report that does not pass or fail aRTA, but allows members to registerconcerns. Finally, some delegations wantgreater diligence in encouraging RTAparticipants to file biennial reports onthe implementation of RTAs.

Although no early harvest on transparencyissues was undertaken for the 2003 CancunMinisterial, negotiations for a (provisional)application of strengthened surveillance mech-anisms are considerably advanced. Suchmechanisms would require more detail in theRTA notification procedures and might in-volve an enhanced role for the Secretariat inpreparing an assessment of each RTA thatwould be provided to WTO members.

Informal discussions on the substantiverules began in June 2004, and there are manyissues under consideration. The fact thatWTO rules on RTAs relate to several otherregulatory areas (some of which are undernegotiation)—including rules of origin, tradefacilitation rules on trade remedies, theGATS—adds to the complexity. Issues underconsideration include:

• Product coverage of RTAs. One centralissue is whether to make more specificthe requirement that “substantially all”trade be covered in each RTA. Lack ofclarity regarding this criterion is viewedby many as a source of the CRTA’s in-ability to reach clear-cut conclusions onWTO compatibility for most RTAs.Many RTAs exclude agriculture, which isa major problem. Some have proposedthat “substantially all” trade should bedefined as a percentage not only of ac-tual trade but also of all the six-digit tar-iff lines listed in the Harmonized System.This approach could ensure that thestandard is set high, but that there is alsosufficient flexibility to set aside productareas that remain sensitive for one rea-son or another. A related issue concernshow (if at all) such a criterion would be


142

applied to RTAs already notified—wouldthey be subject to the same percentage,or grandfathered in under the existingimprecise criterion? Although inherentlyarbitrary, a more specific coverage crite-rion could help to move the reviewprocess for RTAs along. Schiff and Win-ters (2003) suggest adopting a coveragecriterion of 95 percent of the value oftrade after 10 years of operation of theagreement, rising to 98 percent after 15years. This would require that at leastsome sensitive products be included (e.g.,agriculture), but not all, and wouldimply that over time the set of excludedproducts would decrease.

• Policy coverage of RTAs. GATT ArticleXXIV.5 requires that duties and otherregulations of commerce applied bymembers of a RTA are not more restric-tive than those existing prior to its for-mation. In assessing the impact of aRTA, there is no agreement on what thiscovers. In particular, it is unclear to whatextent it covers policies such as safe-guards, anti-dumping measures, mutualrecognition agreements, or rules oforigin—and if they are covered, howthey should be evaluated. Many RTAscontinue to allow for the use of an-tidumping and safeguards on intra-member trade—that is, conflict with thesupposedly free trade objective of RTAs.Only a few RTAs have abolished thereach of antidumping (the EU,ANZCERTA, Canada-Chile). Alterna-tively, some RTAs preclude the use ofsafeguards on goods originating in part-ner countries—for example, Canada andMexico were exempted from the recentU.S. steel safeguard action. This in-creases the negative effect of the actionfor nonmembers.

Whether rules of origin are a regulation ofcommerce has been a key source of disagree-ment for decades. The rules of origin are rele-vant not just for normal trade flows, but also

play a role in the application of safeguardsand other contingent trade policies. Issues re-lated to rules of origin were already being dis-cussed in the early 1970s in working partiesthat were considering RTAs. Thus in connec-tion with the 1972 free trade agreement be-tween the EEC and European Free Trade inEurope (EFTA) member states, the UnitedStates argued that the rules of origin wouldharm nonmembers.15 Not surprisingly, prefer-ential rules of origin are being discussed in thecurrent Doha talks, although expectations arelow that there will be agreement on commondisciplines. Discussions on harmonization ofnonpreferential rules of origin have beenunder way for almost 10 years and have yet tobe concluded.16 In any event, these harmo-nized rules of origin will not apply to regionalagreements or to GSP schemes. Because rulesof origin facilitate the fine tuning of preferen-tial liberalization at the product level, manycountries do not want to see constraints im-posed on their policy freedoms in this arena.

Many proposals have been made in theWTO for stronger rules for RTAs. One sug-gestion is that all RTA members be required toextend their preferential concessions to therest of the world within a specific time frame(Srinivasan, 1998).17 Another suggestion is tominimize the adverse discriminatory conse-quences of RTAs by requiring (or exhorting)members to allow any developing country to“opt in” on the same terms as existing mem-bers, perhaps after a certain time period. Thisgoes back to Viner (1950). As noted, virtuallyevery existing RTA has geographic restrictionson membership and has features that requirenegotiation, so the practical promise of suchopen regionalism is limited. Other suggestionsdo not make good economic sense (see Schiffand Winters, 2003).

Many observers have concluded that thequest for stronger rules is unlikely to succeedbecause many RTAs will not satisfy the rules,which will, in turn, lead countries to prefer thestatus quo. Schiff and Winters (2003) con-clude their discussion of regionalism and theWTO with a section called “Rules Are Not the


143

Answer.” From this perspective, the primaryfunction of the WTO is to act as a negotiatingforum to bring down the discrimination cre-ated by RTAs. The importance (and feasibil-ity) of this depends, in part, on the motivationof governments to pursue RTAs.

