WTO Negotiations and Agricultural Trade Liberalization: The Effect of Developed Countries' Policies on Developing Countries (Cabi Publishing)

WTO NEGOTIATIONS AND AGRICULTURAL TRADE LIBERALIZATION The Effect of Developed Countries’ Policies on Developing Countries

This page intentionally left blank

WTO NEGOTIATIONS AND AGRICULTURAL TRADE LIBERALIZATION The Effect of Developed Countries’ Policies on Developing Countries

Edited by

E. Diaz-Bonilla Inter-American Development Bank Washington DC USA

S.E. Frandsen Danish Research Institute of Food Economics Denmark and

S. Robinson Department of Economics University of Sussex UK

CABI is a trading name of CAB International CABI Head Office Nosworthy Way Wallingford Oxon OX10 8DE UK

CABI North American Office 875 Massachusetts Avenue 7th Floor Cambridge, MA 02139 USA

Tel: +44 (0)1491 832111 Fax: +44 (0)1491 833508 E-mail: [email protected] Website: www.cabi.org

Tel: +1 617 395 4056 Fax: +1 617 354 6875 E-mail: [email protected]

© CAB International 2006. All rights reserved. No part of this publication may be reproduced in any form or by any means, electronically, mechanically, by photocopying, recording or otherwise, without the prior permission of the copyright owners. A catalogue record for this book is available from the British Library, London, UK. A catalogue record for this book is available from the Library of Congress, Washington DC, USA. ISBN-10: ISBN-13:

1-84593-050-9 978-1-84593-050-9

Produced and Typeset in 10pt Palatino by Columns Design Ltd, Reading Printed and bound in the UK by Biddles Ltd, Kings Lynn

Contents

Contributors Preface

vii ix

1

Overview Eugenio Diaz-Bonilla, Søren E. Frandsen and Sherman Robinson

1

2

Review of the EU Common Agricultural Policy Søren E. Frandsen and Aage Walter-Jørgensen

34

3

The Common Agricultural Policy in an Enlarged Europe: Bright or Bleak Prospects for Africa? Birgitte Gersfelt and Hans G. Jensen

57

4

US Agricultural Policy: the 2002 Farm Bill and WTO Doha Round Proposals David Orden

80

5

The Effects of Domestic Agricultural Reforms and Market Access on Trade and Production in Less Developed Countries Sherman Robinson and Karen E. Thierfelder

103

6

Potential Coalitions and Convergence in the Doha Round Kim Martin Lind and Christian Bjørnskov

122

7

Assessing the Harbinson Draft on Modalities in the WTO Agriculture Negotiations Søren E. Frandsen, Hans G. Jensen, Kim M. Lind, Poul P. Melgaard and Wusheng Yu

142

8

Food Security and the World Trade Organization: a Typology of Countries Eugenio Diaz-Bonilla, Marcelle Thomas, Sherman Robinson and Andrea Cattaneo

162

v

vi

Contents

9

A Proposal for Combating Acute Food Shortages Based on Sub-Saharan African Needs Kim M. Lind

184

10 Thinking Inside the Boxes: Protection in the Development and Food Security Boxes versus Investments in the Green Box Eugenio Diaz-Bonilla, Xinshen Diao and Sherman Robinson

207

11 That Was Then but This Is Now: Multifunctionality in Industry and Agriculture Eugenio Diaz-Bonilla and Jonathan Tin

235

12 Trade in Genetically Modified Food: Promises and Pitfalls for the Poor Chantal Pohl Nielsen and Karen Thierfelder

261

13 Is the Everything But Arms Initiative the way to go for Least Developed Countries in the WTO Negotiations? Wusheng Yu and Trine Vig Jensen

282

14 New Regionalism in the Aftermath of Cancún: to the Benefit or Detriment of Developing Countries? Chantal Pohl Nielsen

310

Index

333

Contributors

Bjørnskov, Christian, Department of Economics, Aarhus School of Business, Room P613, Prismet, Silkeborgvej 2, DK-8000 Århus C, Denmark. Cattaneo, Andrea, Economic Research Service, USDA, 1800 M Street NW Room S-4213,Washington DC 20036-5831, USA. Diao, Xinshen, Senior Research Fellow Division, Development Strategy and Governance, International Food Policy Research Institute, 2033 K Street, NW Washington DC 20006-1002, USA. Diaz-Bonilla, Eugenio, Executive Director for Argentina and Haiti, InterAmerican Development Bank, 1300 New York Ave., NW, Room NE1137, Washington DC 20577, USA. Frandsen, Søren Elkjær, Danish Research Institute of Food Economics, Rolighedsvej 25, 1958 Frederiksberg C, Denmark. Gersfelt, Birgitte, Danish Research Institute of Food Economics, Agricultural Policy Research Division, Rolighedsvej 25, 1958 Frederiksberg C, Denmark. Jensen, Hans Grinsted, Danish Research Institute of Food Economics, Agricultural Policy Research Division, Rolighedsvej 25, 1958 Frederiksberg C, Denmark. Jensen, Trine Vig, Ministry of Food, Agriculture and Fisheries, Direktoratet for Fødevare Erhverv, Nyropsgade 30, 1780 Copenhagen V, Denmark. Lind, Kim Martin, Danish Research Institute of Food Economics, Agricultural Policy Resource Division, Rolighedsvej 25, 1958 Frederiksberg C, Denmark. Melgaard, Poul Peter, Danish Research Institute of Food Economics, Agricultural Policy Research Division, Rolighedsvej 25, 1958 Frederiksberg C, Denmark. Nielsen, Chantal Pohl, Danish Research Institute of Food Economics, Agricultural Policy Research Division, Rolighedsvej 25, 1958 Frederiksberg C, Denmark.

vii

viii

Contributors

Orden, David, Markets, Trade and Institutions Division, International Food Policy Research Institute, 2033 K Street NW, Washington DC 20006-1002 USA. Robinson, Sherman, Department of Economics, School of Social Sciences, University of Sussex, Falmer, Brighton, BN1 9SN, UK. Thierfelder, Karen, Economics Department, US Naval Academy, 589 McNair Road, Annapolis, MD 21402, USA. Thomas, Marcelle, Markets, Trade, and Institutions Division, International Food Policy Research Institute, 2033 K Street, NW Washington DC 20006-1002, USA. Tin, Jonathan, 2000 Connecticut Avenue, #719, Washington DC 20008, USA. Walter-Jørgensen, Aage, Former Head of Agricultural Policy Research, Food and Resource Economic Institute, Agricultural Policy Research Division, Rolighedsvej 25, 1958 Frederiksberg C, Denmark. Yu, Wusheng, Danish Research Institute of Food Economics, Agricultural Policy Research Division, Rolighedsvej 25, 1958 Frederiksberg C, Denmark.

Preface

This book presents a selection of studies from a joint research project between the Danish Research Institute of Food Economics (FOI) and the International Food Policy Research Institute (IFPRI). The objective of the project has been to analyse the effects of developed countries’ agricultural policies on developing countries, mostly focusing on food security, poverty and other aspects of interest to the latter, as an input to policy reform scenarios within the WTO trade negotiations. The background for the initiative was the failure of the WTO Ministerial Conference held in Seattle in late 1999. This collapse, which generated much uncertainty about the future global trading system, showed significant divergences on trade and development perspectives and reinforced the need for research-based analyses of those issues. In contrast to past Rounds, by 1999 a large number of developing countries had become members of the WTO, and therefore the specific interests of these countries began to play a larger role in the negotiations. At the same time, civil society in both developed and developing countries became more interested in trade issues and the WTO. But the group of professional economists and trade analysts, on the one hand, and the variety of different groups with interests in trade and development, on the other, seemed to have been talking past each other both in the questions being asked and the answers provided. The joint FOI/IFPRI project has tried to address the main issues raised by those different groups through the application of quantitative techniques and presentation of the results in ways that we considered more usable by policy makers, farmers, NGOs and other interested parties in developing and in industrialized countries. The issues addressed, and the presentation of the material in the book, have mostly the latter audience in mind. This book is not aimed at trade economists and academic researchers in general, although some chapters might appeal to them. ix

x

Preface

The work presented here is in the tradition of the quantitative political economy to the extent that it tries to estimate quantitative impacts of policy choices on different types of countries and social groups within them, and analyses possible alliances and negotiating positions emanating from those diverse interests, focusing on developing countries (i.e. political economy issues in industrialized countries are not the main focus of this book). Both IFPRI and FOI have been using the research material extensively for their own outreach work and for active exchanges of policy analyses and recommendations with different governments, NGOs and producer groups, in industrialized and developing countries. The research has been conducted over the period 2000–2003, and it was concluded by a final workshop held in Copenhagen in February 2004. The project has been funded mainly by the Danish Foreign Ministry, DANIDA. The project has included collaborative work with other institutions in industrialized and developing countries and has also benefited from the participation of a Project Advisory Board, which included representatives from the Danish government, private agricultural and industry sectors and civil society.

1

Overview EUGENIO DIAZ-BONILLA,* SØREN FRANDSEN† AND SHERMAN ROBINSON‡ *Executive

Director for Argentina and Haiti, Inter-American Development Bank, 1300 New York Ave., NW, Room NE1137 Washington DC 20577, USA; †Director, Danish Research Institute of Food Economics, Rolighedsvej 25, 1958 Frederiksberg C, Denmark; ‡Professor, Department of Economics, School of Social Sciences, University of Essex, Falmer, Brighton, BN1 9SN, UK

Background While post-World War II international trade negotiations were effective in reducing industrial protection, spurring technological change and leading to accelerated growth, agriculture has remained highly regulated and protected in many countries throughout the world. The Uruguay Round trade negotiations (1986–1994) succeeded in embedding agriculture in the multilateral trade agreements, but the limited nature of the commitments (that still kept a separate treatment for agriculture) and disappointing implementation resulted in a slow pace of liberalization. Agriculture has consequently emerged again in the current Doha round as one of the most important and difficult areas of negotiation.1 In recent decades trade talks at the multilateral and regional level have been accompanied by heated debates about the costs and benefits for different countries and social groups of global trade expansion, and agricultural liberalization in particular. This book focuses on possible implications of these trends for developing countries. Do all developing countries benefit from industrialized countries’ liberalization in agriculture, or only some of them? How does trade liberalization impact net-food importers, or those countries benefiting from trade preferences that may be eroded by changes in trade policies? Are we considering adequately some of the non-trade concerns from the perspective of developing countries, including issues such as food security or multifunctionality? Are some developing countries better off in a world of preferential access to protected industrialized markets as opposed to multilateral negotiations within the WTO? What are the possible alliances and negotiating positions that result from answering these questions? © CAB International 2006. WTO Negotiations and Agricultural Trade Liberalization: the Effects of Developed Countries’ Policies on Developing Countries (eds E. Diaz-Bonilla et al.)

1

2

E. Diaz-Bonilla et al.

This book tries to contribute to that debate by organizing that analysis into four blocks of topics. The first block looks at developed countries’ agricultural policies and their impacts on developing countries. This is an input to the second block, which analyses possible scenarios of the WTO negotiations and potential coalitions of countries. The third block probes further into specific non-trade concerns from the perspective of developing countries. Finally, trade preferences, and their possible erosion under more liberalized world agricultural markets, are analysed. In order to place the analysis in context, the next section discusses the different groups, positions and alliances in the negotiations highlighting the issue of the heterogeneity across countries, both in agricultural structure and policies. The third section, organized around the four blocks of topics mentioned, provides an overview of the different chapters of this book and the policy issues involved. The final section ends with some conclusions.

Heterogeneity, Negotiating Positions and Alliances Since agricultural negotiations began there has been an active participation from WTO members, including many developing countries, which presented different proposals. Although there is a large variety of positions in the current negotiations, countries can be divided into four main groups (Fig. 1.1), using two criteria: whether countries consider agriculture requires a ‘special’ treatment or not, and whether those countries are high-income economies (with enough financial resources to subsidize agriculture) or not (a North–South axis).2

Industrialized countries The upper right quadrant in Fig. 1.1 is composed of the European Union, Norway, Switzerland, Japan and some higher-income developing countries like Korea. They argue that agriculture is special, in part because of the multifunctional effects created by agriculture.3 From this group only the European Union has presented a full proposal with quantitative commitments. The proposal basically maintains most of the instruments for the EU to protect and subsidize their agriculture and to compete through export subsidies in commercial markets. It provides fuller access to the EU market only to a limited group of developing countries while maintaining competitive exporters in that group under restricted access. Finally, it offers to the developing countries concerned about the competitiveness of their own agriculture an ‘I-protect-so-you-can-alsoprotect’ bargain. The USA and Canada (as well as Australia and New Zealand) appear in the lower quadrant, emphasizing the increasing need to integrate agriculture within the disciplines of the WTO. The USA and Canada,

Overview

3

South Several groups of developing countries

North EU, Japan, Korea, Switzerland, Norway

USA

‘Farmers Not Competitive’ Agriculture is special

‘Farmers Competitive’

Cairns Group 14 developing

3 developed

Agriculture treated as other sectors

Fig. 1.1. Different positions.

however, have at the same time different programmes to protect and subsidize their agriculture. In the case of the USA, its WTO proposal called for an ambitious programme of liberalization in agricultural trade, with hard caps on domestic support, substantial reduction in tariffs and the elimination of export subsidies. But several of the current policies in operation under the 2002 Farm Bill go in the other direction (see Orden, Chapter 4, this volume). The original US proposal, by suggesting a cap on trade-distorting support of 5 per cent of total agricultural production, maximized the adjustment in Europe’s domestic support (where the value of domestic support is larger in absolute terms and as a percentage of the value of agricultural production) while maintaining many options that the USA can utilize to continue offering domestic support to its farmers. The USA was also asking for restraints on state trading enterprises, such as the Canadian Wheat Board, which has made clear inroads into the US wheat market, while, as a competitive exporter further aided by domestic subsidies that will still be operational under its proposal, the USA wanted as much access as possible to developed and developing countries. This has been resisted by different industrialized and developing countries that fear an onslaught from US exports, both directly or indirectly subsidized. Also, the USA did not appear to have taken sufficient account of specific developing country concerns (see Lind and Bjørnskov, Chapter 6, this volume).

4

E. Diaz-Bonilla et al.

These different positions among industrialized countries can be put in better perspective by considering their differences in agricultural structures and policies. In terms of land structure, Japan and Korea have mostly small farms with a more egalitarian land distribution, while the USA and Canada have larger average holdings and more unequal land structure, and the European Union is somewhere in between (see Table 1.1). The USA and the EU (with 15 members) have similar shares of world agricultural production (measured in 1989–1991 world prices) at around 12–13 per cent and are far bigger than the rest of the industrialized economies. Looking at agricultural trade shares (Table 1.2), the USA has been losing its market share in world exports while maintaining its import share, and is expecting that the negotiations will help restore at least part of the lost export share. On the other hand, the change in the net trade position in the European Union during the 1980s – as a result of the impact of the Common Agricultural Policy (CAP) – has been one of the most dramatic developments in world agricultural markets: it significantly increased its Table 1.1. Land structure: average size of holdings and concentration. (From FAO, 2001.)

Africaa Asia Developingb LAC with Argentinac LAC without Argentina USA EUd Japan/Korea Canada

Average size (unweighted)

Average size (weighted)

Gini Index

2.92 2.20 87.09 32.53 186.95 27.27 1.12 349.07

1.20 1.62 63.25 27.66 na 17.91 1.15 na

0.53 0.57 0.82 0.82 0.64 0.59 0.47 0.74

na: not available. a Burkina Faso, Congo (Dem. Rep.), Djibouti, Egypt, Ethiopia, Guinea, Guinea Bissau, Lesotho, Libya, Malawi, Namibia, Reunion, Uganda. b India, Indonesia, Iran, Myanmar, Nepal, Pakistan, the Philippines, Thailand, Vietnam. c Honduras, Panama, Puerto Rico, Argentina, Brazil, Colombia, Paraguay, Peru. d Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, the Netherlands, Portugal, Spain, UK.

Table 1.2. Shares of exports and imports as a percentage of world agriculture. (From FAOSTAT.) Exports

Imports

1960–1980 1981–1994 1995–2001 1960–1980 1981–1994 1995–2001 USA European Union (15) Japan Korea, Republic of Canada

16.3 28.1 0.4 0.2 3.4

15.0 41.0 0.4 0.3 3.2

13.6 42.5 0.4 0.4 3.5

9.7 47.2 7.1 0.7 2.2

8.1 43.8 7.9 1.6 2.0

9.1 41.4 8.2 1.9 2.3

Overview

5

export share, while reducing its participation in world imports. The result of this shift in exports and imports has been that the EU’s net demand for agricultural products from the rest of the world, which amounted to about 30 billion dollars at the beginning of the 1980s (in current dollars), practically disappeared by the end of the 1990s (Diaz-Bonilla et al., 2002). Although the EU is still the world’s largest importer of agricultural products with more than 40 per cent of global imports (but less than half that percentage if intra-European trade is not counted), its net demand lately has been comparable to Korea’s and far smaller than Japan’s. Developed countries also differ in their agricultural policies (see Chapter 2 by Frandsen and Walter-Jørgensen on the European Union and Chapter 4 by Orden on the USA). The Nominal Assistance Coefficient (NAC)4, calculated by the OECD for different advanced economies, shows that farmers in Japan and Korea received, on average, 140–200 per cent more than the world market values in 2000–2003, the EU about 50–60 per cent, the USA some 20–30 per cent and Canada about 20–25 per cent (OECD, 2003, 2004).5 Support to agriculture in those countries has not decreased much, if at all, since the beginning of the 1990s: it reached some temporary lows around 1997 (due to high world agricultural prices in 1996–1997) but began to increase again afterwards. Besides the variations in levels, those countries also differ in the proportion of transfers coming from consumers or from taxpayers. This distinction is important, because the instruments utilized have different implications and treatment under the WTO legal system that considers separately market access, domestic subsidies and export subsidies.6 In Japan and Korea, transfers to farmers come mostly from consumers (more than 90 per cent during the period 2000–2003; see OECD, 2003, 2004). For the European Union and Canada, the share of consumer transfers in the value of producer support (not counting general services) is still around 50–60 per cent (although for Europe it represents a decline from about 80 per cent in the mid1980s), while for the USA it has dropped below 40 per cent in 2000–2003.7 These patterns imply that, for instance, Japan and Korea are more concerned in the WTO negotiations about market access liberalization, while the USA is relatively more interested in maintaining domestic support, with European Countries in-between, concerned both about market access and subsidies.