Making Open Regionalism Workfor DevelopmentDefining the role of the WTONumerous observers and analysts of RTAs,who are interested in enhancing their compat-ibility with the global trading system, proposethat WTO disciplines be applied more strin-gently. Years of discussion in the GATT/WTOon the interpretation of existing criteria, andthe many papers proposing more specific cri-teria have had no impact on the spread or con-tent of RTAs. Improving the enforcement ofArticle XXIV or strengthening/changing WTOdisciplines on regionalism is unlikely to fareany better.

Whether it is helpful to articulate more spe-cific criteria delimiting what “substantiallyall” trade means, or tightening up disciplineson “other regulations of commerce,” dependsin large part on the feasibility of attaininga consensus on specific criteria. Given theplethora of RTAs that are in force, and theshare of global trade notionally coveredthrough such agreements, a strong case can bemade that the horse has already bolted fromthe barn—shutting the barn door will makelittle difference. The problem is a political one;any number of countries will oppose strongerrules because they are already members ofRTAs that would violate them. Indeed, even ifgroups of developed countries—such as theEU—make a case that their agreement satisfieswhatever criteria might be proposed, or thattheir model should become the benchmark,other countries could legitimately argue that ifit was acceptable for the EEC to pass musterin the 1960s and 1970s, it would be hypocrit-ical to impose stronger rules on all WTOmembers today.

Nonetheless, WTO members should lookfor ways that existing disciplines on RTAs canbe reinforced and new disciplines introducedto enhance their development impact. Onearea where knowledge is much more extensivenow, compared to when the original GATTrules were written, regards rules of origin. It isclear that complex and restrictive rules of ori-gin limit the benefits of RTAs for developingcountry participants and divert trade in inter-mediate products. Rules of origin that differacross agreements complicate world tradingconditions and contribute to the emergence ofhub-and-spoke patterns. Thus there would besubstantial benefits from an agreement thatpromoted common, simple, less restrictive,and easy-to-apply rules of origin. That beingsaid, the delays and problems in achieving theobjectives defined in the Uruguay Round forthe nonpreferential rules of origin are not pro-pitious in this case.

From a development perspective, the mostuseful and immediate step for the WTO is toimprove transparency. Information and analy-sis are important inputs for a well-functioningtrading order. Greater monitoring and assess-ment of the impacts of RTA-related policieswould allow more informed and proactive en-gagement by civil society (think tanks, non-governmental organizations, consumers, andtaxpayers) in the policy formation and negoti-ation process. It is true that to reduce protec-tion and protectionist pressures, those thatlose (pay) need to be aware of the costs ofsuch policies. The suppliers of and the clientsfor such analysis and information are not nec-essarily governments, but the constituencies inindividual countries who are affected by pol-icy. In order for trade agreements to promotegood policy-making in member countries,stakeholders must be able to be active in thedomestic policy formation process.

This could be achieved by augmenting thecapacity of the WTO Secretariat andthe CRTA to review, document, and analyzethe effects of RTAs. That is, WTO attention inthis area should focus on gathering informa-tion and analysis. Efforts could concentrate on


144

addressing such questions as: Are RTAs beingimplemented? How? What is the effect onmember countries and on nonmembers? Howmuch trade is covered by the RTA (and is it“substantially all”)? Ideally, stronger surveil-lance would involve those responsible forimplementing each RTA and, in the process,empower them to engage policymakers andstakeholders. Such surveillance and analyticalmonitoring should extend to South-SouthRTAs—which should all be notified to theCRTA.

Strengthening information exchange andmutual (multilateral) dialogue, (includingthrough the establishment of formal monitor-ing mechanisms), would facilitate cooperationon new regulatory issues. If existing or pro-posed policies could be evaluated using objec-tive criteria on their ability to achieve thestated national objectives, countries andstakeholders could assess their efficacy and, ifneeded, adjust the policies. Greater analysis ofthe effects of trade discrimination—both uni-lateral preferences and RTAs—should be partof this agenda.

Information of this type would help RTAstakeholders to hold governments accountablefor outcomes and to assist nonmembers byproviding data that could feed into demandsin the context of WTO negotiations.18 Asargued by Hoekman and Kostecki (2001)among others, the primary means throughwhich the WTO can impose limits on RTAs isby providing a venue for negotiation to reduceMFN barriers, which will automatically limitthe discrimination against outsiders that is in-herent in RTAs.

Priorities for the industrial countries The major industrial countries must strike adifficult balance in their pursuit of bilateraland regional deals with developing countries.On the one hand, there are legitimate reasonsto pursue such initiatives, rather than relyingonly on multilateral channels. The ongoingEU effort to deepen relations with a widerEurope is an important example. Cooperationamong a smaller number of countries may

also be the most effective solution to mitigat-ing environmental externalities, addressingnontrade issues such as labor migration (legalor illegal), or attaining national security objec-tives. But it must be recognized that the ag-gressive pursuit of RTAs with developingcountry partners in all corners of the globeserves the cause of global liberalization poorlyif it delays or halts altogether the progresstoward multilateral liberalization.

Fostering development is increasingly iden-tified as a key rationale for North-SouthRTAs. In part this reflects economicinterests—growing markets abroad are ex-panding export markets—and in part it is dueto the recognition that there is a correlation be-tween sustained economic growth in partnercountries and national security. From this van-tage point, several recommendations emerge.

The highest priority should be to ensurethat the Doha Development Agenda is com-pleted in a manner that provides new marketaccess to exporters in developing countries. Inaddition to their interests in development, thelarge countries have an important historicaland systemic responsibility to ensure that theworld trading system remains as open aspossible.

Supporting open regionalism by encourag-ing partner countries to adopt low externalbarriers would create momentum towardintegration with the world, establish a moreefficient development path, and reduce theadverse negative implications of RTAs on non-members. Because these countries are lesslikely to be preferred RTA partners of theUnited States and the EU, they will continue tohave an incentive to engage at the WTO level.

Adopting the widest possible product cover-age and greatest market access expansionwould increase the development impact. Thismeans that agricultural trade policies should beincluded in RTAs. Excluding agriculture fromRTAs—the dominant practice at present—doesnot promote development. Although inclusionof agricultural market access will increase theincentives for small countries to seek RTAs withthe large players, it will also increase the


145

downside of RTAs for large agricultural pro-ducers in the developing world. The latter willthen have a greater incentive to push for WTO-level MFN reforms; the marginal increases incompetition from preferred partners may helpovercome the domestic interests opposed to re-forms through a process of gradual, piecemealexpansion of access to agricultural markets.

Industrial countries negotiating North-South RTAs should adopt liberal cumulationprovisions in their rules of origin. Because the

least-developed countries (LDCs) already havenearly free access to OECD markets, North-South RTAs will erode such preferences. Cu-mulation will ensure that rules of origin donot impose an additional burden on thesecountries. For member countries, liberal rulesof origin will enhance the benefits of North-South RTAs. A demonstrated willingness ofindustrialized countries to put partner countryinterests before those of national industrygroups (who prefer restrictive rules of origin)


146

In July 1995, Tunisia signed the Association Agree-ment with the European Union (AAEU). The agree-

ment, which came into effect in March 1998, wouldliberalize trade for industrial goods over a 12-yearperiod and ultimately create a free-trade zone. Trade inagricultural goods and services was left out for futurenegotiations. By the mid-1990s, Tunisia had alreadybecome a successful exporting country, thanks to theestablishment, in 1972, of a special offshore system forexporting enterprises that mitigated the anti-exportbias of the highly protective trade regime.* Exports ofmanufactured goods had increased from 4 percent ofGDP in 1975 to 20 percent by 1994. Free market ac-cess to EU markets under the AAEU further enhancedTunisia’s export performance, with manufactured ex-ports increasing to 25 percent of GDP in 2002.

The AAEU gave momentum to trade liberaliza-tion in Tunisia. Average MFN tariffs were reducedfrom 33 percent in 1994 to 26 percent in 2003.However, Tunisia still posts the second highest aver-age MFN tariffs in the broader EU neighborhood,including in the EU accession countries. For exam-ple, only Morocco, with a tariff rate of 30 percent,has a higher border barrier, but countries as diverseas Turkey, Bulgaria, Ukraine, Lebanon, and Moldovahave tariffs of 10 percent or lower.

High MFN tariff differentials in the presence ofEU preferential access risks diverting trade awayfrom the lowest cost sources, denying Tunisian pro-ducers and consumers the benefit of less expensiveimports from outside the preferential trade zone withthe EU. As a result of high MFN tariffs and geo-graphical proximity, about 75 percent of Tunisia’s

Box 6.6 Tunisia’s Association Agreement with theEuropean Union

trade is dependent on the European market—especially in EU neighborhood countries.

A trade strategy linked to the domestic reformagenda would help Tunisia fully realize the develop-ment promise of deeper trade integration. First, am-bitious reduction of MFN tariffs for industrial goodswould prevent trade diversion. Second, removingbeyond-the-border obstacles to trade—by reducingthe still-high trade logistics costs—would enhancefirms’ ability to exploit export opportunities and im-prove their competitiveness; liberalization of back-bone services, especially in transport, ICT, and fi-nance, are also essential. Third, once the free-tradezone with the EU is fully implemented in 2007, pref-erential tax treatment of exporting firms will becomemuch harder to justify, because both offshore and on-shore firms will be equally exposed to foreign compe-tition. The regulatory framework of investment in-centives and trade facilitation will thus have tobecome more even. The EU could also liberalize mar-ket access for Tunisian agricultural exports in prod-ucts in which Tunisia is competitive. Unless marketaccess is improved, the scope for farmers to shift intothese products will remain limited, and sectoral ad-justment in agriculture will continue to be impaired.