Developing countries Developing country members of the WTO have presented a large number of proposals and they show an even greater variety of growth patterns and structural characteristics. Yet they can also be divided into two main groups, depending on whether they consider that agriculture is ‘special’ and different from other productive activities considered under WTO rules (upper left quadrant of Fig. 1.1) or whether agriculture must be treated similarly to other goods (lower left quadrant of Fig. 1.1). This distinction is related to different negotiating positions: one is to ‘play defence’, asking for additional exemptions (i.e. special and differential treatment) to be able to

6

E. Diaz-Bonilla et al.

subsidize and protect agriculture in developing countries; the other is to ‘play offence’, trying to limit the ample legal room industrialized countries have under current WTO rules to protect and subsidize their own agriculture (for which rich countries also have large financial resources). Countries following a ‘defensive’ approach (upper left quadrant of Fig. 1.1) see their agricultural sector as vulnerable, and consider that agriculture is special and requires special treatment by the WTO. Although these perceptions may appear to put them closer to the ‘multifunctional’ arguments of some industrialized WTO members, those developing countries certainly do not want the special problems of agriculture in the South to get confused with the claims by rich countries, which have access to a full array of protectionist devices for their own agriculture. But while they want tighter disciplines on industrialized countries’ agricultural policies, they do not agree that agriculture should be treated similarly to industry. In particular, because they believe that rich countries will not truly and honestly reduce their levels of protection and subsidies (witnessing the tricks utilized during the implementation of the Uruguay Round – see Box 1.1), or because, even if that happens, they think that their agriculture would still not be competitive in a more open environment, they have asked for different ways to maintain protection for their agricultural sector. An approach has been to ask for broad exceptions under the label of a Development or Food Boxes,8 such as allowing developing countries (or a subset of them) to be able to impose higher tariffs for food security reasons, use quicker forms of safeguards and trade remedies against foreign subsidies and dumping, and retain protection for special products. An argument utilized in support of this line of reasoning is that the legal exemptions allowed for developing countries under the Green Box Annex and Article 6.29 of the Uruguay Round Agreement on Agriculture (AoA) are of no use to them, mainly because the policies permitted are very difficult to implement due to the financial, technical and human resource requirements (Solagral, 1999; UNCTAD, 2000). Some of those arguments appear to suggest that trade protection measures are simpler to implement institutionally and have no costs to the economy, compared to the budgetary expenditures required to implement Green Box and other policies allowed under the AoA (this issue is discussed in Chapter 10 by Diaz-Bonilla et al.). Within this group of developing countries there are also some members concerned about maintaining the perceived value of their preferential access to industrialized countries, which may decline if protection in those rich markets is reduced (the issue of preferences is discussed by Yu and T.V. Jensen, Chapter 13, and by Nielsen, Chapter 14). For those that are net food importers, there is the problem of possible increases in the value of food imports in a liberalized world (see simulations in Chapters 3, 5 and 7). Another (smaller) block of developing countries (lower left quadrant in Fig. 1.1) is part of the Cairns Group (14 developing countries out of 17 members in total)10. The Cairns Group, which includes a large group of Latin American and Caribbean (LAC) countries, has focused mainly on

Overview

Box 1.1. Uruguay Round Agreement on Agriculture (URAA) 1994 Market access Non-tariff border measures were replaced by tariffs that provided essentially the same level of protection (‘tariffication’). These tariffs, as well as other tariffs on agricultural products, were reduced by an average of 36 per cent in the case of developed countries and by 24 per cent for developing countries based on the average tariff rates during the period 1986–1988. Reductions were undertaken over a period of 6 years in developed countries (1995–2000) and over 10 years in developing countries. Least-developed countries were not required to reduce their tariffs. Establishment of minimum access quotas (at reduced-tariff rates) where current access was less than 3 per cent of domestic consumption. Minimum access tariff quotas were expanded to 5 per cent over the period 1995–2000. Special safeguard provisions allowed additional duties to be applied in case surging imports threatened domestic producers. Export competition The value of direct export subsidies was reduced to a level 36 per cent below the 1986–1988 base period level over 6 years (1995–2000), and the quantity of subsidized exports by 21 per cent during the same period. For developing countries, the rates of reduction were two-thirds of those for developed countries and the implementation period was 10 years. No reductions were required for the least-developed countries. Where subsidized exports had increased since the 1986–1990 base period, 1991–1992 could be used as the starting point for reduction although the end point remained the same. Domestic support Green Box Measures. Domestic policies that had, at most, a minimal impact on trade (research, disease control, investment in infrastructure, food security and others) were excluded from reduction commitments. Blue Box Measures. Direct payments under production-limiting programmes and assistance to agricultural and rural development in developing countries that represented only a small share of the production value (5 per cent in developed and 10 per cent in developing countries) were excluded from reduction. Amber Box Measures. Total domestic support provided on either a product-specific or non-product-specific basis that did not qualify for exemption was reduced by 20 per cent over the period 1995–2000. For developing countries, the rate of reduction was 13.3 per cent. Least-developed countries were exempted from reduction. Measurement of support The reduction of support builds on a common measure – The Aggregate Measure of Support (AMS) – which was established in connection with the URRA in order to identify the level of support and the reduction commitments by Member States. Source: WTO (2000a).

7

8

E. Diaz-Bonilla et al.

including agriculture in the disciplines of the WTO, with lower levels of protection and domestic subsidies, and the prohibition of export subsidies (‘playing offence’). One thing in common with many developing countries in the other group is their criticism of export subsidies as being unfair and disruptive of international trade. Also, the two groups of developing countries agree on the need to drastically reduce domestic support in rich countries, including the Blue Box and several of the distorting payments to farmers allowed under the current Green Box.11 But the Cairns Group does not consider appropriate the ample SDT provisions favoured by the first group of developing countries because they fear that such an approach could reinforce the protectionist ways of many industrialized countries, while also reducing trading opportunities in other markets. Therefore, the main difference in terms of negotiating strategies between the developing countries of the Cairns Group and those of the first block is the balance between ‘playing offence’ or ‘defence.’ Also, as indicated, there are important differences related to the loss of preferential access (see Yu and T.V. Jensen, Chapter 13 and Nielsen, Chapter 14). Which approach to favour has been a dilemma for many developing countries that, concerned about their own competitiveness, appear ambivalent between accepting the ‘I-protect-and subsidize-so-you-can-too’ offer from some industrialized countries, or, alternatively, challenging the distorting policies of the more advanced countries even though that may mean accepting some limitations in those developing countries’ own policies. As in the case of industrialized countries, it is useful to place those basic positions in the context of the variety of structural and policy conditions. While agriculture in LAC is less important as a percentage of the GDP, and rural population is smaller when compared to total population than in other regions, the region is an important structural exporter of food and agricultural products. Sub-Saharan Africa (SSA) and South Asia, on the other hand, show a greater importance of agricultural production and rural populations, but have smaller export shares (Table 1.3). At the same time, LAC’s agriculture appears more productive (per unit of labour), uses more capital (considering tractors as a proxy) and, after South Asia, is the region better served by roads (the large Amazon area in LAC affects the value of this indicator). Africa and LAC have more available arable land per capita than developing Asian countries (Table 1.3), but average holdings are larger, and land appears to be distributed more unequally in Latin America and the Caribbean than in Asia, with Africa in between (Table 1.1). It is important to note that SSA has an availability of land per capita that is comparable to LAC, but at the same time average holdings are of similar sizes to those in Asia, and the region shows the lowest values for the capital/technology and roads indicators, highlighting some of the opportunities and constraints for expansion of agricultural production in that region. Consequently, more countries in LAC appear to have decided to ‘play offence’ in the negotiations, not asking for special treatment for agriculture, while many African countries seem to have followed a defensive approach, with Asia in between.

Overview

Table 1.3. Structural characteristics. (From World Bank, 2002.) Latin America and Caribbean Agriculture, value added (per cent of GDP) Rural population (per cent of total population) Agriculture value added per worker (constant 1995 US$) Agricultural exports (per cent merchandise trade) Land use, arable land (hectares per person) Agricultural machinery, tractors per 100 ha of arable land Roads, km per square km of total area

7.9 26.5 2915.5 28.3 0.27 118.2 0.141

SubSaharan Africa 17.9 68.4 349.2 23.9 0.26 18.0 0.052

Middle East and North Africa 13.9 43.6 2163.6 4.7 0.21 117.8 0.062

South Asia 28.3 73.2 376.2 17.9 0.16 80.9 0.551

East Asia and Pacific 15.4 67.7 418.4 11.7 0.11 67.9 0.139

All developing countries

Least developed countries

13.2 60.6 589.8 15.3 0.21 102.0

36.7 76.4 239.0 35.3 0.20 8.0

0.123

0.044

9

10

E. Diaz-Bonilla et al.

Another aspect that underlies the positions of many African countries is the fact that agricultural production per capita has not been increasing in that region (particularly in SSA, see Table 1.4). Although all the developing regions are net importers of cereals and dairy products, Africa shows a larger percentage of imports of those products. Therefore, many countries in this region are concerned about food security, an issue that is behind or explains many of their defensive positions.12 On the other hand, in LAC and Asia, production per capita has been increasing, and LAC is a net agricultural exporter while Asian developing countries oscillate between positive and negative values (see Diaz-Bonilla et al., 2002, for a more detailed analysis). Looking at trade flows, developing countries as a whole export a larger share of agricultural exports to developed countries, but the shares and destinations differ by developing region (Table 1.5). Africa exports mostly to the EU (with an important component of preferential access) and to other African countries. The export partners of Latin American developing countries are mostly the EU and USA/Canada, followed by LAC countries, but with large differences from north to south in the continent. Developing countries in Asia, on the other hand, sell mostly to other developing countries in the same region, and, only after that, to Japan and the EU (Diaz-Bonilla and Reca, 2000). Regarding policies, developing countries tend to tax agricultural export products, while at the same time protect, through import taxes and other trade and non-trade barriers, some of the import-substitution-food crops. Due to budgetary constraints and policy changes since the mid to late 1980s, domestic subsidies in most developing countries tend to have lesser significance as a means of supporting local agriculture, while market protection has emerged as the main tool for transferring resources to that sector. The variety of situations in industrialized and developing countries in terms of agricultural structure, performance, trade orientation and policies means that the impacts of world trade liberalization are potentially very different by country, leading to different positions and alliances within the WTO negotiations. The evolution of these negotiations is briefly reviewed below.

Table 1.4. Agricultural production per capita (indices: base 1989–1991; From FAOSTAT.)

China Asia developing w/o China Africa developing SSA LAC Developing countries World

1960s

1970s

1980s

1990s

58.65 82.44 108.84 115.74 84.69 77.78 86.98

64.06 84.71 103.23 108.28 89.94 81.38 91.46

87.57 94.95 96.10 98.32 98.00 93.72 97.82

134.09 105.12 102.10 101.09 108.22 113.11 103.86

Overview

Table 1.5. Direction of agricultural exports (average 1990–1999; From WTO, 2001.) Destination

Origin Africa Asia LAC Middle East Total of all from developing regions

Industrialized countries

North America

EU(15) and EFTA(3)

70.7 51.3 68.3 48.2 60.7

5.4 8.7 25.8 4.2 14.9

57.0 16.8 34.0 36.2 30.2

Japan

Australia and New Zealand

Economies in transition

Developing countries

Africa

3.5 19.8 3.2 1.6 10.0

0.3 1.5 0.3 0.3 0.8

4.5 4.4 5.0 5.9 4.7

29.3 48.7 31.7 51.8 39.3

14.3 4.0 3.3 4.7 5.2

Developing Asia LAC 9.2 36.9 6.4 6.3 19.2

1.4 1.1 17.6 1.0 7.8

Middle East Other 3.3 6.2 3.7 37.2 6.3

1.0 0.5 0.7 2.7 0.8

11

12

E. Diaz-Bonilla et al.

Evolution of WTO negotiations Early in 2003, Stuart Harbinson, the then Chairman of the Agricultural Committee of the WTO, prepared a draft with the modalities to serve as the basis for the agricultural negotiations, trying to bring together the very many different proposals presented up to that point. Although the formal analysis of Lind and Bjørnskov (Chapter 6), suggests that it represented a true compromise, it was not accepted. Subsequently, in August 2003, the USA and the EU presented a joint document with a framework for the modalities. It was widely viewed as signalling a shift in the USA away from its original liberalizing posture and towards the EU position, in exchange for maintaining American subsidies of the 2002 Farm Bill. The convergence of the EU and the USA (across the North–North axis of Fig. 1.1) generated a parallel South–South reaction among different developing countries, leading to the creation of the G-20, that brought together many Cairns Group members (Argentina, Brazil, Thailand, Indonesia, South Africa, Costa Rica) and several of the Development Box countries (Pakistan, Cuba, Tanzania, Egypt), along with countries such as China, India and Mexico – also concerned about the vulnerability of their agriculture. The Cairns Group developing countries accepted a larger component of defensive policies, while the other developing countries decided to reduce their initial aspirations for a stronger SDT component while focusing more on reducing protection and subsidies in industrialized countries. This group challenged the US–EU proposal, following the same format of that document but changing the content: it asked for the elimination of export subsidies, stricter disciplines in domestic support and meaningful market access in the protected markets of the industrialized countries, while considering more generous SDT for developing countries. Developing countries then separated mostly into the G-20, on one hand, and the African group (and other countries with preferential access mainly to the EU), on the other. The latter appeared to have feared the potential negative terms-of-trade effect arising from the possible erosion of preferential market access coupled with projected increases in the price of food imports in liberalized world agricultural markets. The main topic in which they challenged industrialized countries’ policies was related to cotton subsidies – mostly from the USA, affecting Benin, Burkina Faso, Chad and Mali (all of them least developed countries).13 They sought compensation for past subsidies and elimination of future subsidies. This issue certainly did not affect either the perceived rents gained by developing countries from protection in industrialized countries’ markets, or the price of food items in world markets. In the end they opposed any other commitments in rules (such as the Singapore issues of competition, investment, government procurement and trade facilitation). In September 2003, in the Cancún Ministerial, a new framework document, usually referred to as the Derbez Text, emerged during the negotiations, but the different positions could not be bridged, and the talks

Overview

13

collapsed. Almost a year later – in August 2004 – the so-called ‘July Package’ was finally approved; this defined a framework for negotiating modalities in agriculture, as part of a larger effort to secure advances in all areas of negotiations. All these frameworks (those originally proposed by the US–EU on the one hand and the G-20 on the other, the Derbez Text and the July Package) do not have the quantitative precision of the Harbinson Proposal, and it is therefore not possible to estimate their probable impact. As of this point in time the modalities for the agricultural negotiations have not yet been agreed. The expectation that some pre-modalities may be approved in mid-2005, with full modalities agreed in the Hong Kong Ministerial meeting of December 2005, has not materialized. Whatever the negotiating positions, an important analytical issue is to determine the welfare and other effects of following different approaches to reform current agricultural and trade policies. A summary of the analysis of those policy issues in this book is presented next.

Policy Debates As indicated, the policy analysis has been organized into four blocks of topics. First, to help policy makers, negotiators and interested groups in developing countries that may not have full access to the information, there is a section (Developed Countries’ Agricultural Policies and Their Impacts on Developing Countries) with chapters documenting agricultural policies of the European Union (Chapter 2) and of the USA (Chapter 4). The focus is on these two main actors because of the complexity of their policies and the fact that they are big world players in agricultural markets. Other countries, such as Japan, have simpler trade and agricultural policies, based mostly on border protection, and are not big world producers. Chapters 3 and 5 present estimates of the impacts of changes in those policies on different groups of developing countries. While those chapters analyse existing policies, the next section (WTO Negotiations and Coalitions) speculates about negotiating alliances that may change applied policies (Chapter 6) and looks at possible outcomes of the negotiations and their impact (Chapter 7). The third section (Developing Countries’ Interests and NonTrade Concerns) probes further into several specific topics such as food security (Chapters 8, 9 and 10), multifunctionality (Chapter 11), and biotechnology (Chapter 12), from the perspective of developing countries. The fourth section (WTO Negotiations and Preferential Trade Agreements) looks at the issue of trade preferences and their possible erosion under more liberalized world agricultural markets. The answers to those different issues have implications for the political economy of the negotiations, mainly related to the emergence and consolidation of possible alliances across countries with specific policy packages favoured by them. In order to discuss these topics, the chapters in this book utilize different methodologies and approaches. Some of them (Chapters 3, 5, 7, 10 and 13) apply specific computable general equilibrium (CGE) of the

14

E. Diaz-Bonilla et al.

world economy to discuss the trade simulations considered. Also, Chapters 12 and 14 are overviews of results from different modelling exercises (some, but not all, of them from IFPRI and FOI) utilizing CGE models. In general, all those models are based on a similar database (the Global Trade Analysis Project or GTAP)14 and have a common basic structure of interlinked markets. However, those models also present differences (discussed in the different chapters) in the policy experiment analysed, how agricultural policies are modelled, in the level of disaggregation at the country and product levels, and in the macroeconomic closure rules utilized, including the treatment of labour markets, among other things. In interpreting the results, readers must be aware that this category of models is ‘empirical’, in the sense that the models are based on the best available data on the structure and flows of the economy of the countries involved, and utilize parameters, such as different elasticities, selected from different econometric analyses and validated through use by the members of the GTAP consortium.15 Yet these simulations are not ‘projections’ that may be evaluated statistically. They try only to isolate a policy change (trade liberalization of different kinds) while keeping the rest of the policy setting at certain pre-specified levels for the experiment analysed. A different evolution of the non-trade policy framework (such as devaluations or modifications in non-trade taxes or public expenditures) may lead to different trade and policy outcomes. Therefore, the results of those models must be interpreted as conditional simulations. Notwithstanding the differences in the models utilized and reviewed in this book, the fact that some results, presented below, are replicated across simulations provides a control for robustness. Given that the nature of the policy questions varies across chapters, the book also utilizes other methodologies such as cluster analysis (Chapters 6 and 8) and econometric estimation (Chapter 9), while the remaining chapters (2, 4 and 11) are mostly technical descriptions and/or qualitative literature reviews.

Agricultural policies of the European Union and the USA and implications for developing countries As discussed before, any study of the implications of policies in industrialized countries for the rest of the world has to acknowledge the heterogeneity of policies in the countries involved. Chapter 2, by Frandsen and Walter-Jørgensen, reviews in detail the convoluted structure of the Common Agricultural Policy of the European Union, while in Chapter 4, Orden discusses the no less complex details of the US Farm Bill. But what is the impact of these policies on developing countries? A key issue is how delinked from production are the new policy instruments of decoupled support, and whether they are combined or not with proposals to allow meaningful market access.

Overview

15

In Chapter 3, Gersfelt and Jensen analyse the impact on Africa of the recent changes of the CAP – the so-called Medium Term Review (MTR), aimed at partially decoupling domestic support, and of the Eastern Enlargement of the EU. They show that the effects of those changes are positive for Europe, but may be negative for Africa in terms of trade and welfare, although the overall effects are generally small. The analysis also showed that in many cases the effects of the 2003 MTR reform dominate the effects of the Eastern Enlargement. The authors find that the decoupling support is clearly a step forward: to the extent that it reduces EU agricultural policy distortions, there are welfare gains for the EU due to a more efficient economy, and it allows a better targeting of transfers to income support rural development and to environmental concerns. Also, reducing the incentives for overproduction in the EU of cereals, bovine animals and milk benefits mostly countries that are major producers of these commodities, but the developing countries that do not have comparative advantages in those products may not benefit or could eventually be hurt. However, simulations show that even in the case of negative welfare effects, they are far less than the EU benefits, allowing the possibility of compensations to the countries that may be affected. The authors consider that the MTR is a contribution to the WTO negotiations, but probably will not be sufficient for a compromise to be reached, in part because the decoupling is only partial and, mainly, because what counts most for developing countries’ welfare is market access to the EU. Decoupling production must be accompanied by greater market access to have really important positive impacts on developing countries. In fact, one of the implications for developing countries of this analysis is that they should worry less about the level of domestic support, provided it is decoupled, and focus more on market access to industrialized countries, considering that truly decoupled support policies are compatible with open markets. Chapter 5, by Robinson and Thierfelder, includes an analysis of the impact of US policies, and those of other OECD countries, on developing countries. Contrary to most CGE simulations that assume full employment, this chapter follows a long tradition in development analysis (see, for instance, Lewis, 1954), and models labour markets in developing countries with significant unemployment and non-flexible wages. General equilibrium models with full employment can only expand agricultural production by taking labour from non-agricultural activities. Therefore, the positive multiplier effects on the rest of the economy of an expanded agricultural sector are assumed away under full employment specifications. The model also tries to characterize agricultural policies in greater detail by identifying five types of agricultural programmes, including two types in which the payments are not exogenously fixed price-wedges, but the payments rate varies endogenously depending on market conditions. A case in point is the marketing loans in the USA that provide payments to farmers that increase as market prices decline below the loan rates. The chapter also briefly reviews the literature on the effects of decoupled income transfers on agricultural production, and argues that the emerging consensus in the USA suggests that these payments do not seem to have significant production

16

E. Diaz-Bonilla et al.

effects because of the efficiency of US agricultural markets for land, labour and capital, and for managing risk, although this conclusion may change in other countries with less efficient markets. It seems clear that more empirical research is needed in the USA and elsewhere on the impact of decoupled income transfers. Also, the structure of the model (a single aggregate household; no labour–leisure choices; and no intertemporal savings and consumption) does not allow the treatment of other dynamic issues related to income transfers. These are areas of model development that are likely to become increasingly important for agricultural policy analysis. An implication of modelling programmes endogenously is that there can be strong interaction effects when simulating the separate removal of each type of agricultural policy – tariffs, domestic support and export subsidies. The separate reforms are not additive: the change of agricultural support programmes cannot be done in isolation, but rather requires a comprehensive effort. Such comprehensive reform would be greatly facilitated if OECD countries moved toward a system of decoupled income support for their farmers. This would allow market access to developing countries and would meet their demands for not having to face unfair trade competition from production and export subsidies in world markets. Similarly to the previous chapter, this chapter finds that eliminating domestic support programmes without change in market access has little impact on trade with developing countries. Market access is much more important for developing countries than domestic support programmes within the OECD countries, but the two are linked and the developing countries cannot get more access unless the OECD countries achieve coordinated reform of their agricultural support programmes. Once increased market access in the OECD countries is allowed, agricultural exports from the developing countries rise dramatically in the simulations, while their agricultural imports decline modestly. Another finding is that adding trade liberalization among the developing countries to OECD policy reform increases agricultural exports from developing countries by another 50 per cent. In a scenario of comprehensive agricultural policy reform in both OECD and developing countries, two-thirds of the gain in exports for the developing countries is due to OECD reform, while one-third is due to the removal of barriers against trade among the developing countries. The authors also find that unilateral agricultural policy reform by either the USA or the EU is politically difficult because it would lead to declining farm incomes in the reforming region, while the non-reforming region would gain market share. However, all countries gain from a mutually supporting programme of global reform.