*This system covers companies located anywhere in thecountry and it grants duty-free imports of capital and intermedi-ate goods, exemptions from the VAT and excise taxes, and ex-emption from the corporate income tax for the first 10 years ofoperation.


would show that development objectives arebeing taken seriously in RTAs.

Industrial countries could also enhance thedevelopment credentials of their RTAs by tak-ing action to abolish antidumping and similarinstruments of contingent protection. There isa plethora of evidence that antidumping isstraightforward protectionism and that inso-far as there is a rationale for intervention,other policy instruments can be used (i.e.,competition policy). A number of RTAs—ANCERTA, Canada-Chile, the EU itself—have proceeded down this path, which illus-trates that it is feasible.19

Industrial countries should exercise cautionregarding their demands for new regulatorypolicies in RTAs. Behind-the-border, regula-tory policies are critical for a positive impacton development outcomes, but getting the rulesright—ensuring that rules are calibrated to de-velopment capacities and do not detract fromother, more pressing priorities—is as essentialas it is difficult to orchestrate. Applying regu-latory norms in RTAs on a nondiscriminatorybasis will avoid creating another complex setof discriminatory preferences. In other words,RTAs will be supportive of development if thenegotiation and implementation process is de-signed to ensure that such priorities are set ap-propriately, and the preconditions for benefit-ing development are in place. Many RTAs arefar from satisfying this prescription.

Increasing the effectiveness of developmentassistance for trade can help. Aid is an impor-tant part of the equation motivating RTA ne-gotiations, especially for agreements with theEU. Export growth in many LDCs and othersmall and low-income countries is limited bythe lack of supply capacity and the high-costbusiness environment. Firms in these countriesmay also find it difficult to deal with the regu-latory requirements that apply in exportmarkets. Health and safety standards, for ex-ample, are often regulatory barriers to entry;the standards can be excessively strict and thecompliance costly, which weighs dispropor-tionately on producers in low-income coun-tries. Development assistance can help to build

the institutional and trade capacity needed tobenefit from increased trade and better accessto markets. This assistance will be more effec-tive when it is focused more broadly on supplycapacity, and when it addresses the adjustmentcosts associated with reforms. Recently atten-tion has been given to expanding programsthat provide aid for trade, but only when tradeissues are integrated into a nation’s overall de-velopment priorities. Although priorities willdiffer, in many cases assistance will be neededto address trade-related policy and public in-vestment priorities. More could be done to re-place preferential access as the primary carrotfor RTAs with financial transfers. It hasbeen argued at length (e.g., Hoekman,Michalopoulos, and Winters 2004), that tradepreferences should not be a permanent featureof the global trading system. Appropriatelydesigned aid would offer a similar result topreferences and at a lower overall cost.

Challenges for developing countries: Athree-pronged strategy Developing countries would benefit fromadopting a coherent three-part strategy thatintegrates unilateral, multilateral, and regionalinitiatives. A number of middle-income coun-tries have enunciated a clear set of prioritiesand objectives regarding regional and multi-lateral efforts, and have the technical and ne-gotiating capacity to pursue them; however,many developing countries show less evidenceof a coherent strategy on how to use RTAs tomaximum advantage.

At the multilateral level, the Doha Agendanegotiations are the best instrument for mostdeveloping countries to reduce the discrimina-tion they face from the prevailing web ofRTAs. Doha is also critical insofar as it pro-vides the greatest potential new market accessfor the greatest number of the world’s poor. Itis also the only venue in which key policyareas such as agricultural support policies orantidumping can be negotiated in a compre-hensive and substantive manner. As with thehigh-income countries, completing the Dohadeal is the highest priority.


147

Unilateral reforms can increase the benefitsof RTAs. Governments should generally lowerexternal (MFN) barriers to trade in conjunc-tion with the reduction of trade barriersagainst partners and build this commitmentinto the agreements signed. This not only re-duces the scope for detrimental trade diver-sion, it also helps to support multilateralismand increase global welfare. Such commit-ments can be made in the context of the on-going Doha Round, with the advantage of po-tentially generating additional quid pro quoreductions by other WTO members. Thatsaid, strategies that deliberately delay tradeliberalization and domestic reforms in orderto conserve bargaining chips for either RTA orDoha Agenda negotiations are misguided.Most developing countries have only limitedleverage within any such negotiations, and thecosts of delayed reforms outweigh any transi-tory negotiating advantage. The rapid growthrates realized by countries such as India fol-lowing trade liberalization and relatedreforms in the 1990s illustrate this point.

Pursuing bilateral or plurilateral North-South RTA arrangements at the regional levelmay yield short-term market access benefits toparticipating developing countries. But coun-tries should be aware that preferential access islikely to be eroded as more countries sign suchdeals, reducing the value preferences, if not theaccess. Moreover, the impact of commitmentson nontrade regulatory policies must be takeninto account. As noted above, it is importantto get the rules right. For RTAs to be beneficialin the longer term, there needs to be mecha-nisms through which governments and stake-holders are assisted in implementing a set ofreforms that will help sustain growth andreduce poverty. The establishment of these na-tional mechanisms is an important precondi-tion for benefiting from the RTA game.