WTO negotiations In Chapter 6, Lind and Bjørnskov utilize cluster techniques to classify the negotiating positions of a large number of WTO members. They update an early analysis covering country positions up to November 2001 (Bjørnskov

Overview

17

and Lind, 2002) and include an analysis of the March 2003 Harbinson Proposal that intended to be a compromise between the different proposals presented up to that point. The authors ask three main questions: (i) which are the groups of countries that can work together in these negotiations; (ii) whether the positions of the WTO members have converged since the first exercise (up to November 2001); and (iii) whether the Harbinson Proposal reflects a genuine compromise across the different positions. After ranking the positions along 14 dimensions and using cluster analysis to classify countries, they identify some more homogeneous groups. Most of the African group appear to have similar objectives and are grouped together. Most of the rest of the developing countries seem to be closer to positions such as those of Canada and the Cairns Group in general. On the other hand, the European Union and a small group of countries with highly protected agriculture are grouped in a separate cluster at some distance from the rest of the WTO members. In this regard, the authors reconfirm the main finding in Bjørnskov and Lind (2002) that the EU appears isolated in its demands in the round. Regarding the second point, they find mixed evidence on the convergence of the negotiation positions, with some developments in key countries (such as the 2002 US Farm Bill) moving positions further away from a compromise (see Orden, Chapter 4), while some other clusters, mostly of developing countries, appear to be moving closer together. Finally, the Harbinson Proposal, that appears in the average cluster, seemed to express a genuine compromise. They conclude that the success or failure of the Doha Round depends on whether a limited number of rich countries come closer to the consensus reflected in most other positions. While the previous chapter showed, among other things, that the Harbinson Proposal was part of the average cluster, Frandsen et al., in Chapter 7 assess that proposal in terms of the welfare, production and trade impact in a world model.16 The authors estimate that global real income may increase by 100 billion US dollars per year by the year 2013 (in current 2003 prices). About a quarter of those benefits accrue to the European Union-25 itself, due to a better allocation of resources and consumers’ gains from lower prices. The USA would receive 15 per cent of world expanded real income while other developed countries (Canada, Japan, Australia and New Zealand) represent another 40 per cent of the gains. This leaves only about 20 per cent of the additional real income to accrue to a relative large number of developing countries, including the LDCs. However, in relative terms those real income gains are larger as a percentage of the economies of developing countries than the larger absolute values are as a percentage of the developed countries’ economies. Also, the simulations imply that most of the economic gains, mostly in industrial countries, arise from countries’ own liberalizations. The results from the modelling exercise also suggest that while the efficiency gains from a better reallocation of resources of production is in general a positive, for some countries there may be a negative in terms-oftrade effects (the ratio of export to import prices), which affect general

18

E. Diaz-Bonilla et al.

welfare. This price effect depends on the net trade position of those countries and on the value of possible preferential access opportunities to protected developed markets that may be eroded in any trade liberalization scenario. The simulations show that developed countries such as the USA and those of the EU could compensate fully those developing countries that may have welfare losses due to erosion of preferences or adverse terms-oftrade effects, and still be better off, by combining trade liberalization and development aid. In particular, the welfare losses in the LDCs are relatively minor (some 200–250 million dollars a year) when compared to the joint EU–US gain of more than 40 billion dollars. Also, coupling those compensations with domestic reforms in the developing countries that can achieve the dynamic gains indicated before would more than offset the possible negative impacts of changes in world market prices and the erosion of trade preferences.

Developing countries’ interests and non-trade concerns An important non-trade concern for developing countries is food security, and there has been a heated debate about what is the best policy approach to address this issue in the WTO negotiations. Chapters 8 and 9 provide different angles on this debate. For the forthcoming negotiations to consider in detail food security concerns under WTO rules, at least two issues need to be addressed. The first is the relevance of the current classification of countries in the WTO of developed, developing, least-developed (LDCs) and net-food-importing developing countries (NFIDCs) with respect to their food security status. Of these categories, only the NFIDCs are defined with respect to a particular food security indicator (i.e. net food imports), although, as argued below, it may not be the most appropriate. The second issue is whether the current legal texts, which define WTO commitments on the basis of these categories of countries, really address the issue of food security through that differential treatment. Both questions are linked: if the categories are badly defined to capture food security concerns, then it is unlikely that the differential treatment under WTO rules will deal with those concerns in a meaningful way. But even if these categories capture the variety in the situations of food (in)security, the question of the adequacy of current and future WTO rules and commitments to treat these differences must still be answered. Chapter 8, by Diaz-Bonilla et al., contributes to the first issue of the adequate classification of countries, using various dimensions of food security and three clustering methods. A methodological novelty is the application of the theory of ‘fuzzy sets’ in conjunction with more traditional methods of cluster analysis. This study classifies 167 countries encompassing all levels of income into 12 clusters using five indicators of food security: food production per capita, the ratio of total exports to food imports, calories per capita, protein per capita and the share of the non-

Overview

19

agricultural population. Developing countries appear scattered across all levels of food security and insecurity, except in the very high food-secure group, while developed countries are all in food-secure clusters. Therefore, granting food security exceptions to developing countries as a whole fails to discriminate among them.17 Regarding developed countries, the typology confirms the common sense perception that all of them are food secure, weakening the claims of some industrialized countries related to food security concerns as the basis for their use of protection and subsidies. Also, the NFIDCs split about 60 per cent/40 per cent between food-insecure and food-neutral groups. Being a net food importer appears to be only a weak indicator of food vulnerability.18 LDCs, on the other hand, include mostly countries suffering from food insecurity, but some countries that appear in food-insecure categories are neither LDCs nor NFIDCs. Limiting the special and differential treatment related to food security problems only to LDCs or to food-insecure NFIDCs would leave them out. The main conclusion for the negotiations is that the current WTO classification of countries, although a starting point, may not be the most adequate one to capture food security concerns. The authors also identify some policy conclusions from the different profiles of some food-insecure countries. Some are predominantly rural (mostly in Africa and South Asia), whereas for others the urban population is more important (like many countries in Latin America and those that were part of the former Soviet block). Obviously, the same policy (such as maintaining high prices for producers because of security concerns) will have different impacts in these two types of countries. Also, in some cases countries are food insecure mostly because of low levels of calories and proteins per capita, although they do not use large percentages of their exports to buy food, while other countries are a mirror image: they use a large percentage of their exports to buy food but their current levels of calories and proteins per capita are close to the average for all countries considered. Again, the policy options for these two types of countries are different, to the extent that the first group may increase imports to improve availability of calories and proteins, whereas the second group appears more constrained. Whatever the classification of developing countries and the possible special and differential treatment related to food security, there is a concern regarding whether further liberalization of world agriculture may lead to the decline of food stocks (because of reductions in production subsidies, mainly in industrialized countries), thus endangering world food security, particularly in more vulnerable countries. Lind, in Chapter 9, analyses anew the issue of what should be the adequate level of food stocks for insurance purposes. He utilizes a stochastic approach based on calorie availability, and applies the results to Sub-Saharan Africa, one of the developing regions suffering the most from food deficits and variability in food supply. The conclusions suggest the need for lower levels of stocks than those usually recommended by studies which follow the traditional methodology of projecting requirements, taking as given those historical stocks created by

20

E. Diaz-Bonilla et al.

agricultural subsidies and protection (see FAO, 1983). This chapter also studies the viability of a financial fund for food security purposes, as an alternative to maintaining physical stocks. It finds that the savings in maintenance and management of the latter would endorse implementation of the approach based on a financial fund, although the logistical issue of how to physically move the food to affected regions remains. In the negotiations some developing-country members of the WTO have proposed expanding the special and differential treatment options available to them arguing that ‘one size does not fit all’, and that they need more expanded instruments to promote their own agriculture. That special and differential treatment has been packaged under the notions of a ‘Development Box’ or a ‘Food Security Box’. Part of the issue, as discussed before, is that if the WTO is to address food security concerns, it will need to improve its current classifications of countries. The other part of the discussion is what should be the content of that special and differential treatment. There have been different proposals (under a Food Security Box and similar ideas) that argue the need for increased agricultural protection, or at least for some special crops, as a way of promoting food security. As indicated before, these suggestions are usually accompanied by the misleading argument that protection ‘does not cost money’ and is easier to implement in poor countries than alternative policies such as investments in agricultural technology, extension and infrastructure. In Chapter 10, Diaz-Bonilla et al. evaluate this debate related to the protection component of the Development or Food Security boxes in a world trade model. As the model considers technological change in the production functions for agricultural sectors linked to government expenditures in agricultural research, the authors can also analyse alternative policies, such as investments in agricultural technology that are allowed under the current Green Box. This chapter considers two scenarios. In the first one import taxes on grains (as representative of food security crops) are increased by an arbitrary amount (selected just to make the simulation comparable) in the countries/regions that support the notion of a Development Box or a Food Security Box. Then, the cost of that increase in import tariffs is calculated as a tax equivalent, based on the consumption of those commodities for which protection is increased and on the price wedge between world prices and higher domestic prices. In the second scenario, this protection cost is transformed into an explicit tax on consumers, collected by the government, for the same amount, and is utilized to finance additional investments in agricultural technology. The first simulation shows that the increase in agricultural protection results in a negative but relatively small effect on GDP and employment for the whole economy. Although there is more production of the goods protected, there is less consumption and utilization overall, because prices increase and imports are curtailed. Production of the industries depending on the protected crops also shrinks. Measured by consumption of food and not by production, food security declines with increased protection. Agricultural trade among developing countries,

Overview

21

including those applying the higher levels of protection, also declines by about 300 million dollars in the simulations, suggesting that concerns that Development of Food Security Boxes could hurt South–South agricultural trade may be well founded. The numbers are not large, but they indicate the direction of the possible results of generalized use of protectionist approaches to address food security concerns. In the second scenario, an increase in investment in agricultural research financed by an equivalent tax calculated from the first scenario shows consistent increases in GDP, employment, agricultural production and consumption in general, including, particularly food. These results refer basically to increases in permanent protection (as different from contingent protection), and do not necessarily invalidate the need for temporary policy instruments (such as a special safeguard or streamlined countervailing procedures) to confront sudden negative trade shocks and unfair trade practices.19 While the notion of multifunctionality has been advanced mainly by developed countries, Diaz-Bonilla and Tin (Chapter 11) explore possible implications of that notion for developing countries. The authors compare the current notion of multifunctionality in agriculture with the set of ideas behind support for industry in developing countries, particularly in the first decades after World War II. They argue that, presently, developed countries invoke similar arguments for agriculture. Although it is clear that subsidization and protection of agriculture in industrialized countries is imposing costs on their own economies, as inefficient industrialization did in the case of several developing countries, rich countries are in a better position to absorb the costs of expanding the multifunctionality of agriculture beyond its non-subsidized levels. However, while protected industrialization in developing countries affected mostly those pursuing such strategies but not the world in general, the expansion of the agricultural sector in industrialized countries fostered through protection and subsidies affects agriculture globally due to its larger share in the world, hurting agricultural production in countries that do not have the resources to subsidize that sector. The level of production displaced in developing countries due to subsidies and protection in developed countries may be substantial (see also Gersfelt and Jensen, Chapter 3, Robinson and Thierfelder, Chapter 5 and Frandsen et al., Chapter 7). This chapter concludes that the arguments in support of a legal treatment under the WTO that supports and does not hinder rural and agricultural development in developing countries do not need new and debated notions such as multifunctionality, and can be more effectively based on traditional arguments linked to growth dynamics, poverty alleviation, food security and environmental issues, as they apply to those less favoured countries. Further, the notion of multifunctionality may not only be unnecessary for developing countries to support the policies needed for rural development, but may also be harmful. This would be the case if it were utilized to justify additional support to agricultural production in industrialized countries. In that case, agricultural production in developing countries (and the multifunctional effects linked to it) would contract even

22

E. Diaz-Bonilla et al.

further because of the excess of subsidized production in industrialized countries. The authors argue that the proponents of ‘multifunctionality’ must face the fact that an important effect of those policies in rich countries is that agriculture in other countries, many of them poor, is forced to contract. While Chapter 10, by Diaz-Bonilla et al., emphasizes the importance of investing in technology to create a dynamic agricultural sector in developing countries, Nielsen and Thierfelder, in Chapter 12, further discuss the challenges and opportunities of modern biotechnology for developing countries. The authors address a rather contentious international food trade issue, namely the production and export of genetically modified (GM) food by a number of large countries such as the USA, Argentina and China. On the one hand, developing countries need to increase productivity to help with food security and nutrition concerns. On the other hand, developing countries worry about access to their traditional markets if they increase their involvement in GM products and there is a subsequent consumer backlash in their key export markets. There are strong political economic forces that are causing different countries to adopt very different national approaches to the regulation of GM products, including import bans. Moreover, at the international level it is not yet clear whether and how trade in GM products should be regulated. In particular, it is possible that the so-called Cartagena Protocol on Biosafety, which came into force in September 2003 and is supposed to ensure safe transboundary movement of such products, may conflict with existing WTO agreements. The GMO (genetically modified organism) issue is almost certain to affect the way the WTO’s Sanitary and Phytosanitary (SPS) Agreement is interpreted and, should it come under sufficient pressure, re-negotiated. In addition to a review of national and international GM regulations and their potential areas of conflict, the chapter summarizes the results of a number of simulation studies about the trade and welfare effects of SPS policy reactions to the GMO issue. The models surveyed in this chapter differ in country and sector focus, the assumptions about how GM technology affects markets, the possible shifts in consumer attitudes toward GM commodities, how segmented are markets, what are the productivity gains from the new technology, and how this technology disseminates across countries. Those models incorporate the limited empirical evidence available and use educated speculation about how these trends might evolve in the future. Notwithstanding all those differences, the authors reach some broad conclusions. First, adopters of the more productive GM technologies gain, while non-adopters or regions that restrict GM technologies either do not gain at all or have only small benefits (through changes in world prices), depending on how strongly they segment their markets. Second, under the assumption of moderate costs of labelling and market segmentation, world markets adjust to the introduction of the new technologies, with significant, but not dramatic, changes in prices, production and trade flows. Third, effects on developing countries, while following the adopter/non-adopter pattern mentioned above, also reflect the size of their domestic market and

Overview

23

their trade orientation. Large developing countries, like India and China, gain from the new technology and are not much affected by changes in trade regime. Countries in sub-Saharan Africa, on the other hand, are very dependent on EU markets and, if they decide to expand GM production, could potentially be strongly affected by EU trade restrictions. The authors emphasize the need for further empirical research in order to have a clearer view of the future of the new technology, including several key issues such as: what are the potential benefits (in terms of new traits, productivity and cost gains) of the next generation of GM goods; what are the potential costs associated with labelling requirements, identify preservation, and segmentation of markets; what are the implications of different intellectual property rights regimes and the evolution of market structures for the dissemination of the new technologies and for the distribution of the benefits; what will be the reaction of consumers; and how will the institutions of the world trading system, including the WTO, adapt to the evolution of these new technologies. From the perspective of the developing countries the authors emphasize the need for further research to develop productivity-enhancing crops that are specific to problems facing developing countries, and the possible intervention of the public sector in subsidizing relevant research and in entering into partnerships with the private sector to make the technology available and affordable for poor small-scale farmers. Also, the public sector in developing countries must improve its own institutional and regulatory system related to assessment and management of risks to human and animal health, the environment and biodiversity; to anti-trust legislation; and to securing of intellectual property rights and farmers’ rights. Developing countries will require financial and technical support from donor countries to develop the capacities needed to tackle the challenges of modern biotechnology.

Erosion of preferences and preferential trade agreements The issue of preferential agreements looms large in current negotiations. First, looking at existing agreements, erosion of preferences, as discussed before, is a main concern for several developing countries. Second, and now considering the future, industrialized countries have been actively pursuing bilateral preferential trade agreements, and the USA, in particular, has indicated that, if there are no advances in the WTO negotiations, it is prepared to follow the bilateral approach. Third, current or future preferential agreements may create a constituency for protection and tradedistorting practices (see Diaz-Bonilla et al., 2003). The two chapters in this section look at different aspects of this topic. In Chapter 13, Yu and T.V. Jensen analyse the ‘Everything But Arms’ (EBA) initiative by the European Union, aimed at providing free market access to the LDCs. The authors find that EBA has effects on only a limited number of products, mainly food and agricultural goods

24

E. Diaz-Bonilla et al.

(including sensitive products such as sugar, rice and bananas), due to the fact that preferences on other products have already been granted to the LDCs prior to the EBA. In addition, current exports of the EBA products from the LDCs to the EU are small. Therefore, they find only limited positive effects for EBA in the concerned LDC countries: total welfare gains are calculated at less than US$300 million for all the LDCs concerned in the study. Furthermore, most of the gains come from the three sensitive products that are subject to gradual liberalization over time, especially sugar. On the other hand, the negative impact on the EU and developing countries seems to be quite small, indicating that granting quota and duty-free access to the LDCs does not cost the EU very much and does not pose a big threat to exports from developing countries. The authors clarify that the quantitative assessment of the benefits of the EBA is based on the assumption that the safeguard provisions and the rules of origin attached to the EBA initiative would not play the role of restricting exports from the LDCs. If the safeguard clause is invoked (say for sugar), then even the modest benefits of the EBA shown in this study may not materialize. The authors also analyse the issue of erosion of preferences: what would happen if the EBA were implemented and the WTO negotiations led to agricultural liberalization? They find that the EBA preferences will be diminished under different plausible WTO reform scenarios in which the EU cuts its market access barriers and subsidies. In fact, some LDCs may suffer welfare losses under those scenarios. However, the authors argue that using this result as an argument for maintaining protection in the EU or in other industrialized countries would cause much bigger welfare losses to the whole world and, ultimately, would also harm the LDCs. As in other simulations discussed, the reforms carried out by the EU are estimated to generate sizeable gains for the world as a whole, which far exceed the potential losses of some countries. Based on the quantitative analysis, the authors conclude with two sets of policy implications from the perspective of the LDCs. First, regarding the EBA initiative, they consider it important that the trading opportunities provided through the EBA are available to the LDCs without the uncertainties and complexities associated with the safeguard measures and rules of origin of the EU. LDCs should negotiate unconditional quota and duty-free access to the markets of the EU and other developed countries within the current Doha Round, and support the abolition of special safeguard provisions of Art. 5 of the Agreement on Agriculture. Second, concerning the possible negative impact of the WTO reforms, LDCs should insist on receiving the technical and financial assistance and cooperation needed to solve the many supply side constraints they face. The authors argue that the EU and other developed countries are in the position to offer such assistances as they stand to benefit the most from further multilateral liberalization. Nielsen, in Chapter 14, analyses the more general issue of the desirability of preferential trade agreements from a developing country perspective and in light of the ongoing multilateral WTO negotiations.

Overview

25

Notwithstanding the theoretical arguments and empirical evidence in favour of a multilateral approach to global trade liberalization, the present world trading system is characterized by a complex web of regional and preferential trade agreements. Based on the survey of recent empirical studies of selected PTAs, the author reaches several conclusions. First, the welfare impact on participating countries is generally non-negligible and positive, but tends to be small. Second, trade diversion can be an issue, particularly for specific sectors and countries. Third, the greatest gains for developing countries lie in trade liberalization of products that are politically sensitive for the preference-giving countries. Fourth, multilateral liberalization is preferred, and OECD agricultural policy reform is also desirable, but then erosion of preferences emerges as an issue. For all the uncertainties, the author concludes that both economic theory and quantitative empirical analyses can provide guidelines as to how real-world preferential trade agreements should be constructed so as to help achieve the benefits of trade liberalization in a second-best world.

Some Policy Conclusions An overall conclusion (shared with studies before, during and after the Uruguay Round) is that the combination of domestic support, market protection and export subsidies in industrialized countries appears to be displacing agricultural production and exports from developing countries (Chapter 3, by Gersfelt and H.G. Jensen; Chapter 5, by Thierfelder and Robinson and Chapter 7, by Frandsen et al.). A substantial reduction of those distorting policies should allow developing countries to develop a more dynamic agricultural sector, presumably with positive implications for poverty alleviation and food security.20 However, expansion of agricultural production does not guarantee that total welfare effects for the country as a whole are positive. In fact some simulations, including some in this book (for instance, Chapters 3 and 7), suggest that even though agricultural production in developing countries would expand with agricultural liberalization in industrialized countries, there may be aggregate negative welfare impacts for some developing countries that are net food importers and/or have preferential access to protected markets in rich countries. The short-term, static impact of liberalization in industrialized countries may have a negative terms-of-trade effect on some developing countries, if, for instance, they have access to high-price industrialized countries’ markets for some products (say sugar) and buy cheap subsidized temperate products (say wheat) from developed countries. Simulations that try to capture ‘pure’ trade effects (usually by utilizing full employment assumptions or not allowing for capital accumulation, land expansion or technological change) show smaller positive results for developing countries as a whole from agricultural liberalization in industrialized countries, and then negative terms of trade may dominate other possible positive effects for a net welfare loss.

26

E. Diaz-Bonilla et al.

Should, therefore, agricultural trade liberalization in industrialized countries be slowed down or stopped because of these negative welfare effects on some developing countries, especially if they may affect some poor countries? This debate is at the centre of the formation of coalitions across developing countries. As mentioned, the convergence of the EU and USA generated a parallel South–South convergence among different developing countries, leading to the creation of the G-20 that brought together many Cairns Group members and several of the Development Box countries, along with countries such as China, India and Mexico. Developing countries then separated mostly into the G-20, on one hand, which is pressuring developed countries to liberalize their markets, and the African group (and some developing countries that have preferential access particularly to the EU market, and/or are food importers), on the other. The latter appears concerned about potential negative terms-of-trade effects arising from the possible erosion of preferential market access, coupled with projected increases in the price of food imports in liberalized world agricultural markets. The possible negative static result of agricultural trade liberalization for some developing countries has been highlighted early in trade studies (see, for instance Koester and Bale, 1984) and has received some attention lately (see Panagariya, 2004). This debate must be placed in its historical perspective. Depressed world prices of many food products during the 1980s and 1990s related to agricultural protectionism and subsidies in industrialized countries appear to have discouraged investments in the rural sector of many developing countries that came to depend on cheap and subsidized food from abroad, and contributed to turning many of them, like different countries in sub-Saharan Africa, from net food exporters into net importers. Even the World Bank and other development banks seem to have cut loans to agricultural and rural development projects influenced, in part at least, by low world agricultural prices that both reduced the expected returns of future projects and depressed the expost results of evaluated projects (Lipton and Paarlberg, 1990). Low food prices may have also pushed several developing countries into a more extreme specialization in tropical products than would have been the case under a different set of relative prices. Given that distorted starting position, the static negative terms-of-trade effect follows. A different set of prices, resulting from world agricultural liberalization, may well lead to an increase in developing countries’ production of temperate-zone staples and close substitutes, reducing or changing the net food import position. In addition, using simulations that go beyond pure trade effects and include employment multipliers (as in Chapter 5, by Robinson and Thierfelder) and/or allow for capital accumulation, land expansion or technological change, the number of developing countries with negative welfare effects declines significantly or disappears completely (see, for instance, World Bank, 2001; Diao et al., 2003). Besides, simulations in this book (Chapters 3, 5, 7 and 14) show that, even in the case of negative welfare effects, a better alternative would be to

Overview

27

proceed with the liberalization (given the large aggregate positive welfare effects in the industrialized countries and also at world level) and then compensate the cases of negative welfare impacts, particularly when they affect poor countries. This is particularly relevant in the case of preferences that can be more clearly quantified and compensated. Following the same approach applied by industrialized countries to compensate domestic producers, the value of trade preferences lost can be transformed into equivalent foreign aid flows for the poor developing country losing the preferences during a pre-specified period of time. In general, considering that agriculture and agro-industry are the main economic activities in many poor developing countries regardless of their net trade position, and that such activities usually have significant growth multipliers for the whole economy, the level of non-realized dynamic benefits for those countries due to distortions in industrialized countries’ agriculture may be substantial. A separate implication for vulnerable developing countries, in Africa and elsewhere, is the need to invest in their own agricultural sectors (which would require stepped-up funding from multilateral organizations and donors) to be able to capture the opportunities that may arise from reductions in current distortions in agricultural markets (Chapter 10, by Diaz-Bonilla et al.; see also Chapter 12, by Nielsen and Thierfelder on biotechnology). The persistence of protection and subsidies in rich countries also affects rich countries through higher food costs and a larger tax burden on citizens. US and EU farm policies also increase the costs to each other, suggesting the need for a coordinated approach to reducing distortions (see Chapters 5 and 7). Agricultural liberalization generates important gains in industrialized countries that, as argued, can be used to compensate the more limited cases in value and number, where some developing countries may suffer welfare losses due to terms-of-trade effects and erosion of preferences. Completing the unfinished business of the Uruguay Round, to allow for broad-based economic growth in developing countries and help materialize the important welfare gains in industrialized countries, would require that industrialized countries transform their domestic support towards truly decoupled approaches, along with important reductions and hard caps by commodities (as percentage of the value of production) on trade-distorting support; provide meaningful market access on a MFN basis along with full free access (including quotas and tariffs) for LDCs; and abolish export subsidies and similar measures. Of those measures, as argued by Gersfelt and H.G. Jensen in Chapter 3, market access is more important for developing countries than decoupling subsidies, although the latter may facilitate the former (see also Chapters 5 and 7). Concerns about the fate of small farmers in rich countries can be better addressed through decoupled income support and safety nets tailored to those farms rather than protecting and subsidizing specific productions that end up benefiting mainly large producers to the detriment of everybody else.

28

E. Diaz-Bonilla et al.

Developing countries also need to carefully consider their own domestic policies in general, as well as their agricultural policies in particular. For years many of them have maintained macroeconomic and trade policies that discriminated against agriculture, and currently, although general policy biases may be gone, a large percentage still do not invest enough in agriculture and rural development. At the same time, however, several developing countries indicated concerns during the current WTO agricultural negotiations that further trade liberalization could create problems for their large and predominantly poor agricultural populations. Poor countries have argued for a slower pace in reducing their tariffs on the premise that industrialized countries should first eliminate their higher levels of protection and subsidization. A related concern is how to manage sudden negative impacts from subsidized exports, or from import surges. Poor producers may see their livelihoods irreparably damaged by unfair trade competition and drastic shocks if, for instance, they are forced to sell productive assets or take children from school (see, for instance, Lipton and Ravallion, 1993). The requests for longer transition periods and the design of some policy instruments in the WTO framework that are better tailored to poor countries’ capabilities to manage unfair trade practices and shocks seem appropriate. Food-insecure countries may need streamlined procedures for taking action against export and other unfair trade practices, and to manage import surges or sudden price declines (such as a food security safeguard for a reduced number of products and during a limited period). While those are temporary instruments, there have also been arguments for even further and more permanent agricultural protection in developing countries to ease poverty and promote food security. As indicated, sometimes this suggestion is accompanied by the argument that protection ‘does not cost money’ and is easier to implement in poor countries than alternative policies such as investments in agricultural technology, extension, and infrastructure. As the simulations in Chapter 10 show, these notions are mistaken. Contrary to common perceptions that protection is a tax paid by foreigners and collected by governments, much of the (implicit) tax is paid by domestic consumers and collected privately by producers in the form of higher prices. This tax on food has an obvious negative impact on poor households, which in many developing countries spend more than half of their income in feeding themselves, and is mainly received by large agricultural producers, which have larger volumes of products to sell. It must be remembered that landless rural workers, the increasing number of poor urban households, and also many poor small farmers, are net buyers of food. Within developing countries there is also a variety of groups (different types of farmers, landless workers, poor urban consumers, women-headed farms and households, and other vulnerable groups) with different poverty profiles. The problems faced by those groups of poor farmers and poor consumers are better addressed through policies and investments targeted at them directly. Special and Differential Treatment under WTO rules granted at the national level, or at the level of staple crops

Overview

29

considered relevant for food security, is not necessarily the most effective and equitable way to address problems of poverty and hunger. Instead, poor countries need adequate policies that operate at the household and individual levels. More investment should be targeted to the poor and vulnerable, rather than to support crops in general, which usually benefits larger farmers. Also, as discussed in Chapter 10 as well, some of the suggested changes for a Development or Food Security Box that maintain or increase high levels of protection may reduce trade among developing countries. In these negotiations the heterogeneity among developing countries must be recognized, acknowledging that ‘one-size policy does not fit all’. In particular, if the issues of poverty and hunger are to be properly addressed, then quantitative and objective indicators should be utilized to identify the really vulnerable countries (Chapter 8). At a more systemic level, although world price instability during recent years does not seem higher than in the previous decades, it is in the interest of all countries to make sure that volatility does not increase in world markets. Particularly for poor and vulnerable countries, it is important that food aid is managed in a counter-cyclical way and that adequate financial facilities are operated by the international organizations (Chapter 9). The best approach for developing countries is to eliminate biases against that sector in the general policy framework and to maintain a neutral trade policy reducing protection overtime, while fully using longer transition periods to be negotiated in the WTO to increase investments in human capital, land tenure, water access, technology, infrastructure, nonagricultural rural enterprises, organizations of small farmers, better designed safety nets based on poverty indicators and other forms of social capital and political participation for the poor and vulnerable. Developing countries do not need controversial notions such as ‘multifunctionality’ to pursue those policies (Chapter 11), to the extent that they are not constrained under the existing WTO Agreement on Agriculture, and that current negotiations may give even more policy room to developing countries in the area of targeted policies to vulnerable groups. The claims for more protection out of concern for small farmers while under-investing in rural development and poverty alleviation would otherwise ring hollow. This targeted support requires additional financial resources from the international community and a pro-poor global environment, coupled with peace, good governance and the right political priorities in developing and developed countries. All countries must also follow prudent macroeconomic policies that minimize the risk of economic crises, which tend to be harsher on the poor. Overall, industrialized countries, because of the size of their economies and the impact of their economic, environmental, political, diplomatic and military decisions, bear a larger responsibility in creating a pro-poor world economy. A balanced result of the WTO negotiations can help the poor and hungry. But without addressing those other key factors, any modification in the WTO agreements may have limited benefits.

30

E. Diaz-Bonilla et al.

Notes 1

Agricultural negotiations were resumed in March 2000, as mandated in Article 20 of the Agreement on Agriculture, and reaffirmed in November 2001, at Doha, Qatar, when WTO member countries decided to launch a new round of trade negotiations on topics including agriculture, as part of a single undertaking. In Box 1.1 there is a summary of the Uruguay Round Agreement. 2 A more technical analysis of negotiating groups based on cluster methodology is in Lind and Bjørnskov, Chapter 6, this volume. 3 The notion of multifunctionality argues that agriculture creates public goods for the economy, whose importance goes beyond the market value of food and fibre as such. These effects are considered inseparable from production and may require the maintenance of subsidies and protection for agriculture. As discussed in Diaz-Bonilla and Tin (Chapter 11), a similar notion of positive externalities, with a parallel debate about what were the adequate policies, was behind the drive of industrialization in developing countries. 4 The NAC shows how much of the final price received by farmers comes from the market, valued at world prices, and how much the farmer receives as transfers from consumers (in the form of higher prices due to trade protection) and from taxpayers (as payments from the government). A value of 1 indicates that all the income received by farmers comes from the market, and any number above 1 indicates transfers from consumers and/or taxpayers. It does not include the government’s expenditures on general services such as research and extension, pest control and the like. Another indicator of the OECD, which is quoted more often, is the Producer Support Estimate (PSE). In simplified terms, the PSE shows transfers (from consumers and taxpayers) as a percentage of the full price received by the farmer (i.e. transfers (TR) divided by the sum of world price (WP) plus transfers, or TR/(WP+TR), then shown in percentages), while the NAC is the ratio of the sum of world price and transfers divided by the world price ((WP+TR)/WP). 5 The NAC, whether it comes from consumers or taxpayers, also varies by products in the various countries (OECD, 2003). Unfortunately, similar calculations for other commodities of importance to developing countries, such as fruit and vegetables and cotton, are not available. 6 Transfers from consumers are caused by tariffs, quotas and similar instruments that, within the WTO framework, are discussed under market access issues. Transfers from taxpayers are in the form of farm payments more or less linked to production, which are discussed under the rubric of domestic support or subsidies. Export subsidies have usually aspects of transfers both from consumers (by maintaining domestic prices higher than would have otherwise been the case) and from taxpayers (due to the payment involved). 7 In the case of the EU, the 2003 reform of the EU CAP will change this pattern somewhat (see Chapter 2, by Frandsen and Walter-Jørgensen). 8 The ‘Development Box’ (presented by Sri Lanka, Dominican Republic, Pakistan, Cuba and others) combines a series of existing exceptions and additional proposals for ‘special and differential treatment’ for developing countries in the areas of market access, domestic support and export subsidies (see Diaz-Bonilla et al., Chapter 10, this volume). 9 The Green Box (in fact, Annex 2 of the Agreement on Agriculture) enumerates a series of expenditures related to general services for agriculture as well as direct payments considered to be less distorting. Article 6.2 of the AoA allows the use of some domestic support for low-income and resource-poor producers.

Overview 10

31

The members of the Cairns Group are Argentina, Australia, Bolivia, Brazil, Canada, Chile, Colombia, Costa Rica, Guatemala, Indonesia, Malaysia, New Zealand, Paraguay, the Philippines, South Africa, Thailand and Uruguay. Although it is usually considered that Cairns Group members are large and competitive exporters, in fact, some of the developing countries in this group (the Philippines, Bolivia, Guatemala) have a profile of food-insecure countries (see Diaz-Bonilla et al., Chapter 8). 11 It should be noted that there are no references in the AoA to boxes or colours. The notion of Amber, Blue or Green Boxes is just a concise way of referring to specific categories of policies. A simple way to think about the domestic subsidies in the AoA is to divide them into: (i) those prohibited (Amber) that are paid to specific products in relation to current production; (ii) those permitted under restrictions (Blue) that are paid to specific products in relation to historic (not current) production (and therefore are delinked from current volume of supply) but are effective only if farmers continue producing those products; and (iii) those permitted (Green) that are not based on the farmers effectively producing specific products. While it is clear that the first two categories induce additional production (albeit Amber more than Blue), there is a debate regarding whether some payments considered under the Green Box are really non-distorting. 12 Diaz-Bonilla et al. in Chapter 8 discuss the heterogeneity of food security conditions across developing countries using cluster analysis on a world sample of developed and developing countries. See also McCalla and Valdes (1999). 13 Minot and Daniels (2002) report that US payments to cotton producers in 2001–2002 amounted to more than 50 per cent of world prices, maintaining high production, depressed world prices and increasing poverty in some African countries. 14 The GTAP project, administered by the Department of Agricultural Economics, Purdue University, West Lafayette, Indiana, is the result of collaboration between a consortium of international and national agencies from around the world (including the World Bank, the US Department of Agriculture, ABARE and others) to develop, maintain and update a global database of production, consumption and trade by countries and regions. IFPRI and FOI, also members of the consortium, have collaborated with data for specific countries for the world database. That database is permanently being updated and refined, particularly in relation to developing countries where data are weaker. IFPRI, FOI and other members of the consortium have devoted much effort in improving available data for developing countries. The simulations are mostly based on the latest version of the database available at the time of writing (GTAP 5). 15 Elasticities in FOI models also come from the GTAP model, while IFPRI has been sharing data mostly with the Economic Research Service of the US Department of Agriculture. 16 Other proposals, such as the EU–US joint framework, the G-20 counter offer or the Derbez Text, mostly lack specific numbers for the commitments, and therefore cannot be evaluated quantitatively. 17 Concerns about the wide variety of countries that are self-identified as developing countries and receive some special treatment because of that have existed for some time in the General Agreement on Tariffs and Trade (GATT), and now in the WTO. Those concerns are borne out by this analysis, at least for the limited issue of food security. 18 Some NFIDC, such as Venezuela, Trinidad and Tobago, are important oil exporters. Others may be net food importers just because of a dominant tourist industry (like Barbados, which has an income of US$7,000 per capita, the highest of all NFIDCs). However, although being a NFIDC may not be a good indicator of serious food security problems, this classification, negotiated during the Uruguay Round, constitutes an acquired right that has implications under the Ministerial Decision on

32

E. Diaz-Bonilla et al. LDCs and NFIDCs (‘Decision on Measures Concerning the Possible Negative Effects of the Reform Program on Least-developed and Net Food-importing Developing Countries’, GATT, 1994). 19 The creation of a new special safeguard in agriculture has been proposed by India (G/AG/NG/W/102 January 2001, India) and the African Group (G/AG/NG/W/142, March 2001) among others, while the streamlining of countervailing measures against subsidized exports, but also of domestic subsidies, has been suggested by several developing country members of the Cairns Group (‘Special and Differential Treatment for Developing Countries: Transitional Instruments to Expeditiously Countervail Subsidized Imports (SDCM)’, Special Session of the Committee on Agriculture, Informal Meeting, 24–26 September 2001, by Argentina, Bolivia, Paraguay, the Philippines and Thailand). 20 For instance, Delgado et al. (1998) have shown that an agricultural-led growth strategy may have larger dynamic multipliers for the rest of the economy than other alternatives in poor developing countries. For the links between agricultural growth and poverty reduction see Eastwood and Lipton (2001).

References Bjørnskov, C. and Lind, K.M. (2002) Where do developing countries go after Doha? An analysis of WTO positions and potential alliances. Journal of World Trade 36(3), 543–562. Delgado, C., Hopkins, J. and Kelly, V.A. with Hazell, P., McKenna, A.A., Gruhn, P., Hojjati, B., Sil, J. and Courbois, C. (1998) Agricultural Growth Linkages in SubSaharan Africa. Research Report 107. International Food Policy Research Institute, Washington DC. Diao, X., Diaz-Bonilla, E. and Robinson, S. (2003) How Much Does It Hurt? The Impact of Agricultural Trade Policies on Developing Countries. IFPRI, August 2003. http://www.ifpri.org/media/trade20030826. htm Diaz-Bonilla, E. and Reca, L. (2000) Trade and agroindustrialization in developing countries: trends and policy impacts. Agricultural Economics 23(3), 219–229. Diaz-Bonilla, E., Thomas, M., Robinson, S. and Yanoma, Y. (2002) WTO, Agriculture, and Developing Countries: A Survey of Issues. Trade and Macroeconomics Division. Discussion Paper No. 81, January 2002. International Food Policy Research Institute, Washington DC. Diaz-Bonilla, E., Robinson, S. and Swinnen, J.F.M. (2003) Regional agreements and the

World Trade Organization negotiations. American Journal of Agricultural Economics 85(3), 679–683. Eastwood, R. and Lipton, M. (2001) Pro-poor Growth and Pro-growth Poverty Reduction: What Do they Mean? What Does the Evidence Mean? What Can Policymakers Do? Paper presented at the Asia and Pacific Forum on Poverty: Reforming Policies and Institutions for Poverty Reduction. Asian Development Bank, Manila, the Phillippines, 5–9 February 2001. FAO (1983) Approaches to world food security. FAO economic and social development paper 32, Food and Agriculture Organization of the United Nations, Rome. Koester, U. and Bale, M. (1984) The Common Agricultural Policy of the European Community: a blessing or a curse for Developing Countries? World Bank Staff Working Papers No. 630, Washington DC. Lewis, W.A. (1954) Economic Development with Unlimited Supplies of Labour. Manchester School of Economic and Social Studies, Vol. 22, May 1954. Lipton, M. and Paarlberg, R. (1990) The Role of the World Bank in Agricultural Development in the 1990s. Washington DC: International Food Policy Research Institute.

Overview Lipton, M. and Ravallion, M. (1993) Poverty and Policy. In: Behrman, J. and Srinivasan, T.N. (eds) Handbook of Development Economics (3B). North-Holland, Amsterdam. McCalla, A.F. and Valdes, A. (1999) Issues, interests and options of developing countries. Paper presented at the Conference on Agriculture and the New Trade Agenda in the WTO 2000 negotiations. Geneva, Switzerland. Minot, N. and Daniels, L. (2002) Impact of Global Cotton Markets on Rural Poverty in Benin. MTI Discussion Paper No. 48, IFPRI, November 2002. OECD (2003) Agricultural Policies in OECD Countries: Monitoring and Evaluation. Paris. OECD (2004) Agricultural Policies in OECD Countries: At a Glance. Paris. Panagariya, A. (2004) The Miracles of Globalization. Foreign Affairs, September/ October 2004 Solagral (1999) Paper presented at the WTO Round and Food Security for USAID Partner Countries: an economic growth and agricultural development training

33 workshop. 1–2 November 1999. Washington DC. UNCTAD (United Nations Conference on Trade and Development) (2000) Impact of the reform process in agriculture on LDCs and Net Food-Importing Developing Countries and ways to address their concerns in multilateral trade negotiations. Background note by the UNCTAD secretariat. TD/B/COM.1/EM.11/2, Geneva, Switzerland. WTO (2000a) Legal tests: the WTO agreements – A Summary of the Final Act of the Uruguay Round (http://www.wto.org/english/docs_e/legal_e/ursum_wp.htm). WTO (2001) Agricultural Trade Performance By Developing Countries, 1990–99. Background Paper by the Secretariat. G/AG/ NG/S/6/Rev.131, January 2001. Geneva, Switzerland. World Bank (2001) Global Economic Prospects and the Developing Countries, 2002. Washington DC. World Bank (2002) World Development Indicators. Washington DC.