South-South RTAs, especially regionalagreements among neighboring countries, cangenerate substantial benefits to participants innon-trade policy areas such as border cross-ings, infrastructure, standards and regulatoryframeworks, and related enforcement institu-

tions. Emphasis should be directed toward en-hancing cooperation in areas where there is astrong public goods dimension—collaborationhere can yield large gains. Such cooperationcan have positive spillovers on the multilateralprocess by helping to identify what type ofmultilateral rules might be beneficial.

Only by adopting a three-part integratedstrategy built around unilateral, multilateral,and regional components can developingcountries ensure that their trade policy willcontribute the most to growth and poverty re-duction. To be fully effective, each trade pol-icy lever must be used to its best purpose, andcoherent trade policy and associated technicalassistance resources have to be integrated intoa national development strategy. In the case oflow-income countries, the Poverty ReductionStrategy Paper (PRSP) is an example of such asinstrument. Governments and stakeholdershave an interest in building such a three-partstrategy into their development strategies. Inthis way, open regionalism can help countriesimprove their standards of living.

Notes1. For a summary of earlier World Bank research

on the linkage and compatibility of regionalism andmultilateralism, see World Bank (2000) or Schiff andWinters (2003).

2. The results reported herein refer to so-called“static” gains, that is, it is assumed that trade reformhas no impact on productivity. Thus the gains aremainly associated with the efficiency gains from re-moving the trade (and perhaps other) distortions.

3. This grouping is referred to as the Quad-plus, orQuad�, because of the inclusion of Australia and NewZealand.

4. Note that under the bilateral agreements, do-mestic distortions are not removed because they arenot specific to any individual trading partner.

5. Note that some of the high-income Asiancountries—for example, the Republic of Korea andSingapore—are excluded from the bilateral agreementsand therefore tend to lose, in part, because of tradediversion effects.

6. The excluded countries/regions are Brazil,China, India, Mexico, Russia, rest of East Asia, andrest of South Asia.


148

7. The three North-South agreements include abroad East Asian region that encompasses both thehigh-income and developing countries, the FTAA in theWestern Hemisphere, and a broad free trade area cen-tered on the EU, including the new accession countries,extending to the Middle East and North Africa and Sub-Saharan Africa. The South-South agreements include aLatin American-wide RTA (LAFTA), a developing EastAsia RTA (AFTA), a Europe and Central Asia RTAexcluding the EU-accession countries (CIS), a Sub-Saharan African RTA (SSA), the Middle East and NorthAfrica (MNA), and a South Asian RTA (SAFTA).

8. In a result not shown in these figures, South Asiais not included in any of the three broad North-Southagreements, but is simulated to create its own RTA. Inisolation, the gains from SAFTA, the South Asian re-gional trade agreement, yields a positive gain of some0.2 percent of baseline income, as shown in figure 6.3.However, SAFTA’s exclusion from the 3-block North-South agreements induces a loss of 0.3 percent of base-line income, largely due to trade diversion and a loss inthe terms of trade.

9. The simulated South/South agreements reflectfree trade among geographic neighbors where the dif-ferences in comparative advantage are slight. Reduc-tions in South/South trade barriers across more distantcountries have significantly more potential for increas-ing trade and incomes.

10. Concerns about the impact of discriminationon multilateralism in trade has a long history, e.g.,Patterson (1965). There is voluminous literature onthis issue, much of it conceptual. Interested readers arereferred to Bhagwati (1991), De Melo, Panagariya andRodrik (1993), Winters (2000, 2001), and Schiff andWinters (2003).

11. Much of the early theoretical literature that an-alyzed the incentive effects of RTAs for multilateralismassumed, for simplicity, that the latter implied freetrade—see Levy (1997), and Krishna (1998). [Winters(2000) provides a thorough review of the theory.] As aresult, the focus was on a binary choice between mul-tilateral free trade and no MFN liberalization. How-ever, in practice countries can choose to conclude amultilateral round with considerable liberalization orwith very little. Thus an empirical perspective that fo-cuses on whether RTAs change the probability of con-cluding a trade round is too narrow. Moreover, it is vir-tually impossible to test this in a systematic waybecause one never observes simultaneously a countrywith and without RTAs. Therefore an empirical assess-ment of the stumbling block hypothesis must useBhagwati’s definition and investigate if RTAs lead toless multilateral liberalization.

12. Testimony before the House Committee onWays and Means, May 8, 2001; accessed June 2003 on

http: / /waysandmeans.house.gov/legacy/trade/107cong/5-8-01/5-8chri.htm; U.S. tariff schedule 1994and 2003. See Limão (2003).