2

Review of the EU Common Agricultural Policy SØREN E. FRANDSEN AND AAGE WALTER-JØRGENSEN Danish Research Institute of Food Economics, Rolighedsvej 25, 1958 Frederiksberg C, Denmark

The purpose of this chapter is to present an overview of the EU Common Agricultural Policy (CAP) and to show how the policy has evolved over the past decade in response to external and internal pressures for reform. The international trade negotiations in the World Trade Organization (WTO), aiming at liberalizing world trade, exert pressure on the EU to reduce border protection and production-linked support for agriculture. Internally, there is pressure for reforms to emphasize rural economic development, protection of the environment and consumer preferences for food quality and animal welfare. Also, the enlargement of the European Union with ten new Member States has been instrumental in the reform process of the CAP. The chapter is organized as follows: first, the objectives of the CAP are outlined, indicating changes in objectives over time; second, the interaction between reforms of the CAP and the trade negotiations in the WTO are analysed using the McSharry Reform, the 2000 Agenda and the 2003 CAP reform as examples; third, an overview of the CAP is presented, giving emphasis to the main principles of market policy and rural development, including a brief review of the EU Commission’s proposal for rural development programmes for the years 2007–2013; fourth, the policy regimes for individual EU markets are analysed, giving details of the functioning of the CAP; and, finally, the perspectives for development of the CAP are discussed.

Objectives of the CAP The CAP has evolved over time to cope with changes in the economic and political environment and the challenges arising from enlargement of the Community.1 At the beginning, the CAP focused solely on market 34

© CAB International 2006. WTO Negotiations and Agricultural Trade Liberalization: the Effects of Developed Countries’ Policies on Developing Countries (eds E. Diaz-Bonilla et al.)

Review of the EU CAP

35

regulation aimed at enhancing productivity and stabilizing the supply of food, but as the EU became self sufficient in more and more farm products the emphasis was gradually shifted toward supply management and rural development. For many years, however, the main focus of the CAP has been to provide reliable food supplies and stable farm incomes. This line of policy prevailed until the beginning of the 1990s, when a major reform of the CAP was implemented (the McSharry reform), leading towards an integrated, coordinated and more focused effort to improve the conditions for rural life in the EU. The principles of the McSharry reform have been carried forward in two successive reforms: Agenda 2000, which has defined the policy framework for 2000–2006, and the 2003 CAP reform that spells out the long-term requirements for sustainable agriculture, taking into account the enlargement of the EU. The main objectives of the CAP today are to (European Commission, 1999a): ● ● ●

Ensure the competitiveness of the European Union agricultural sector, both within the Community market and in growing export markets. Promote ways of farming that contribute to the maintenance and enhancement of the rural environment and landscape. Contribute to sustaining the livelihood of farmers while promoting the economic development of the wider rural economy.

While pursuing these objectives, the CAP should also rise to the challenge of heightened consumer interests in food safety and quality, animal welfare and the protection of the environment.

Reforms of the CAP – Interaction with the WTO For over a quarter of a century, from the establishment of the CAP in 1962 to the beginning of the 1990s, the line of policy pursued by the EU with respect to agriculture has changed very little. The CAP provided protection for EU agriculture against world market fluctuations and support of farm incomes through high product prices while, at the same time, trying to cope with increasing production surpluses through market intervention and production management (quotas). As a result, the EU was facing soaring budget problems and criticism from consumers, who valued food quality and other related concerns such as food security, animal welfare and the environment, rather than quantity. The situation changed significantly when agriculture was brought into the international trade negotiations of the GATT in the late 1980s, forcing members of the WTO to review their farm policies. In response to the external and internal pressures for reform, the EU embarked on a major revision of the CAP in the early 1990s, using proposals from the GATT negotiations as a guideline. The revision, known as the McSharry reform, was implemented in 1993, two years before the Uruguay Round Agreement of Agriculture (URAA) became effective.

36

S.E. Frandsen and A. Walter-Jørgensen

This line of policy was carried forward in the Agenda 2000 reform, aiming at a coherent policy framework for agriculture and rural development in the EU. In response to the Doha Round of negotiations in the WTO, and influenced in particular by the prospective enlargement of the EU, the European Council agreed on a further reform of the CAP in 2003. The latter reform – also known as the mid-term review (MTR) – entails important changes for the CAP and its incentive structure from 2004 onwards.

The McSharry reform The McSharry reform contained two main elements: reduction of price guarantee levels and introduction of direct payments to producers as compensation for the lower prices. In the case of cereals, oilseeds and protein crops, producers were granted acreage support, and a similar system based on a direct payment per animal (headage premiums) was applied to the production of feeding cattle, sheep, lambs and goats. Introduction of support for compulsory land setaside was also part of the reform. The introduction of acreage and headage support was a compromise between the USA and the EU in the Uruguay Round negotiations, where the two parties agreed to reduce price support and to compensate farmers though direct income support. In the USA this involved replacement of the so-called deficiency payments by acreage payments based on historical land use, whereas the EU payments were linked to the actual use of land or to the current number of cattle. The US payment scheme is therefore considered to have less of an impact on production than the EU scheme2. Nevertheless, both schemes were placed in the URAA’s ‘Blue Box’, for which there was no commitment to reduce support.

Agenda 2000 In 1999, with a view to longer-term developments and the continuation of trade negotiations in WTO, EU Heads of State agreed on a new agricultural strategy comprising a coherent policy framework for agriculture and regional development in the EU. Covering the period 2000–2006, the strategy built on the principles of the McSharry reform but included also the development of a comprehensive strategy towards the wider needs of Europe’s rural communities. The main guidelines for the reform were (European Commission, 2001a): ●

●

Continued competitiveness should be ensured by sufficiently large price cuts that would guarantee growth of home-market outlets and increased participation by Community agriculture in the world market. A new division of functions between the Commission and the Member States concerning compensation in the form of direct payments or rural development measures.

Review of the EU CAP ●

●

37

Simplification of rules such as the new rural development regulation and the market-management regulations, in particular with respect to arable crops. Rural development becomes the second pillar of the CAP.3

The reform stressed the intention to consider rural development in a wider context, i.e. by including agriculture and forestry, as well as other occupational interests in rural areas. The 2003 CAP reform With a view towards making the CAP more market oriented, to provide better quality and healthy foods, to further environmental and animalfriendly production methods, and to sustain the natural living conditions and care for the countryside, yet another reform of the CAP was adopted in 2003. Like the McSharry reform a decade earlier, this reform is responding to external pressures from the world trade negotiations. At the same time, the reform reflects the needs for adjustment of the CAP in the light of the EU Council’s decision in December 2002 to go ahead with the Eastern Enlargement of the European Union. In particular, the 2003 reform seeks to better accommodate the concerns of developing countries and to bring CAP expenditures in accordance with the overall budgetary framework for the enlarged European Union until 2013. Thus, the reform is clearly in line with the objectives of Agenda 2000 and is meant to complete the reform process in some areas and establish a stable policy framework in others. The reform contains five main elements, namely (European Commission, 2003): ● ● ●

●

●

Continuation of the Agenda 2000 approach (revisions of the market policy, e.g. reductions in intervention prices for dairy products). Decoupling of direct support (introduction of a single farm income payment based on historical payments). Introduction of compulsory cross-compliance (reduction of direct payments in cases of non-respect of EU standards related to the environment, food safety, animal health and welfare). Strengthening rural development by enhancing rural development instruments to meet new standards and by redistributing funds from the 1st to 2nd pillar of the CAP (modulation). A mechanism for fiscal discipline to ensure that the CAP expenditures do not exceed the budgetary limit.

Continuation of the Agenda 2000 approach For cereals, the intervention price is maintained and the monthly increments are reduced by 50 per cent. The area premiums (to cereals, oilseeds, proteins and setaside) remain unchanged at €63/t. For protein crops the existing supplement of €9.5 will be maintained and converted into a crop-specific area payment of €55.57/ha. For durum wheat, the specific

38

S.E. Frandsen and A. Walter-Jørgensen

aid is reduced by 17 per cent from 2004 to 2006 in the traditional zones and it is decoupled in 2005. For rice, the intervention price is cut by 50 per cent and for nuts, starch potatoes and dried fodder, the market policies will be reformed as well. Reforms for olive oil, tobacco and cotton will, according to the decision, also be undertaken in the future. For dairy, the intervention price for butter will be reduced by 25 per cent, which is an additional price cut of 10 per cent compared to the Agenda 2000 reform. For skimmed milk powder, prices will be reduced by 15 per cent as agreed in Agenda 2000. It was decided to maintain the milk quotas until 2014/15, with an increase in the quotas for Greece and Portugal in addition to those agreed in Agenda 2000 (+2.39 per cent). To compensate for the price reductions, a direct payment per ton of milk is to be introduced from 2004 onwards. The single farm payment (decoupling) will apply only in the dairy sector once the reform is fully implemented (2007), unless Member States decide to introduce it earlier. Decoupling of direct support A significant share of the direct support is to be decoupled from production decisions. The single decoupled farm income payment from 2005 onwards includes prior payments for cereals, oilseeds, protein crops, flax, hemp, linseed, durum wheat supplement, starch potatoes (40 per cent), grain legumes, rice, dried fodder, beef and veal, sheep and goats, and milk (which will be initialized in year 2006/07). Each farmer, having received a direct payment under the existing support schemes, will be eligible to receive a single farm payment in the form of a number of payment entitlements, defined as the number of hectares on which the single farm payment may be claimed in the future (see below). The support per unit of entitlement is determined by: ● ●

A historic reference amount (corresponding to the amount of direct payments received during the reference period 2000, 2001 and 2002). A total reference area (the number of hectares generating the reference amount (including all forage area) in the reference years).

The entitlement (€/ha) is calculated by dividing the total reference amount by the total reference area. In a given year, payments will be granted only for those entitlements for which a farmer has eligible hectares at his disposal. Thus, in order to get the full payment, he must have as many eligible hectares as he has entitlements. The eligible area includes any type of agricultural land, with the exception of areas under permanent crops. The eligible area can be used for any agricultural activity defined as the production, rearing or growing of agricultural products as long as that land is maintained in good agricultural condition. Farmers may sell or lease land with or without entitlements and they may also sell their entitlements with or without land. But in all cases, farmers need to have an area of eligible hectares equal to or greater than the number of payment entitlements they hold in order to fully activate their

Review of the EU CAP

39

single farm payment. A national reserve for hardship cases, including problems of transition and new-entrance farmers, is to be established. In addition, Member States have the following implementation options at the national or regional levels (this is the partial implementation of decoupling): ● ● ●

●

Retention of 25 per cent of the hectare premium or, alternatively, up to 40 per cent of the supplementary durum wheat aid. Continuation of up to 50 per cent of the current sheep and goat premiums. Retention of up to 100 per cent of the suckler cow premium component and up to 40 per cent of the slaughter premium component or, instead, retention of either up to 100 per cent of the slaughter premium or up to 75 per cent of the special male premium component. Disbursement of additional payments for purposes of encouraging specific types of farming which are important in the protection or enhancement of the environment and in improving the quality and marketing of agricultural products.

Compulsory cross-compliance With reference to the concept of cross-compliance, the direct payment can be reduced in the case of non-respect of EU standards related to the environment, food safety, animal health and welfare, and to maintaining land in good agricultural and environmental condition. Strengthened rural development policy The term modulation refers to the way in which total direct aid payments for individual farms will be reduced. A franchise of €5000 is applied. The modulation will start with a rate of 3 per cent in 2005, 4 per cent in 2006 and 5 per cent in 2007 and beyond. The saved amount will be used for rural development purposes and the money will be spent according to size of the agricultural area, agricultural employment and income across member states. Member States will nevertheless receive at least 80 per cent of their contribution to the modulation scheme. Fiscal discipline Finally, fiscal discipline will be invoked from 2007, so that direct payments will be reduced if the budgetary ceiling is overshot by a security margin of €300 million.

Overview of the CAP In order to give the reader an impression of the functioning of the CAP and the problems involved in alleviating the access to the market for developing countries’ exports of farm products, we have to look at the actual formulation

40

S.E. Frandsen and A. Walter-Jørgensen

of the CAP. In view of the overall objective of the publication, the main focus here is on market policy, giving emphasis to the effect of the policy on market access in the EU. However, for the sake of completion, the rural policy of the EU is also reviewed, presenting the EU Commission’s recent proposal for a revision of the rural policy to be implemented during the period 2007–2013.

Market policy The Market Policy of the CAP rests on three main principles, which have guided intervention in the EU markets for agricultural products since the establishment of the CAP in 1962. These are: A Common Market It is a major objective of the EU that labour, capital, goods and services should be allowed to move freely between member states in order to promote equal conditions for producers and fair competition in the market. This is pursued through the EU Customs Union that provides free trade among member states and common rules for external border protection. Internal border control remained in existence right up to 1986, when the member states agreed upon common rules for the internal market but, in principle, agricultural trade between member states was liberalized with the establishment of the CAP. Internal trade is, nevertheless, subject to sanitary and phytosanitary regulations, which will not be dealt with in this chapter. Preferential treatment The EU gives priority to EU-produced agricultural products over imported products. This is pursued through a combination of external border protection, intervention purchases in the internal market and export restitutions. Until the implementation of the GATT agreement in 1995, EU markets for major agricultural products were protected by variable import levies that raised the price of imported products to a predetermined level (reference or target price). Today, import of agricultural products is subject to custom tariffs as determined by the Uruguay Round Agreement on Agriculture, including the use of Tariff Rate Quotas in a number of cases. Figure 2.1 gives an indication of the development in market price support and total support in EU agriculture. In 2003, market price support corresponded to 20 per cent of the sector’s revenue from production and support. If we include public expenditures on other support4 to agriculture, the share of support in total revenue amounted to 35 per cent that year. Market price support has accounted for a declining share of total support but, as indicated by the Figure, the overall level of support was about the same in 2003 as in 1995 when the URAA was implemented. The fall in support in 1996 and 1997 is explained mainly by high world market prices of cereals that led to the introduction of export taxes in the EU.

Total support

2003

2002

2001

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

41

1989

1988

1987

50.0 40.0 30.0 20.0 10.0 0.0

1986

Support in % of total revenue

Review of the EU CAP

Market price support

Source: OECD (2004) and own calculations. Note: Total revenue = value of production (at farm gate) + total support estimate (producer support estimate, PSE). Market price support is measured as the difference between the internal price and the world market price. Other support elements are measured by budgetary expenditure. As market price support is included in the value of production, the percentages are calculated as follows: Market price support % = (MPS/(VP
MPS + TS))*100; Total support % = (TS/(VP
MPS + TS))*100 where MPS = market price support; VP = value of production (at farm gate) and TS = Producer support estimate (PSE). Fig. 2.1. Level of support in EU agriculture.

Table 2.1 provides an overview of the average size of import barriers in the EU for various agricultural products measured as import tariff equivalents in percentage of the world market price. The figures are based on applied tariff duties, taking into account exhaustively regional agreements and trade preferences for the 12 aggregate commodities shown. The figures indicate that the markets for dairy products, ruminant meat and Table 2.1. Average import protection in the EU. (From GTAP Database, Version 6, 2004.) Import tariff equivalents, % of world market price Primary agricultural products Wheat Other grains Vegetables, fruit, nuts Oilseeds Plant-based fibres Wool Processed agricultural products Beef and veal, sheep meat Other meat Vegetable oils and fats Dairy products Processed rice Sugar

0.9 18.0 16.6 0.0 0.0 0.0 46.3 18.8 11.9 38.1 108.4 110.1

42

S.E. Frandsen and A. Walter-Jørgensen

sugar are highly protected, whereas the EU allows free imports of oilseeds, plant-based fibres and wool. The average rates of protection for other meat (pigs and poultry) and for fruit and vegetables were in the range of 20 per cent and 12 per cent, respectively, measured relative to world market prices. Common financing The third pillar of the CAP is the principle of common financing that implies sharing the cost of the policy by Member States. The main sources of revenue are contributions from Member States and the so-called Traditional Own Resources (TOR), consisting of customs duties, agricultural levies and a Value Added Tax (VAT)-based levy system. In 1988, a new Own Resource based on Member States’ GNP was introduced, setting a ceiling for total payments at 1.2 per cent of GNP by 1992 and reducing the basis for the VAT levy (European Commission, 1999b). The funds are administered through the European Agricultural Guidance and Guarantee Fund (EAGGF). In particular, the Fund’s Guarantee Section finances expenditures related to the agricultural market organizations, rural development measures that accompany market support, rural measures outside regions whose development is falling behind, certain veterinary expenditures, and, finally, information measures relating to the CAP. The Guidance Section finances other rural development expenditures not covered by the Guarantee Section. Out of a total budget for the EU of €101,051 m. in 2004, €48,361 m. was accounted for by expenditure on agriculture (Table 2.2), of which more than 80 per cent related to market policies and 17 per cent to rural development. Rural development policy – multifunctional agriculture The rural development policy that resulted from the Agenda 2000 strategy aimed to complement reforms in the agricultural market sector by promoting a competitive, multifunctional agriculture and encouraging Table 2.2. Budgetary expenditures on the CAP, 2004. (From European Commission, 2004.) Million euro EAGGF – Guarantee Plant production Animal production Ancillary expenditures Rural development Monetary reserves EAGGF – Guidance Other agricultural expenditures Total

26,789.4 12,264.5 903.5 4,803.0 0.0

44,760.4

3,564.7 35.4 48,360.5

Review of the EU CAP

43

alternative sources of income in rural areas, while giving proper consideration to environmental conditions. Thus, it has been the objective to introduce a sustainable and integrated rural development policy, to ensure better coherence between rural development and the prices and market policies of the CAP, and to promote all aspects of rural development by encouraging the participation of local actors. To achieve these objectives, a number of measures were put at the disposal of the Member States, who were allowed to choose measures that responded best to the needs in their rural areas. The instruments applied can briefly be grouped as (EU Commission, 2004): ● ● ● ● ● ● ●

Investments in farm business. Human resources: young farmers, early retirement, training. Support for less favoured areas and areas subject to environmental constraints. Agri-environmental measures. Support for processing and marketing of agricultural products. Afforestation of agricultural land. Measures promoting the adaptation and development of rural areas.

EU support for rural development is co-financed by the EAGGF and Member States. Agri-environment measures, aid for early retirement, afforestation of agricultural land and support for less favoured areas and areas subject to environmental constraints (measures accompanying market policies) are co-financed by the EAGGF Guarantee section. Local activities (LEADER+ projects), designed to help rural actors improve the long-term potential of their local region, are funded from the EAGGF Guidance section. For other rural development measures, the source of EU funding depends on the region concerned: support to regions classified as less favoured areas (Objective 1 areas) is financed by the EAGGF Guarantee section, whereas support to regions outside Objective 1 areas is covered by the EAGGF Guidance section. As described above, the 2003 CAP reform further strengthened rural development activities by transferring funds from market and income support to rural development. At the same time, the scope of instruments was expanded in order to respond to growing public concerns on food quality, environmental protection and animal welfare. For this purpose, a number of measures were added to the list of measures already in place, including: ● ● ● ● ● ● ● ●

Food quality measures (participation in quality schemes). Standards related to the environment, health (public, animal and plant), animal welfare and occupational safety. Animal welfare (beyond good animal husbandry practice). Support for young farmers (reinforced). Support for the implementation of Natura 2000 (Birds and Habitats Directive). Support for forestry (widened scope). Increased EU co-financing rate for agri-environment and animal welfare. Less favoured areas (increased maximum level of support).