13. These are described in chapter 2.14. A practical problem faced by the drafters of

Article XXIV was that in the case of a customs union,changes in the external tariffs of member countrieswould occur as they adopt a common external tariff.The rule that applies to customs unions is that dutiesand other barriers to imports from outside the unionmay not be, on the whole, higher or more restrictivethan those preceding the establishment of the customsunion (Article XXIV:5a). The interpretation of thisphrase became a source of much disagreement amongGATT contracting parties. However, except for cus-toms unions, (in which a common tariff structurewould be adopted), the rule for RTAs was unambigu-ous. Duties applied by each individual member countrymay not be raised (Hoekman and Kostecki 2001).

15. The RTA was argued to lead to “trade diver-sion by raising barriers to third countries’ exports ofintermediate manufactured products and raw materi-als. This resulted from unnecessarily high requirementsfor value originating within the area. In certain cases ...the rules disqualify goods with value originating withinthe area as high as 96 percent. The rules of origin lim-ited non-originating components to just five percent ofthe value of a finished product of the same tariff head-ing [for] nearly one-fifth of all industrial tariff head-ings” (GATT, 1974:152-53, cited in Hoekman andKostecki 2001).

16. The WTO Agreement on Rules of Origin aimsto foster the harmonization of the (nonpreferential) rulesused by members. The objective is that common rules beapplied equally to all nonpreferential trade policy instru-ments by WTO members—tariffs, import licensing, an-tidumping, and so forth. The agreement calls for a workprogram to be undertaken by a Technical Committee, inconjunction with the WCO, to develop a classificationsystem regarding the changes in tariff subheadings(CTH), based on the Harmonized System (HS); thiswould constitute a substantial transformation. In caseswhere the HS nomenclature does not allow substantialtransformation to be determined by a CTH test, theTechnical Committee is to provide guidance regardingthe use of supplementary tests such as value-added crite-ria. Although the harmonization program was to becompleted by July 1998, deadlines have also beenmissed, in part, reflecting lack of consensus among mem-bers over the formulations to be adopted for sensitiveproducts, especially agriculture, textiles, and clothing.

17. As the implied move to full global free trade isunlikely to be politically feasible, the implicationwould be to impose a ban on most if not all RTAs—also unlikely to be acceptable.


149

18. Clearly there would be budgetary implicationsassociated with a stronger surveillance role, and aprecondition is that WTO members accept that theSecretariat be given the independence to undertake theanalysis and form an explicit judgment of the effects ofspecific agreements. However, if the required resourcesor willingness for the WTO to undertake the task can-not be found, this in itself would be a good indicationof the importance that is accorded by WTO membersto the spread of RTAs.

19. See Hoekman (1998) for a discussion.

ReferencesAghion, Philippe, Pol Antras, and Elhanan Helpman.

2004. Negotiating Free Trade. NBER WorkingPaper 10721, National Bureau of Economic Re-search, Cambridge, MA.

Baldwin, Richard. 1993. A Domino Theory of Region-alism. CEPR Discussion Paper No. 857. Centerfor Economic Policy Research, Washington, DC.

Bhagwati, Jagdish. 1991. The World Trading System atRisk. Princeton, NJ: Princeton University Press.

Bhagwati, Jagdish, and Arvind Panagariya. 1996. TheEconomics of Preferential Trading Agreements.Washington, DC: AEI Press.

Dee, P., and J. Gali. Forthcoming. The Trade and In-vestment Effects of Preferential Trading Arrange-ments. In NBER East Asian Seminar in Economics14 Proceedings. Chicago: University of ChicagoPress.

De Melo, Jaime, Arvind Panagariya, and Dani Rodrik.1993. Regional Integration: An Analytical andEmpirical Overview. In New Dimensions in Re-gional Integration, eds. J. De Melo and A. Panagariya. New York: Cambridge UniversityPress.

Finger, J. Michael. 1993. GATT’s Influence on RegionalAgreements. In New Dimensions in RegionalIntegration, eds. J. De Melo and A. Panagariya.New York: Cambridge University Press.

Harrison, Glenn, Thomas Rutherford, and David Tarr.2002. Trade Policy Options for Chile: The Im-portance of Market Access. World Bank Eco-nomic Review 16(1).

Hoekman, Bernard. 1998. Preferential Trade Agree-ments, In Brookings Trade Forum 1998, ed. RobertLawrence. Washington, DC: Brookings Institution.

Hoekman, Bernard, and Denise Konan. 2001. Deep In-tegration, Nondiscrimination and Euro-Mediterranean Free Trade. In Regionalism inEurope: Geometries and Strategies After 2000,eds. Jurgen von Hagen and Mika Widgren.Kluwer Academic Publishers.

Hoekman, Bernard, and Michel Kostecki. 2001. ThePolitical Economy of the World Trading System,2d ed. New York: Oxford University Press.

Hoekman, Bernard, Constantine Michalopoulos, andL. Alan Winters. 2004. Special and DifferentialTreatment of Developing Countries: MovingForward After Cancún. The World Economy 27:481–506.

Konan, Denise, and Keith E. Maskus. 2003. Quantify-ing the Impact of Services Liberalization in a De-veloping Country. Policy Research Working Paper3193, World Bank, Washington, DC.