44

S.E. Frandsen and A. Walter-Jørgensen

As described earlier, the 2003 CAP reform introduced a new system of compulsory modulation, designed to switch funds from market policy to rural development. Member States are free to use such funds to finance the introduction of the new rural development measures or to reinforce existing measures. The recurrent review has clearly made the EU’s rural development policy complicated and difficult to administrate. With a view to reinforcing the development activity and simplification of its implementation, the EU Commission adopted a proposal for a reform of the policy in July 2004. The purpose of the reform is to bring the policy under a single funding and programming instrument in order to increase its coherence, transparency and visibility. The proposed reform pivots around three major policy objectives (see Table 2.3): improving competitiveness of farming and forestry (Axis 1); enhancing the environment and countryside through support for land management (Axis 2); and strengthening the quality of life in rural areas (Axis 3). Furthermore, a fourth implementation axis (LEADER) mainstreams the local development strategies developed through a bottom-up approach, which were previously financed under the LEADER initiative. To balance these activities, the EU Commission suggests that a minimum of 15 per cent of the total EU budget for rural development should be allocated to axes 1 and 3, whereas axis 2 should get at least 25 per cent of the total budget. Seven per cent of the budget should be allocated to the LEADER axis. European Union co-financing should not exceed 50 per cent for axis 1 and 3 activities (75 per cent for convergence regions) or 55 per cent for axis 2 and LEADER activities (85 per cent for convergence regions). Special rules apply to ultra-peripheral regions. The preliminary EU budget for rural development is set at €11,724 m. in 2007, increasing to €13,165 m. in 2013 (2004 prices).

Policy Regimes – Individual EU Markets The EU market policy is based on a system of border protection and market intervention, as indicated by Table 2.4. For some products, farmers are guaranteed a minimum price, which is guided by politically determined intervention prices. For other products, the price may be determined by market forces or a combination of market forces and elements of support. Production aid is given for olive oil, seeds, wine, tobacco and fruit and vegetables. Quantitative restrictions on production are used in a number of markets either in the form of restrictions on the use of land, as production quotas or as restrictions on the number of animals eligible for support. Public support for storage and exports may, moreover, be used to balance supply and demand. Producers of arable crops receive acreage payments as compensation for the reduction of prices following the McSharry reform, and a similar system applies to beef, sheep and lamb production.

Review of the EU CAP

45

Table 2.3. Axes and strategies in the proposal for rural development programmes. (From European Commission, 2004.) Objective

Strategies and rural development programmes

Axis 1

Human resources Vocational training and information Setting up young farmers Early retirement for farmers and farm workers Advisory services for agriculture and forestry Restructuring physical potentials Modernization of farms Improving the economic value of forests Adding value to primary agricultural and forestry production Improving infrastructures Natural disaster Restoring agricultural production potentials Improving product quality Helping farmers to adapt to product standards Participation in food quality schemes Information and promotion of products under quality schemes Transitional measures for the new Member States Supporting semi-subsistence farms Setting up of producer groups

Improving competitiveness of farming and forestry

Axis 2

Enhancing the environment and countryside through support for land management

Axis 3

Strengthening the quality of life in rural areas

LEADER

Sustainable use of agricultural land Natural handicap payments to farmers in mountain areas Payments to farmers in handicap areas other than mountain areas NATURA 2000 payments Agri-environmental and animal welfare Support for non-productive investments Sustainable use of forestry land First afforestation of agricultural land First establishment of agriforestry systems on agricultural land First afforestation of non agricultural land NATURA 2000 payments Forest-environment payments Restoring forestry production potential Support for non-productive investments Diversifying rural economies Diversification to non-agricultural activities Creation and development of micro-enterprises Encouragement of tourism Maximizing the potential of the natural heritage Improving the quality of life in rural areas Essential services for the economy of the rural population Renovation and development of villages; preservation and restoration of rural heritage Local development strategies Implementation of local development strategies Implementation of cooperative projects Education

46

Table 2.4. Instruments applied by the CAP market policy. (From European Commission, 2001b.) Minimum or guidance price X

X X

Storage support

Export support

Direct supportb

Xj Xj Xj Xk

X

X

Xl Xl Xl

X X

X X

Xj Xj

X

X

Xk

X X

X X X

Xf Xg

X Xh Xc Xd X Xe

Quantitative restrictions

Xi

X X

a Direct

h Premium

b Compensatory

i

support based on output. payments for reduction in product prices. c Minimum prices for certain processed fruits and vegetables. d Intervention prices for butter and skimmed-milk powder. e Basic prices for fresh and chilled sheep carcasses. f Based on actual production. g Flat rate production aid.

graduated by type of product. Community withdrawal compensation. j Restrictions on land use. k Restrictions on output. l Acreage support. m Headage support.

Xm Xm Xm S.E. Frandsen and A. Walter-Jørgensen

Cereals Oilseeds Protein crops White sugar Olive oil Seeds Wine Tobacco Fruit and vegetables Milk and dairy products Beef Sheep and lamb Pork Eggs

Production aida

Review of the EU CAP

47

To be eligible for acreage support, farmers are committed to set aside farmland. The amount of land to be set aside by producers is calculated as a proportion of the area under arable crops (reform area). The total reform area, and thereby the total budget for hectare premiums, is fixed on a Member State basis. Expansion of the planted area beyond the defined base area will lead to a proportionate reduction in the premium per hectare. This also applies to the premium for setaside. From the 2000–2001 marketing year up to the 2006–2007 marketing year, the reference rate of compulsory setaside is 10 per cent, but the percentage may be adjusted should the supply situation change.

Cereals The market regime for cereals covers wheat, rye, barley, oats, maize, grain sorghum, buckwheat, millet and canary seed. The regulation is based on border protection, public storage and export restitutions to ensure guaranteed prices to producers. As the EU is a net exporter of cereals, price guarantees to producers depend ultimately on the administration of export restitutions. Producers receive acreage payments as compensation for price reductions following the McSharry Reform. To be eligible for compensation farmers must set aside a certain percentage of their productive land, as described above. The setaside scheme has been an important policy instrument in restricting production. In recent years, total support for wheat has amounted to 46 per cent of the value of production (Table 2.5). Thirteen per cent of the support is provided as market price support, whereas 75 per cent is in the form of payments based on area (mainly compensatory payments for reduction of guaranteed prices since 1993). The payments also comprise a number of other support programmes, including disaster payments. Payments based on input use (e.g. support for the restructuring of agriculture, organic production, vocational training and support for setting up young farmers) amounted to 5 per cent of the total. Payments based on input constraints include agri-environmental and forestation measures. For maize, the level of support is slightly lower than for wheat, with a larger share of the support being provided through the market. As mentioned above, the 2003 CAP reform provides certain adjustments to the support mechanisms for cereals. The intervention price for cereals is retained, but the monthly increments are reduced by 50 per cent. Rye will be excluded from the intervention system, but Member States whose rye production is significant can receive additional modulation payments within the framework of rural development measures. The EU is a net exporter of cereals, with a self-sufficiency rate in the range of 105 per cent to 125 per cent in recent years. The EU accounted for 14 per cent of global exports of cereals and 3 per cent of global imports in 2001.

48

Table 2.5. Level and composition of support, average 1997–2001. (From OECD, 2003 and own calculations.) Composition of support, %

Market price support

46 39 38 52 48 80 58 19 31

13 39 0 91 90 59 30 78 89

75 49 81 3 1 28 61 2 1

Note: General Services Support Estimate (GSSE) is not included.

Payments based on historical entitlement

Payments based on input use

Payments based on input constraints

0 0 0 0 1 0 0 1 1

5 5 6 5 6 6 9 15 9

8 7 13 1 1 6 0 5 1

S.E. Frandsen and A. Walter-Jørgensen

Wheat Maize Oilseeds Sugar Milk Beef Sheep and goats Pigs Poultry

Percentage PSE

Payments based on area/number of animals

Review of the EU CAP

49

Oilseeds Oilseeds comprise seeds of rape, colza and sunflowers, and soybean. The EU market in oilseeds has been open to duty-free imports since the McSharry reform in 1993, but producers still receive acreage payments. The latter used to be considerably higher than for cereals, but following the Agenda 2000 reform acreage payments have been aligned progressively to the level applicable to cereals. Following the alignment, the limit on area grown with oilseeds has been raised as of the 2002–2003 marketing year. Total support amounts to about 38 per cent of the production value, 81 per cent of the support being mainly acreage payments in compensation for institutional price reductions since 1992 (Table 2.5). Thirteen per cent of the support is linked to input constraints, most of which is compensation for setaside. The production of rapeseeds has increased by nearly 10 per cent per annum during the 1990s, expanding exports by more than 20 per cent per year. The market policy for oilseeds is not affected by the 2003 CAP reform. Protein crops Protein crops (peas, field beans and sweet lupins for feed) are part of the arable crops sector eligible for acreage support. No trade barriers are applied to these crops but producers receive acreage payments and a yieldbased supplement. As a result of the 2003 CAP reform, the latter will be converted into a crop-specific area payment. Sugar The production of sugar in the EU is subject to considerable border protection, the price to consumers being set by a politically determined intervention price. Producers are guaranteed minimum prices for sugar produced within the so-called A- and B-quotas and subject to levies to cover the costs of subsidizing export of A- and B-sugar. Exports of C-sugar (production in excess of the A- and B-quotas) are sold at the world market price. The EU covers extra costs associated with the re-export of sugar imported on preferential terms. Support has amounted to 52 per cent of the production value in recent years, 91 per cent of which is market price support (Table 2.5). Sugar is not part of the arable crop scheme providing acreage support, but 5 per cent of the support is linked to input use, mainly investment support. Neither the Agenda 2000 reform nor the 2003 CAP reform brought changes to the policy regime for sugar, but as part of the Everything But Arms initiative taken in 2001 the EU has decided to open the market for duty-free imports of sugar from the world’s 49 Least Developed Countries. Also, in the autumn of 2003, the EU Commission wrote up a sugar policy option paper in which several alternatives for the future of the EU sugar regime were discussed. Among the options is the suggestion to abolish the sugar quota system, supplemented by a significant cut in the intervention price of white sugar.

50

S.E. Frandsen and A. Walter-Jørgensen

The EU is one of the largest producers of sugar in the world and one of the largest net exporters. EU sugar production has been relatively stable for a number of years, the rate of self-sufficiency being in the range of 113 to 115 per cent.5 Olive oil The market regime for olive oil is based on border protection, production aid and export restitutions. In 1998 direct payments to smallholders – based on the number of trees – were abolished and replaced by production-linked direct payments; consumption aid was abolished, and the Community maximum guaranteed quantity of olive oil eligible for production aid was apportioned among Member States. Since 2001 production aid has been granted only to oil obtained from olive groves planted before 1 May 1998. Reforms regarding the so-called Mediterranean products will nevertheless be finalized in the years to come. Expenditures related to such reforms will be kept within the existing budget and will be based on the objectives and the approach of the 2003 reform package. Olive oil is produced in the Mediterranean region by a large number of primarily small producers. Production has increased by more than 10 per cent during the 1990s, resulting in an increase in exports of 7–8 per cent and a similar fall in imports. The rate of self-sufficiency has increased steadily from 93 per cent in 1992/93 to 130–140 per cent in recent years. Seeds Producers of grass seeds receive flat rate production aid graduated by the type of product. The market for grass seeds is not subject to border protection and export restitutions are not provided for exports of seeds. Wine The EU market regime for wine is based on border protection, intervention mechanisms to support producer prices and measures to reduce winegrowing areas. The market protection has, in certain periods, resulted in considerable surpluses of wine, leading to the introduction of a distillation scheme for table wine. The market policy for wine was revised in 1999, resulting in the following changes: restrictions on the planting of certain wine varieties; introduction of support for permanent abandonment of wine-growing areas; and support for private storage of table wine. The market regime has not been affected by the 2003 CAP reform. The EU occupies a leading position on the world wine market, but total wine production has been declining during the 1990s, resulting in increased imports. During the same period, the self-sufficiency rate has varied between 133 and 108 per cent.

Review of the EU CAP

51

Tobacco The market regime for tobacco is based on a combination of import barriers, production quotas and a premium scheme, allowing for production-based payments. As in the case of olive oil, the regime for supporting the production of tobacco is also expected to be changed in the future, in line with the objectives and the approach taken in the 2003 CAP reform. The EU accounts for about 4 per cent of world tobacco production, nearly 30 per cent of all imports, and around 10 per cent of global exports. Greece and Italy are the major producing countries in the Community, accounting for about three-quarters of the total output of raw tobacco. Fruit and vegetables The EU is a net exporter of vegetables but a net importer of fruit. The market regime for fruit and vegetables builds on border protection, intervention purchases of surplus production and export restitutions. The EU also provides support for market organization and clearance of orchards. Intervention in the internal market is administered by producer organizations that can withhold products delivered by Members and receives support for alternative disposals. The intervention scheme was revised in 1996, and the amount of products qualifying for such support has been falling ever since. In 2002 intervention purchases were restricted to 5 per cent of market production of citrus fruits, 8.5 per cent of market production of apples and pears and 10 per cent of other products. The products that are withheld can be used for charity purposes or they may be destroyed. The market regime for fruit and vegetables is not affected by the 2003 CAP reform. Dairy products The market regime for dairy products combines quantitative restrictions (quotas) on milk production at the farm gate level with border protection and market intervention. The internal market for dairy products is heavily protected, with protection coefficients in the range of 85–90 per cent of the world market price in recent years. Intervention in the internal market is based on public storage of butter and milk powder combined with export subsidies. Total support for milk production has amounted to 48 per cent of production value in recent years, 90 per cent of which is market price support (Table 2.5). Six per cent of this support consists of payments based on input use such as support for organic production, and restructuring and modernization. As part of Agenda 2000, the intervention prices for butter and skimmed-milk powder were to be reduced by 15 per cent in three stages, starting from the 2005/06 marketing year. In compensation, producers would receive direct payments based on their individual quota. The amount of direct payments per premium unit would be increased in four stages from 2005 to 2007 and the total milk quota raised by 2 per cent in four stages from 2001.

52

S.E. Frandsen and A. Walter-Jørgensen

The 2003 CAP reform brings further changes to the milk sector, cf. above. As described earlier, the intervention price for butter will be reduced by 10 per cent in addition to what has been already foreseen as a result of Agenda 2000, and intervention purchases of butter will be suspended above a predetermined level. In compensation, direct payments will be increased by about twice the amount specified in Agenda 2000. The general quota increase decided in the Agenda 2000 reform is scheduled to take place from 2006 onwards. The EU is one of the largest exporters of dairy products, but also imports dairy products on a small scale, notably from developed countries. Production has remained constant since the introduction of milk quotas in 1984. Beef and veal The production of beef and veal in the EU is linked mainly to milk production. Nevertheless, the market for these products has its own policy regime that builds on principles similar to that of dairy products: protection of external borders; intervention in the internal market; and export restitutions in order to maintain minimum prices to producers. In addition, producers are paid headage premiums as compensation for product prices reductions. The production of beef and veal is subject to a variety of direct support programmes, some of which are linked to production (special premiums, suckler cow premiums, slaughter premiums and additional payments). Others are aimed at levelling seasonal variation in slaughtering or extensification of production. Some of the programmes have been in existence for quite a number of years, but payments have been enhanced in compensation for reductions in product prices. As part of the Agenda 2000 reform, the intervention price was reduced by 25 per cent from 2000 to 2002. In compensation, direct support to producers of bovine animals was increased by way of extending various support schemes (Walter-Jørgensen and Jensen, 2001, p. 72). The market regime for beef and veal is not affected by the 2003 CAP reform. According to the OECD, total support amounts to nearly 80 per cent of the value of production, of which 59 per cent is market price support (Table 2.5). Headage payments, mainly compensation for price reductions, have accounted for 28 per cent of total support, and payments based on inputs and input constraints for 6 per cent each. The production of beef and veal has more or less followed the trend in milk production with a rate of self-sufficiency ranging between 100 and 115 per cent. The EU is a net exporter of beef and veal, but also imports quality beef, primarily from South America, South Africa and the USA. Sheep, lamb and goats The market regime for sheep, lamb and goats is based on border protection and market intervention to protect the Community market against price fluctuations on the world market. Producers are granted a premium per

Review of the EU CAP

53

head of ewe based on a politically determined basis price for sheep meat in the internal market. A similar system applies to producers of goat meat in less favoured areas. Compared to beef, the level of market price support is lower for sheep, lamb and goats. Direct payments, on the other hand, are much higher for sheep than for beef. The market regime is not affected by the 2003 CAP reform. Total support amounts to about 58 per cent of the production value, 30 per cent of which is market price support. Direct payments (headage support) account for 61 per cent of the support and payments based on input use account for 6 per cent (Table 2.5). The EU is the second largest producer of meat from sheep and goats measured on a global scale. Yet, the EU is only 80 per cent self-sufficient in these products, the largest share of imports coming from New Zealand and Australia. Pork The EU market regime for pork is based on border protection and export restitutions. No price guarantees are provided for the production of pork. Export subsidies were formerly intended to compensate producers for the higher price of cereals on the internal market compared with the world market price, but this is no longer the case. Today, export restitutions are administrated on an ad hoc basis as relief/aid to farmers in periods of high supply and low market prices. The EU maintains considerable border protection for pork, with the overall rate of tariffs being in the range of 30 per cent of the world market price. Total support amounts to 19 per cent of the production value, 78 per cent of which arises from border protection (Table 2.5), with 15 per cent being provided in the form of input payments (mainly restructuring and modernization of production). The EU accounts for about 20 per cent of the global output of pork, about 6 per cent of the production being exported to foreign markets. The production of pork has increased steadily in recent years, resulting in a yearly increase in exports close to 4 per cent during the 1990s. Imports of pork, which used to be almost non-existent, have increased since 1993 due to the commitments imposed by the URAA. The rate of self-sufficiency has been stable over a number of years at about 107 per cent. Poultry and eggs The market regimes for poultry and eggs are identical. The EU does not provide price guarantees or direct support for these products, but the internal market is protected by import tariffs. The EU also provides export restitutions on a restricted basis. Total support amounts to 31 per cent of the production value (Table 2.5). Eighty-nine per cent is market support and 9 per cent is payments linked to input use (notably restructuring and modernization of production). The EU is self-sufficient in poultry, meat and eggs, but preferential imports have increased following the URRA.

54

S.E. Frandsen and A. Walter-Jørgensen

Discussion and Perspectives The EU is under continuing pressure, notably in the WTO, to reduce domestic support and to open its markets for imports of agricultural products from developing countries. So far, the pressure has had limited effects on the level of support, mainly because reforms of the CAP have involved shifts in support rather than reduction of support. An example is the introduction of area payments to compensate for reduction in price support in connection with the McSharry reform. As a result, production management in the EU has mainly been administered through quantitative restrictions on production (quotas) and factor use (setaside of farming land), while maintaining a high level of border protection and productionlinked support. By emphasizing rural community issues, the Agenda 2000 reform widened the focus of the CAP, putting greater emphasis on the protection of the natural environment and allowing non-farming aspects to become an integral part of rural policy in the EU. Likewise, Agenda 2000 opened up a new division of functions between the Commission and Member States concerning compensation in the form of direct payments or rural development measures. It is the intention of the reform that local governments should be responsible for such support, subject to Common Market regulations. Otherwise, Agenda 2000 has continued the line of policy laid out in the McSharry reform, compensating farmers for price reductions through direct payments coupled to production or land use. The 2003 CAP reform takes a different approach, which entails a radical reform of the CAP and its incentive structure from 2004 onwards. By making direct payments dependent on the amount of support received in a given reference period (2000–2002), the so-called ‘single payment scheme’ will provide support independent of that what farmers actually produce, forcing them to consider actual market conditions when planning production. Furthermore, direct payments are being linked to the fulfilment of certain environmental objectives. Yet, the payment is still tied to the total agricultural area and thereby to total agricultural production. In that respect the new single farm payment is coupled to the level of agricultural activity and production. The single payment scheme became effective on 1 January 2005 and will be fully implemented by 2007. The reform also leaves a back door open for Member States to maintain a proportion of direct aids to farmers in their existing form, notably where they believe there may be disturbance to agricultural markets or abandonment of production as a result of the move to the single payment scheme. The escape clause applies to a significant part of the direct payments for cereals and arable crops, sheep and goat schemes, and direct payments in the bovine sector. Furthermore, within specified limits, Member States are entitled to grant additional payments to support agricultural activities that are important for the protection or enhancement of the environment or for improving the quality and marketing of agricultural products.

Review of the EU CAP

55

The 2003 CAP reform places considerable emphasis on cross-compliance. To be eligible for direct payment, farmers must comply with European standards for the environment, food safety and animal health and welfare. The linking of direct payments to good farming practices and environmental conditions continues the line of policy laid down in Agenda 2000. However, it remains to be seen how cross-compliance can be handled, taking the need for control and investigation of the impact of farming practices on the environment, food safety and animal health and welfare into account. With a view to reinforcing the development activity and to simplify its implementation, in July 2004 the EU Commission adopted a proposal for a revision of the administration of rural development policies in the EU. The goal is to increase the coherence, transparency and visibility of the policy by bringing structural policies under a single funding and programming instrument, giving Member States a central role in drawing up their rural development programmes and implementing them. Neither the targets nor the choice of instruments will be affected by the reform. It goes beyond the present analysis to judge the efficiency of this reform. It appears, however, that a more precise targeting of the policy would be needed to achieve these objectives. From a developing country perspective, as well as in the context of the WTO trade negotiations, the 2003 Common Agricultural Policy Reform is a step towards lowering support for EU agriculture. However, the most important issue for developing countries appears to be market access, as discussed in Chapters 3, 5 and 7. The studies also reveal that trade liberalization in agriculture has the potential to create economic benefits, notably in developed countries, that could be used to assist growth and development in the poorer regions of the world. As pointed out by Yv and Jensen in Chapter 13, developing countries are facing many supply side constraints and bottlenecks, which require technical and financial assistance and cooperation to overcome. Thus, the 2003 EU reform would have to be supplemented by economic assistance to help the developing countries harvest the benefits of trade liberalization.

Notes 1

Up to May 2004, the EU covered 15 member states: Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, the Netherlands, Portugal, Spain, Sweden and the UK. In May 2004, ten new member states joined the EU: Cyprus, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Malta, Poland, Slovak Republic and Slovenia. 2 Acreage payments linked to land use will raise the land rent and thereby enhance the competitive position of the supported crop. 3 The 1st pillar is market policy. 4 General Service Support (GSSE) not included. 5 European Union agriculture ministers agreed on 24 November 2005 to reform the EU sugar policy. The EU-guaranteed price for white sugar will be cut by 36 per cent over a 4-year period beginning in 2006/07. Farmers will be compensated for, on average, 64.2 per cent of the price cut through a decoupled payment. Countries that

56

S.E. Frandsen and A. Walter-Jørgensen give up more than half of their production quota will be entitled to pay an additional coupled payment of 30 per cent of the income loss for a period of 5 years. ‘A’ and ‘B’ quotas will be merged into a single production quota and a 4-year voluntary restructuring scheme, consisting of a payment to encourage factory closure of EU sugar factories and isoglucose and inulin syrup producers, will be established. Intervention buying of surplus production will be phased out after 4 years.

References European Commission (1999a) CAP reform – a policy for the future (http://europa. eu.int/comm/dg06/publi/fact/policy/en.pdf). (Accessed November 2000.) European Commission (1999b) Financing the European Union – Commission report on the operation of the own resources system, Directorate General XIX, 7. October 1999. European Commission (2001a) Agenda 2000, (http://www.europa.eu.int/scadplus/leg/en/ s60000.htm). (Accessed February 2001.) European Commission (2001b) The agricultural situation in the community, 1999 Report, Directorate-General for Agriculture and Rural Development, Brussels/ Luxembourg. European Commission (2002) The agricultural situation in the community, 2002 Report, Directorate-General for Agriculture and Rural Development, Brussels/Luxembourg. European Commission (2003) CAP reform summary, Newsletter Special Edition, Directorate-General for Agriculture and Rural Development, July 2003. GTAP (2000) Database, version 5, pre-release 3 at www.gtap.agecon.purdue.edu. OECD (2000) AMAD – Agricultural Market Access Database, Paris. OECD (2002) Producer and consumer support estimates, OECD Database 1986–2001, Paris.

UNCTAD (2000) Handbook on the GSP scheme of the European Community (http://www.unctad.org/Gsp/eu/euhtml/poa rt1.asp) (Accessed December 2000). Walter-Jørgensen, A. and Jensen, T.V. (2001) The CAP and the international trade negotiations, Rapport nr. 123, Statens Jordbrugs- og Fiskeriøkonomiske Institut, Copenhagen, Denmark, 119 p. Walter-Jørgensen, A., Jensen, H.G. and Frandsen, S.E. (2001) Reform af EU’s sukkerpolitik – konsekvenser for EU og udviklingslandene, Rapport nr. 126, Statens Jordbrugs- og Fiskeriøkonomiske Institut, Copenhagen, Denmark, 89 s. WTO (2000a) Legal tests: the WTO agreements – a summary of the final act of the Uruguay Round (http://www.wto.org/english/docs_e/legal_e/ursum_wp.htm). (Accessed December 2000). WTO (2000b) Tariff quota administration methods and tariff quota fill, G/AG/NG/S8. (http://www.wto.org/WTO/ddf/ep/public. htm). (Accessed December 2000). WTO (2003) Draft Cancun Ministerial Text, (www.wto.org/english/thewto_e/minist_e/ min03_e/draft-decl_e.htm). (Accessed September 2003).

3

The Common Agricultural Policy in an Enlarged Europe: Bright or Bleak Prospects for Africa? BIRGITTE GERSFELT* AND HANS G. JENSEN† *Danish

Research Institute of Food Economics, The Royal Veterinary and Agricultural University and Department of Economics, University of Copenhagen, Rolighedsvej 25, 1958 Frederiksberg C, Denmark; †Danish Research Institute of Food Economics, The Royal Veterinary and Agricultural University, Rolighedsvej 25, 1958 Frederiksberg C, Denmark

Introduction The Common Agricultural Policy (CAP) of the European Union has been undergoing important changes during recent years. In December 2002 the EU decided to go ahead with the so-called Eastern Enlargement of the European Union, with plans to expand the EU with ten new member countries in 2004, which would also extend the domain of the CAP significantly. In June 2003 the EU subsequently reached an agreement on the so-called mid-term review (MTR) of the CAP, which entailed radical reforms of the CAP and its incentive structure from 2004 onwards. As the EU is one of the major players on the agricultural world markets as well as an important trading partner for many developing countries, the expansion of the CAP and the MTR reform may have a significant impact on world trade in agricultural commodities, thereby also affecting developing countries. But how extensive will this impact be, and will the overall effects be positive or negative for the developing countries? That is the question this chapter seeks to answer, focusing specifically on the impact on Africa, trying to provide a quantitative assessment of how the Eastern Enlargement and the MTR reform will affect agriculture in the African countries in terms of output, trade flows and overall welfare. This is accomplished by using a modified version of the global general equilibrium model GTAP, where special attention has been given to the modelling of the CAP. However, before turning to this quantitative analysis, we present a brief summary of the agricultural aspects of the Eastern Enlargement and the MTR reform of the CAP (for a more comprehensive treatment see Chapter 2). © CAB International 2006. WTO Negotiations and Agricultural Trade Liberalization: the Effects of Developed Countries’ Policies on Developing Countries (eds E. Diaz-Bonilla et al.)

57

58

B. Gersfelt and H.G. Jensen

The Eastern Enlargement of the EU After a decade of negotiations and preparations a historic agreement was reached at the European Council meeting in Copenhagen on 13 December 2002, where the accession to the EU of eight Central and Eastern European Countries (Poland, Hungary, the Czech Republic, the Slovak Republic, Estonia, Latvia, Lithuania and Slovenia) as well as Cyprus and Malta was decided. The so-called Eastern Enlargement took place on 1 May 2004. Accession to the EU has significant consequences for the agricultural sectors in the ten accession countries. The remaining border protection between the existing and the new members is removed, and the new member countries adopt the same level of border protection against Third World countries as that existing in the EU. The latter implies that for some accession countries existing tariffs on certain commodities rise significantly, while other accession countries must lower their tariffs on various commodities (see Jensen and Frandsen, 2003a). Thus, in some cases the Eastern Enlargement reduces developing countries’ access to the markets of new members, while in other cases, and other things being equal, it increases access to these markets.1 CAP market measures such as the intervention price system and the export subsidies are extended to the new member countries from day one.2 Naturally these measures are also accompanied by an extension of the EU production quota system to the new member countries. Direct payments in the form of area and animal premiums are also extended to the new member countries, but these will be phased in over a period of ten years (see European Commission, 2002). Furthermore, due to a simplified implementation scheme in the initial years as well as to the MTR reform, the majority of the premiums in the new member countries must be decoupled from production. The extension of the CAP to the new member countries is obviously costly. However, in October 2002, the EU leaders agreed on a financial framework, which stabilizes EU spending on CAP expenditures in the period up to 2013 (the so-called Schroeder–Chirac deal). The total expenditures on market measures and direct payments in the enlarged EU, comprising 25 countries (henceforth EU-25), will consequently be bounded by this financial framework.

The MTR Reform of the CAP On 10 July 2002 the European Commission presented its proposal for the so-called mid-term review (MTR) of the CAP. This review had been undertaken in accordance with the stipulations of the Agenda 2000 reform of the CAP, and was as such expected, but apparently the radical content of the proposal came as a surprise to many. However, after a year of tough negotiations the European Council finally agreed on how to reform the CAP, in June 2003.

CAP in an Enlarged Europe

59

The details of the MTR reform are presented in Chapter 2 by Frandsen and Walter-Jørgensen. Here, we focus only on some key principles in this reform to better present the simulations. The continuation of the Agenda 2000 approach to revisions of the market measures entails a reduction in the intervention prices for a number of commodities, in some cases linked to increases in the direct support payments for these commodities.3 In the case of dairy products, the intervention price for butter will be reduced by 25 per cent (which is an additional 10 per cent price cut compared to the Agenda 2000 reform), while intervention prices for skimmed-milk powder will be lowered by 15 per cent (as agreed in the Agenda 2000 reform). The price reductions are compensated through introduction of a direct payment per ton of milk quota from 2004 and onwards. Milk quotas are to be maintained to 2014–2015 (with an increase in the quotas for Greece and Portugal). For cereals, on the other hand, there will be no change in the intervention price, but the monthly increments are reduced by 50 per cent. The area premium for cereals (as well as for oilseeds, protein and setaside) remains unchanged. The most radical and innovative feature of the MTR reform is no doubt the decision to decouple some of the direct support payments from production.4 This is achieved by replacing most of the previous area and animal premiums (including the premiums for cereals, oilseeds and protein crops, as well as the premiums for beef, veal, sheep, goats and milk) with a single decoupled farm income payment from 2005 (in the case of milk, 2006–2007) onwards. This single farm income payment, which is based on historical reference of payment, is tied to the land, but the land can be used for any agricultural activity – save for certain exceptions regarding production of fruit, vegetables and table potatoes – as long as it is maintained in good agricultural condition.5 The decoupling scheme covers the majority of the direct support premiums, and if implemented fully it would have significant consequences for EU agricultural production. However, the member states are granted some discretion regarding the extent of decoupling, as they are allowed to opt for a partial implementation of the decoupling scheme.6 The direct payments are also affected by another of the main elements in the MTR reform. Modulation entails a percentage reduction in the individual farm’s direct payments after allowing for a franchise of €5000. The saved funds will be used for rural development measures (see Chapter 2, by Frandsen and Walter-Jørgensen). As modulation affects all direct support payments, this specifically also entails a reduction in the productspecific direct support payments.

Modelling the Future Common Agricultural Policy In order to investigate empirically the impact on developing countries of the Eastern Enlargement and the MTR reform, we incorporate these policy initiatives into a global CGE model and database – the Global Trade

60

B. Gersfelt and H.G. Jensen

Analysis Project (GTAP).7 The GTAP model and database are specifically tailored to the analysis of global trade issues, and our analysis is based on an explicit representation of the CAP in this model, where all the EU-25 member countries are individually represented. The GTAP model is a standard multi-regional, static, computable general equilibrium (CGE) model. Like any other applied economic model, this model is, of course, based on specific assumptions both in terms of theoretical structure as well as on the specific parameters and data used. Thus, regional production is produced according to a constant return to scale technology in a perfectly competitive environment, and the private demand system is represented by a non-homothetic demand system (a Constant Difference Elasticity function).8 The foreign trade structure is characterized by the Armington assumption, which implies imperfect substitutability between domestic and foreign goods (see Jensen and Frandsen, 2003b). The macroeconomic closure used is a neo-classical closure, where investments are endogenous and adjust to accommodate any changes in savings. This approach is adopted at the global level and investments are then allocated across regions to equalize the marginal rate of return in all regions. Although global investments and savings must be equal, this does not apply at the regional level, where the trade balance is endogenously determined as the difference between regional savings and regional investments. This is valid as regional savings enter the regional utility function. In the labour market the wage rate adjusts to ensure full employment in the model. The numeraire used in the model is the global primary factor price index (see De Melo and Robinson, 1989 and De Melo and Tarr, 1992). The following analysis uses version 5 of the GTAP database (with 1997 as the base year), modified specifically to reflect the CAP. To keep the model focused and within computational bounds the database is aggregated to 40 countries and 24 commodities, of which 12 are primary agricultural goods and seven are secondary agricultural goods.9 The GTAP model and database were then used to run three simulation scenarios, and some of the results were then further aggregated across selected commodity and country groupings. The first of these simulations is a baseline for the period 1997–2013, which serves as the benchmark against which the impacts of the Eastern Enlargement and the MTR reform are measured. The baseline features projections of the world economy and incorporates the effects of changes in the CAP as outlined in the Agenda 2000 reform, the Everything But Arms initiative (EBA),10 and the EU preferential market access for bovine meat products and other meat products from the Accession Countries (see Baker, 2002). The impact of the Eastern Enlargements and the MTR reform are then analysed in two scenarios. In the first scenario we study the effect of enlarging the EU with the ten new member countries under the current CAP regime of Agenda 2000. Thus, this so-called Agenda 2000 scenario illustrates the consequences of simply enlarging the EU without further

CAP in an Enlarged Europe

61

reforms of the CAP (although the budgetary ceiling from the Schroeder– Chirac deal is observed). The second scenario, on the other hand, features the Eastern Enlargement as well as the MTR reform of the CAP. Hence, this MTR scenario in fact illustrates the combined effects of the two events. Both scenarios thus entail the integration of the ten new member countries into the EU and the CAP, with simulations presented for the year 2013. Enlargement of the EU implies that all tariffs and export subsidies, as well as non-tariff barriers between the EU and the new member countries, are to be abolished. At the same time, all sectors in the new member countries are given the same level of protection against developing countries as that found in the EU at the time of accession. As previously mentioned, this implies that import tariffs in all the new member countries change for almost all agricultural commodities – and often significantly so. Export subsidies are also introduced for certain agricultural commodities. Regarding domestic support, the expansion of the CAP to the new member countries follows the outlines for domestic support (in terms of direct payments, production quotas and other supply management instruments) laid down by the Copenhagen Agreement. This means that in our analysis the new member countries receive 100 per cent of the CAP direct payment level, as these payments will be fully phased in by the year 2013. The baseline, as well as the specific changes in tariffs and export subsidy rates and outlines for domestic support, are fully documented in Jensen and Frandsen (2003a,b). In the Agenda 2000 scenario there are no further changes to the CAP other than those implied by enlargement of the EU with the ten new member countries. This means, among other things, that the domestic support payments, which are introduced in the new member countries, will be coupled to current land use and livestock production in this scenario. In the MTR scenario on the other hand, the CAP in the enlarged EU is reformed in accordance with the stipulations of the MTR reform. First of all, this entails a number of adjustments to the market measures for certain commodities. For instance, the supplementary durum wheat payments are reduced to €285/ha in traditional areas, €0/ha in well-established areas. Also, rice and dairy intervention prices are reduced (modelled as import tariff/export subsidy reductions) and direct payments for rice, as well as dairy premiums and additional payments to milk producers, are increased. Milk quotas in Portugal and Greece are also expanded. Second – and more importantly – domestic support payments are also decoupled in the MTR scenario in accordance with the rules laid out in the MTR agreement. However, as outlined in the previous section, EU members will have a certain amount of discretion in the implementation of the MTR reform, which means that we have to make specific assumptions about how to model the MTR reform implementation in each member country. In order to make the scenario as realistic as possible, we have based these assumptions on the political debate taking place in each member country about the options for decoupling direct payments. This has resulted in the following assumptions, as categorized in Table 3.1.

62

B. Gersfelt and H.G. Jensen

Table 3.1. Coupling/decoupling assumptions for domestic support in the MTR scenario. Cattle France

100% veal slaughter premiums 100% suckler cow and 40% slaughter premiums given to bulls, steers, cows and heifers from the age of 8 months

Denmark

75% coupling of special male premium

Sheep/ Goat

Durum wheat

Hectare premiums

50% coupling

D

25% coupling

50% coupling

D

D

Portugal

100% coupling of suckler cow and 40% coupling of slaughter premiums

50% coupling

D

D

Finland

75% coupling of special male premium

50% coupling

D

D

Greece

D

50% 40% coupling coupling

Italy

100% coupling of slaughter premiums

D

D

40% coupling

D

100% coupling of suckler cow and 40% coupling of slaughter premiums

D

40% coupling

D

100% coupling of suckler cow and 40% coupling of slaughter premiums

D

D

D

Belgium/Luxembourg 100% coupling of veal slaughter premiums

D

D

D

Germany

100% coupling of veal slaughter premiums

D

D

D

Netherlands

100% coupling of veal slaughter premiums

D

D

D

Sweden

75% coupling of special male premium

D

D

D

Spain Austria

Ireland

Full decoupling

UK

Full decoupling

All accession countries

Full decoupling

Note: D is equal to 100% decoupling of payments.

The decoupled payments of the MTR reform are modelled by converting direct aid payments in each member country into a uniform hectare payment given to all utilized agricultural areas. The results in this chapter consequently illustrate the situation where direct support is decoupled from production without enforcement of any restrictions on land use. Furthermore, all direct payments have been reduced in accordance with MTR stipulations on modulation, and the amounts saved have been reallocated to rural development following the MTR guidelines (see Chapter 2, by Frandsen and Walter-Jørgensen).11 Figure 3.1 summarizes the experimental design adopted in the analysis. The Agenda 2000 enlargement scenario thus illustrates the effect of enlarging the EU with the ten new member countries without reforming the CAP,

CAP in an Enlarged Europe

63

Baseline Scenario Projection of adjusted V5 GTAP Database from 1997 to 2013 Shocks to GDP, factor endowments and population Sector-specific shocks to total factor productivity Capital stock endogenously determined Implementation of Agenda 2000 reform of the CAP, Everything But Arms initiative, and EU preferential access for bovine and other meat products from the future member countries

Agenda enlargement EU enlargement with ten new countries Scenario 1

Implementation of the old CAP with coupled domestic support in the new member countries No change in CAP in old member countries

MTR enlargement EU enlargement with ten new countries Scenario 2

The isolated effect of the MTR reform of the CAP is quantified by subtracting scenario 1 from scenario 2

Implementation of the new CAP with decoupled domestic support in the new member countries Implementation of the MTR reform of the CAP in the old member countries

Fig. 3.1. Experimental design.

whereas the MTR-plus-enlargement scenario shows the effect of enlarging the EU with the ten countries while also reforming the CAP in accordance with the MTR agreement. As shown in Fig. 3.1, the pure effect of the MTR reform in an enlarged EU may be derived by comparing the MTR scenario results with the Agenda enlargement scenario results, as any difference between these two sets of results can be attributed to the MTR reform. It should be noted that the scenario of primary interest is the MTR enlargement scenario, whereas the Agenda 2000 scenario is mainly an auxiliary scenario, which helps us in decomposing the results in the MTR enlargement scenario.