Krishna, Pravin. 1998. Regionalism and Multilateral-ism: A Political Economy Approach. QuarterlyJournal of Economics 113: 227–51.

Lawrence, Robert Z. 1991. Emerging RegionalArrangements: Building Blocks or StumblingBlocks? In Finance and the International Econ-omy 5: The AMEX Bank Review Prize Essays,23-25, ed. Richard O’Brien. New York: OxfordUniversity Press.

Levy, Philip. 1997. A Political-Economic Analysis ofFree-Trade Agreements. American Economic Re-view 87(4): 506–19.

Limão, Nuno. 2003. Preferential Trade Agreementsas Stumbling Blocks for Multilateral TradeLiberalization: Evidence for the U.S. Paper. Uni-versity of Maryland, College Park.

Mansfield, and Reinhardt. 2003. Multilateral Determi-nants of Regionalism: The Effects of GATT/WTOon the Formation of Preferential Trading Arrange-ments. International Organization 57(4): 829–862.

Milward, Alan. 1992. The European Rescue of theNation State. Berkeley: University of CaliforniaPress.

Patterson, Gardner. 1966. Discrimination in Interna-tional Trade: The Policy Issues, 1945–1965.Princeton, NJ: Princeton University Press.

Saggi, Kamal. 2004. Preferential Trading Arrange-ments and Multilateral Tariff Cooperation.Southern Methodist University Working Paper.Dallas.

Saggi, Kamal, and Halis Murat Yildiz. 2004. BilateralTrade Agreements and the Feasibility of Multilat-eral Free Trade. Southern Methodist Universityand Ryerson University Working Paper. Dallas.

Schiff, Maurice. 1997. Small is Beautiful: PreferentialTrade Agreements and the Impact of Country Sizeand Market Share. Journal of Economic Integra-tion 12: 359–87.

Schiff, Maurice, and L. Alan Winters. 2003. Regional-ism and Development. New York: OxfordUniversity Press.

Schott, Jeffrey. 2004. Assessing US FTA Policy. In FreeTrade Agreements: US Strategies and Priorities,


150

ed. Jeffrey Schott.Washington, DC: IIE (Institutefor International Economics).

Snape, Richard. 1993. History and Economics ofGATT’s Article XXIV, In Regional Integration andthe Global Trading System, eds. K. Anderson andR. Blackhurst. London: Harvester-Wheatsheaf.

Srinivasan, T. N. 1998. Developing Countries and theMultilateral Trading System: From GATT to theUruguay Round and Beyond. New York: HarperCollins.

Stoeckel, Andrew, and B. Borrell 2001. PreferentialTrade and Developing Countries: Bad Aid, BadTrade Canberra: RIRDC (Rural IndustriesResearch & Development Corporation).

Tojo, Yoshiaki. 2004. The WTO and FTAs. Presenta-tion at the JEF-IIE Conference, 16 March.

U.S. GAO (United States General Accounting Office).2004. International Trade: Intensifying FreeTrade Negotiating Agenda Calls for Better Allo-cation of Staff and Resources. Report to Congres-sional Requesters, Washington, DC.

Viner, Jacob. 1950. The Customs Union Issue. NewYork: Carnegie Endowment for InternationalPeace.

Winham, Gilbert. 1986. International Trade andthe Tokyo Round Negotiation. Princeton, NJ:Princeton University Press.

Winters, L. Alan. 2000. Regionalism versus Multilater-alism. In Market Integration, Regionalism andthe Global Economy, eds. R. Baldwin,Daniel Cohen, André Sapir and Anthony J. Venables. London: Centre for Economic PolicyResearch.

. 2001. Post-Lomé Trading Arrangements:The Multilateral Option. In Regionalism inEurope: Geometries and Strategies After 2000,eds. Jurgen von Hagen and Mika Widgren.Kluwer Academic Publishers.

World Bank. 2000. Trade Blocs. Washington, DC:World Bank.


151

As an important annual publication from the World Bank, GlobalEconomic Prospects presents the Bank’s view of the global economyand analyzes prospects for poverty reduction around the world.This year’s 15th edition examines the proliferation of regional trade agreements that has fundamentally altered the world trade landscape.

While the multilateral WTO talks remain mired in contentious debate,new bilateral and regional preferential trading arrangements are mushrooming all over the

globe. Will these agreements promote or stifle development? Global Economic Prospects 2005 addressesthis and other questions, such as:

• What type of regional and bilateral arrangements are most beneficial?• Do these agreements inspire deeper integration that multilateral trade agreements cannot? • Do they contribute to—or detract from—incentives for countries to engage in the multilateral Doha

trade talks?

WORLD BANKPublications

Visit our website at www.worldbank.org/publications

See other side for order form.

New! Global Outlook onlineYour source for:• Macroeconomic indicators and forecast data, from 1980

to 2006• Insightful interactive calculators and simulation tools• One-page PDF briefs summarizing countries’ external

financial position and trade• 16 individual commodity reports and price forecasts• Timely analysis of worldwide economic prospects and risks

Please visit the new interactive Global Outlook website at www.worldbank.org/globaloutlook

The reference of choice on development

Global Economic Prospects 2005 Trade, Regionalism, and Development

The World BankGLOBALECONOMIC PROSPECTS

Name _________________________________________________

Title ___________________________________________________

Organization ____________________________________________

Address________________________________________________

City ___________________________________________________

State __________________________________________________

Zip/Postal Code__________________________________________

Country ________________________________________________

Phone _________________________________________________

Fax ___________________________________________________

E-mail _________________________________________________

�� Charge $__________________ to my:

�� American Express �� Mastercard �� Visa

�� Card no._____________________________________________

Expiration date _____________ /____________

Name _________________________________________________as it appears on the card

Signature _______________________________________________required for all credit card charges

�� Check no._______________

in the amount of $__________________ is enclosed. When ordering directly from the World Bank, make check payable in U.S. funds drawn ona U.S. bank to: The World Bank. Please send your check with your order.

Institutional customers in the U.S. only:�� Bill me. Please include purchase order.

Mail order to: World Bank Publications

P.O. Box 960, Herndon, VA 20172-0960, USATelephone: 703-661-1580 or 800-645-7247

Fax: 703-661-1501www.worldbank.org/publications

Questions? E-mail us at [email protected]

Order Form

DCGP5

PAYMENT METHODOrders from individuals must be accompanied by payment or credit card information. Credit cards are accepted only for orders addressed to the World Bank. Check with your local distributor about acceptable credit cards. Please do not send cash. For customers in developing countries, geographic discounts apply. Please see http://publications.worldbank.org/discounts for details.

WORLD BANKPublications

Visit our website at www.worldbank.org/publications

* Shipping and Handling charges are $8.00 per order. If a purchase order is used, actual shipping will be charged. For air mail outside the U.S., charges are US$7.00 per order plus US$6.00 per item.

Title Stock # Price Qty. Total US$

Global Economic Prospects 2005: Trade, Regionalism, and Development D15747 US$38November 2004. (ISBN 0-8213-5747-6).

Subtotal

Shipping and Handling*

Total US$

The reference of choice on development

�� YES, please send me Global Economic Prospects 2005: Trade, Regionalism, and Development(Please Print)

The proliferation of regional trade agreements is fundamentally altering the world trade landscape.The number of agreements in force surpasses 200 and has risen eight-fold in two decades.

Today as much as 40 percent of global trade takes place among countriesthat have some form of reciprocal regional trade agreement.

Global Economic Prospects 2005:Trade, Regionalism, and Development addressestwo questions:

• What are the characteristics of agreements that most promote—or hinder—development for member countries?

• Does the proliferation of agreements pose risks to the multilateral trading system, and if so, how can these risks be managed?

The report argues that agreements leading to open regionalism—that is,deeper integration of trade as a result of low external tariffs, increased services competition, and efforts to reduce cross-border and customs delayscosts—are effective as part of a larger trade strategy to promote growth.Such regional agreements can complement a strategy that, on the onehand, includes autonomous liberalization to promote productivity gainsand, on the other hand, leverages domestic reforms to enhance marketaccess.

Although regional agreements can prove beneficial to member coun-tries, they can have adverse effects on excluded countries.The lowering of border barriers around the world is crucial to minimizing these effects.The completion of the Doha Development Agenda by all countries in the World Trade Organization will reduce the risk of trade diversion associated with regional agreements and will decrease the trade losses of countries excluded from agreements.

Chapter 1 of this study presents the World Bank’s view of the globaleconomy. It analyzes both short-term dynamics for the global economy and long-term prospects for reductions in poverty around the world.Chapter 2 introduces the issues associated with regional trade agreementsand provides an overview of regional trading trends. Subsequent chaptersexamine the effects of regional agreements on trade creation (chapter 3);trade facilitation (chapter 4); and services, investment, intellectual propertyrights, and labor mobility (chapter 5). Chapter 6 analyzes ways that regionalagreements can be made more compatible with a nondiscriminatory multilateral system.

For additional information and to access the Global Outlook online publication, please visit www.worldbank.org/prospects.

ISBN 0-8213-5747-6

A wave of regional agreements is sweeping the trade world.Both North–South bilateral

arrangements, such as the United States–Chile free trade agreement, and South–South

plurilateral arrangements, such asthe just-proposed South Asia

Free Trade Area, pose challengesand opportunities for developing

countries.These agreements not only influence trading

opportunities for people inmember countries and elsewhere

but also violate the most basicprinciple of the multilateral

trading system: nondiscrimina-tion. Global Economic Prospects

2005 takes on two urgent questions: can these agreements

be designed to promote development in its broadest

sense, and can they do so without hurting countries left out?

—François BourguignonSenior Vice President and

Chief Economist

global economic prospects - world bankpubdocs.worldbank.org/en/295201443469726534/global... · 1.4...

Documents