Results for Production in EU-25 As the effects of the Eastern Enlargement and the MTR reform on developing countries are a result of the impacts these reforms have on EU agricultural production and trade, the natural starting point for this analysis is to consider the changes in EU-25 agricultural production.12

64

B. Gersfelt and H.G. Jensen

The first column of results in Table 3.2 shows the aggregate supply response for EU-25 in the MTR enlargement scenario, while the second and third columns of results serve to decompose these aggregate supply responses by showing which amount of the change is attributable to the enlargement and which to the MTR reform. Considering the example of cereals, we see that the combined effect of enlarging the EU and implementing the MTR reform will reduce the index of cereal production to 93.7 from 100, i.e. a 6.3 per cent decline in EU-25 production of cereals. Decomposing this result, we see that an enlargement of the EU with no reform of the CAP (i.e. the Agenda scenario) would result in a 1.8 per cent increase in EU-25 production of cereals, while a subsequent MTR reform in the enlarged EU-25 would bring about a 8.1 per cent reduction in the aggregate cereal production (relative to the baseline production level). In total, this means we get a 1.8 + (
8.1) =
6.3 percentage change in total cereal production in the MTR enlargement scenario. Looking at the overall supply response in the MTR enlargement scenario, we see that the most dramatic changes relate to the three important commodity groups: cereals, bovine animals and untreated milk, as volume of production for these three commodities declines by, respectively, 6.3, 7.2 and 7.1 per cent. Furthermore, the decomposition shows that in the case of cereal and bovine animal production these declines are the result of the MTR reform. It is not surprising that decoupling of the cereal area premiums and the bovine animal premiums under the MTR reform leads to a significant decline in the production of these commodities. However, it may seem strange that an extension of the old and coupled Agenda 2000 CAP to the new member countries apparently would not result in large increases in the production of cereals and bovine animals, despite the prevalence of partially coupled support for these commodities under Agenda 2000. However, the aggregate EU-25 response conceals the fact that production of these commodities in the new member countries increases under an Agenda 2000 enlargement, but at the same time cereal and bovine animal productions is decreased slightly in the existing EU member countries, resulting in minor or non-existent effects on aggregate production. Turning to the case of untreated milk, we see that the decline in production in the MTR enlargement scenario is primarily attributable to the enlargement effect. This is due to the fact that extension of the CAP to the new member countries also entails the introduction of milk quotas in these countries. These quotas are binding in all the new member countries and lead to an overall reduction in their untreated milk production of 35 per cent, which results in the total EU-25 production of untreated milk declining by 6 per cent. The reason why the MTR reductions in intervention prices and decoupling of direct payments to milk result only in an additional 1.1 per cent drop in untreated milk production is that the milk quotas continue to be binding in most countries after the implementation of the MTR reform (see Jensen and Frandsen, 2003b).

MTR Enl. Primary Agricultural Commodities

Index

Cereals (incl. paddy rice) Vegetables, fruit, nuts Oilseeds Sugar cane and beet Plant-based fibres Other crops Bovine animals Other animals Untreated milk Wool

93.7 103.4 99.6 99.9 97.9 103.3 92.8 100.7 92.9 106.2

Contribution from Enl.

Change in index 1.8
1.0 0.1
0.4
3.5
2.0 0.0
0.3
6.0
2.5

MTR Enl.

MTR


8.1 4.4
0.5 0.3 1.4 5.2
7.2 1.0
1.1 8.7

Contribution from Enl.

MTR

CAP in an Enlarged Europe

Table 3.2. Change in volume of production in EU-25, baseline 2013 = index 100.

Change in index

Secondary Agricultural Commodities

Index

Bovine meat products Other meat products Vegetable oils and fats Dairy produce Sugar Other processed foods (incl. rice)

95.7 99.7 99.4 97.2 100.0 100.1

0.1
0.6
0.5
1.6
0.1 0.1


4.4 0.3
0.1
1.2 0.1
0.1

Other Commodities Beverages and tobacco Textiles/wearing apparel Natural resources Manufactures Services

99.0 103.5 99.9 100.2 100.0


1.1 3.5
0.1 0.1
0.1

0.1
0.1 0.0 0.1 0.0

Note: The reported EU-25 results are constructed by aggregating the results for the 25 individual EU member countries. The results for wheat and other grains (which form the major part of the aggregate ‘cereals’) and the results for bovine animals, other animals and untreated milk are aggregated using FAOSTAT quantities, while the remaining results are aggregated using GTAP value shares.

65

66

B. Gersfelt and H.G. Jensen

While decoupling under the MTR reform pulls resources out of the sectors which used to receive coupled direct payment, it also pushes resources into sectors which were previously unsupported or which continue to receive coupled support, as these sectors have become relatively more profitable following the decoupling under the MTR reform. The MTR enlargement scenario consequently shows an increase in the production of e.g. other crops and vegetables and fruit and nuts of 3.3 and 3.4 per cent, respectively.13 However, it should be noted that our modelling of the MTR reform builds on the assumption that there will be no restrictions on the use of agricultural land receiving the single farm payment (which is, however, not the intent of the MTR reform when it comes to production of vegetables, fruit and nuts; see the above section on MTR reform of the CAP). Should these restrictions on land use be enforced, our analysis overestimates the expansionary effect in the vegetables, fruit and nuts sector. However, the analysis does show that the MTR reform will increase the incentives to produce these commodities, and if the area restrictions are not stringently enforced, EU-25 production will indeed increase. Summing up the EU-25 supply response results for the MTR enlargement scenario, we see that for most commodities the effect of MTR reform dominates the effect of enlarging the EU (with the production of untreated milk being a notable exception). The changes in composition of EU-25 agricultural production are therefore characterized by the shifts induced by the decoupling scheme, which entail a shift of production away from sectors which used to receive coupled direct payments (e.g. cereals and bovine animals), and into sectors which were previously unsupported or remained unreformed (e.g. other crops and vegetables, fruit and nuts). These changes in EU-25 agricultural production do not affect only the EU: they also affect the rest of the world, including developing countries, through international trade.

Results for EU-25 Trade with Africa Due to historical ties and geographic proximity, the EU is typically an important trading partner for African countries. For many of these economies agricultural exports are a significant source of export revenue. However, at the same time a number of African countries are not selfsufficient in food production, making food imports a vital necessity. It is therefore reasonable to speculate whether the changes in European agricultural production following the Eastern Enlargement and the MTR reform will have significant impacts on the African countries through their trade with the EU. In analysing the European–African trade flows it is useful to distinguish between EU exports to Africa and EU imports from Africa, as these trade flows may to some extent be influenced by different policy instruments and mechanisms. Therefore, Fig. 3.2 shows the change in value of EU agricultural exports to Africa, while Fig. 3.3 shows the change in

CAP in an Enlarged Europe

67

0 –100 –200 –300 –400 –500 –600 –700 North Africa

South African Customs Union Agenda enlargement

Rest of Africa

MTR enlargement

Fig. 3.2. Change in value of EU agricultural exports to Africa, US$ million (1997).

0 100 –200 –300 –400 –500 –600 –700 North Africa

South African Customs Union Agenda enlargement

Rest of Africa

MTR enlargement

Fig. 3.3. Change in value of EU agricultural imports from Africa, US$ million (1997).

value of EU agricultural imports from Africa.14 In order to distinguish between the effect of the Eastern Enlargement and the effects of the MTR reform, the figures depict these changes in both the Agenda 2000 scenario and the MTR scenario. As mentioned previously, the scenario of interest is the ‘realistic’ MTR enlargement scenario. However, the Agenda 2000 enlargement scenario shows us which part of the effect in the MTR enlargement scenario that is attributable to the Eastern Enlargement, while the difference between the effects in the two scenarios is attributable solely to the MTR reform. The first thing to note, when looking at these two figures, is that in both scenarios we see a reduction in both EU agricultural exports to, and EU agricultural imports from, Africa. Furthermore, these reductions are larger in the MTR enlargement scenario, implying that both the Eastern Enlargement and the MTR reform in itself led to reductions in the value of

68

B. Gersfelt and H.G. Jensen

EU trade with Africa. The fact that the MTR reform effects are generally significantly larger than Eastern Enlargement effects mirrors the fact that the MTR reform effect dominated the supply response results for EU-25 in the previous section. The total effect in the MTR enlargement scenario amounts to the EU reducing the value of its agricultural exports to Africa by 1997 by US$480 m., while reducing the value of its agricultural imports from Africa by US$773 m. Following the Eastern Enlargement and the MTR reform, the African agricultural trade balance vis-à-vis the EU-25 thus deteriorates by US$293 m. In order to account for the differences in level of development across the African continent, Africa has been disaggregated into three regions – North Africa (comprising Morocco, Algeria, Tunisia, Libya and Egypt), the South African Customs Union (comprising South Africa, Lesotho, Swaziland, Namibia and Botswana) and the Rest of Africa. Overall, this categorization broadly reflects the level of development, with the countries of North Africa and the South African Customs Union being middle-income countries, whereas most of the countries in the category Rest of Africa are low-income countries. Given these differences in levels of development, as well as differences in resource endowment across the three regions, we might expect the regions to be affected differently by the changes to EU agriculture, and the figures confirm this suspicion. In terms of EU agricultural exports to Africa, the North African region is clearly more affected in absolute terms than the two other regions, whereas in the case of EU agricultural imports from Africa, the region Rest of Africa experiences the largest absolute impact. In the remaining part of this section we will consequently focus on these two examples, starting with the issue of EU exports to North Africa.

EU exports to North Africa In order to further analyse the decline in the value of EU-25 agricultural exports to North Africa, Fig. 3.4 shows the decomposition of the change in value of EU-25 exports by commodities. Figure 3.4 shows that the changes in EU-25 agricultural exports generally mirror the changes in EU-25 agricultural production. The changes in value of EU-25 agricultural exports are thus dominated by the effect of the MTR reform, and in the MTR enlargement scenario we consequently see a decline in the value of EU cereal exports and bovine meat exports, whereas the value of EU exports of other crops and vegetables, fruit and nuts increases slightly. However, in the case of EU dairy exports, it is the Eastern Enlargement effect which is primarily responsible for the decline in value of EU agricultural exports. The decline in value of EU-25 agricultural exports to North Africa in the MTR enlargement scenario reflects a decline in the volume of EU-25 agricultural exports to this region. This in turn means that North Africa’s import demand for these commodities from other countries increases,

CAP in an Enlarged Europe

69

200 150 100 50 0 –50 –100 –150 –200 –250 –300 Cereals Vegetables, (incl. paddy fruits, nuts rice)

Other crops

Bovine Other meat Dairy meat products products products

Agenda enlargement

Sugar

Other processed foods (incl. rice)

MTR enlargement

Fig. 3.4. Change in value of EU-25 exports to North Africa, US$ million (1997, selected commodities).

which raises the average North African import price on cereals, bovine meat products and dairy products by 0.7, 3.9 and 4.7 per cent, respectively. Following the Eastern Enlargement and the MTR reform, North Africa will thus have to cover some of its import needs for cereal, bovine meat products and dairy products from other sources and at higher prices or, alternatively, increase its own production of these commodities.

EU-25 imports from the Rest of Africa Figure 3.3 showed that the decline in value of EU-25 agricultural imports from Africa stems primarily from the decline in value of EU-25 imports from the Rest of Africa. Figure 3.5 decomposes this latter trade flow by commodity. Not surprisingly, we once again see that the majority of the changes in the value of EU-25 agricultural imports from the Rest of Africa are concentrated in a few commodity groups, namely sugar, the category ‘other food products’ and the category ‘other crops’. In the case of other crops the decline in the value of EU-25 imports of US$256 m. in the MTR enlargement scenario (relative to the 2013 baseline scenario) is clearly more than 100 per cent attributable to the MTR reform, as an Eastern Enlargement under Agenda 2000 would actually have produced a US$144 m. increase in the value of EU-25 imports of other crops. This change in the trade flow of other crops reflects the change in EU-25 production of other crops, as an Eastern Enlargement under Agenda 2000 would result in declining EU-25 production of these commodities, while the MTR reform, on the other hand, causes an increase in the production of other crops due to the decoupling of direct payments for many of the other commodities.

70

B. Gersfelt and H.G. Jensen

200 150 100 50 0 –50 –100 –150 –200 –250 –300 Cereals Vegetables, (incl. paddy fruits, nuts rice)

Other crops

Bovine meat products

Agenda enlargement

Other meat products

Dairy products

Sugar

Other processed foods (incl. rice)

MTR enlargement

Fig. 3.5. Change in value of EU-25 imports from Rest of Africa, US$ million (1997, selected commodities).

However, we see a different pattern in the case of sugar and other processed food products. For these commodities, the decline in the value of EU-25 imports from the Rest of Africa that we see in the MTR enlargement scenario is due to the enlargement effect, as this decline would also occur in the case of EU enlargement under the Agenda 2000 CAP.15 It therefore appears that we have a case of trade diversion, which occurs when the formation – or in this case the extension – of a preferential trade agreement (PTA) causes a PTA member to replace imports from non-members with imports from a less cost-effective PTA member due to the PTA tariff preferences. Thus, the developing country loses its export to the PTA member as the PTA tariff preferences provide other PTA members with an artificial competitive advantage in the PTA market (see Chapter 14, by Nielsen, which provides a general discussion of PTAs). However, as usual, reality is more complicated, as we are in fact dealing with multiple PTAs in this case. The reason is that a number of countries in the Rest of Africa are classified as least-developed countries (LDCs), and the EU’s Everything But Arms initiative (EBA) entails a non-reciprocal removal of all EU restrictions on imports from LDCs (except for imports of arms) (see Yu and Jensen, 2003). Prior to the Eastern Enlargement, the African LDCs consequently had better access to the old EU-15 market than the new EU member countries did. The EBA initiative grants the LDCs duty and quota-free access to the EU markets for virtually all products (see Chapter 13, by Yu and Jensen). Under the EBA scheme ‘potential’ EU tariff revenue is thus transferred to LDC exporters through their duty-free access to the EU markets, which implies that these LDC exporters receive higher prices on their EU exports than non-preference-receiving exporters.16 However, the LDC exports are subject to rules-of-origin restrictions under the EBA scheme, which implies that not all exports coming out of the EBA countries will qualify for duty and quota-free access to the EU (see Brenton, 2003).

CAP in an Enlarged Europe

71

In the present analysis, the EBA agreement is modelled as duty-free access for the LDCs to the EU internal market. Consequently, the LDC exporters are receiving higher prices (rents) for their exports to the EU than their non-EBA competitors, as the EU imposes import tariffs on the exports of the latter, thereby effectively reducing the export prices received by these non-EBA countries. In terms of trade volumes, the EBA agreement results in the LDCs exporting greater volumes than they otherwise would have, since receiving a higher export price than their competitors obviously affords them a competitive advantage in the EU markets.17 However, it should be noted that we have not accounted for the effects of rules-of-origin restrictions in our analysis, and we have implicitly assumed that the EBA preferences are fully utilized in the year 2013. Ceteris paribus, this will tend to overemphasize the significance of the EBA scheme. The special preference obtained by the LDCs under the EBA initiative obviously rests on the fact that non-LDCs generally do not have duty- and quota-free access to the EU markets. However, in the course of the Eastern Enlargement, the LDCs’ special preferences in the EU-15 market are partly eroded as the new member countries become part of the internal EU market. This implies that the LDC exports of, e.g. sugar to the old EU member countries may be replaced by sugar from the new EU members. However, this would clearly not be due to the new EU members gaining an unfair advantage over the LDCs, rather it entails a levelling of the playing field for the LDCs and the new EU member countries in the old EU-15 market.18 On the other hand, the classic PTA trade diversion effects may arise for those African countries not classified as LDCs, if they do not enjoy other types of preferential access to the EU markets. The decline in the MTR enlargement scenario of EU-25 demand for other crops, sugar and other food products from the Rest of Africa causes the region’s export prices for these commodities to decline by 1.0, 1.1 and
0.4 per cent, respectively. As the following section will show, this in turn has consequences for the production of these commodities in the Rest of Africa.

Derived Impact on Production and Welfare in North Africa and the Rest of Africa As we have seen, the Eastern Enlargement of the EU and the MTR reform of the CAP affects the EU’s agricultural trade with Africa, and this will subsequently affect not only African agricultural production but also the overall measure of welfare in African economies. Agricultural production in Africa We first consider the MTR enlargement scenario consequences for agricultural production in Africa. Table 3.3 shows the changes in volume of production in the two regions North Africa and Rest of Africa, as well as the changes in value of EU-25 exports to and imports from these regions, in

72

Table 3.3. Changes in volume of agricultural production in North Africa and the Rest of Africa. North Africa Change in Value of EU-25 Exports to

Imports from

US$ mill.

Rest of Africa

Change in production Contributions MTR enlarg.

Enlarg.

2013 = index 100

Change in production

Change in Value of EU-25

MTR

Exports to

Change in index

Imports from

US$ Mill.

Contributions MTR enlarg.

Enlarg.

2013 = index 100

MTR

Change in index


218 35 0 0 4 23 8 0
1 0

5
47 0 0
2
5
1 0 4
1

100.9 99.2 100.0 100.2 99.8 97.6 99.9 99.9 100.5 99.1


0.1 0.1 0.0 0.2 0.1 0.3 0.0 0.0 0.4
0.9

0.9
0.8
0.1 0.0
0.3
2.8 0.0 0.0 0.1 0.0


63 13 0 0 0 13 2 0
1 0


22
52 2
1 3
256 0
1 0 0

99.7 99.7 100.5 98.4 100.7 98.8 100.2 100.1 100.3 100.2


0.4 0.1
0.1
1.7
0.4 0.7
0.1 0.0 0.1 0.0

0.1
0.4 0.6 0.1 1.1
1.9 0.3 0.1 0.2 0.2

Bovine meat products Other meat products Vegetable oils and fats Dairy products Sugar Other processed foods (incl. rice)


55 0 6
171 14 9

0 0
7 1
1
25

116.7 98.8 99.0 107.1 99.7 99.8


0.1 0.3
0.1 5.0
0.1
0.1

16.8
1.5
0.9 2.1
0.2
0.2


23
3
3
27 4 2

9 2 1 1
162
161

102.9 100.3 100.5 105.1 97.5 99.6

0.1 0.1 0.0 1.4
2.7
0.4

2.8 0.2 0.5 3.7 0.2 0.0

Total


345


78

–

–

–


86


637

–

–

–

B. Gersfelt and H.G. Jensen

Cereals (incl. paddy rice) Vegetables, fruit and nuts Oilseeds Sugar beet and cane Plant-based fibres Other crops Bovine animals Other animal products Raw milk Wool

CAP in an Enlarged Europe

73

order to illustrate clearly the connection between changes in trade flow and derived changes in production. Starting with the results for the North African region, the table shows that as the EU-25 exports of cereals, bovine meat products and dairy products to this region decline, and the region’s import prices of these commodities rise (by 0.7, 3.9 and 4.7 per cent, respectively), domestic production becomes more profitable and the volume of domestic production increases by 0.9 per cent for cereals, 16.7 per cent for bovine meat products and 7.1 per cent for dairy products.19 The table furthermore shows that changes in the EU–North African trade flow also result in a marginal decrease in the production of vegetables, fruit and nuts, as the value of EU exports of these commodities to North Africa increases by US$35 m., while the value of EU imports from North Africa declines by US$47 m. (amounting to a net trade effect of US$–82 m. for North Africa). Turning to the results for the Rest of Africa we see that the decline in EU-25 imports of other crops, and the ensuing drop in the region’s export price for these commodities of 1.0 per cent, results in the production of other crops declining by 1.2 per cent in the region. In the case of sugar we also see the decline in EU-25 imports from the Rest of Africa manifested in a 2.5 per cent decline in production of processed sugar. Furthermore, this effect is transmitted to the region’s primary production of sugar cane and beet, which declines by 1.6 per cent. The decline in EU-25 imports of other food products results only in a minute decline in the Rest of Africa’s production of these commodities, of 0.4 per cent. In the previous section on EU trade with Africa we focused on the effects of EU-25 agricultural imports from the Rest of Africa. However, changes in the regional volume of production stem from the combined effects of changes in the region’s exports and imports. Table 3.3 shows that the value of EU-25 exports of cereals, bovine meat products and dairy products to the Rest of Africa decline by US$63, 23 and 27 m. (1997), respectively (parallel to the pattern we saw for North Africa, although the magnitudes are smaller for the Rest of Africa). In the case of bovine meat and dairy products these changes translate into increases in domestic production of these commodities, whereas cereal production in the region is left virtually unchanged. Summing up, the analysis shows that the changes in EU-25 trade flows to and from the African regions translate into corresponding changes in African production of the affected commodities. However, it is also clear from Table 3.3 that the Eastern Enlargement of the EU and the MTR reform of the CAP are not going to cause dramatic changes to the African agricultural production patterns, as the changes in volume of production are indeed minor for most commodities.

Welfare implications for Africa So far we have considered how the Eastern Enlargement and the MTR reform affect the African regions in terms of changes in the key variables

74

B. Gersfelt and H.G. Jensen

relating to agricultural trade and production. But how do these changes affect the African economies overall? One way of answering this question is to consider the impact which the Eastern Enlargement and the MTR reform have on economic welfare, measured as the money metric value of the equivalent variation (EV)20. Table 3.4 consequently shows the absolute changes in EV in the MTR enlargement scenario, as well as which part of the changes in EV stems from the MTR reform and which part stems from the Eastern Enlargement. Furthermore, Table 3.4 also shows the relative (i.e. percentage) change in EV in the MTR enlargement scenario. Table 3.4 shows that the enlarged EU stands to gain approximately US$13 bn in the MTR enlargement scenario, while Africa as a whole will realize a minor welfare loss of US$484 m. The region Rest of Africa accounts for 49 per cent of the African welfare loss, while North Africa accounts for 38 per cent. As the statistics show, the EU would clearly be able to compensate the African economies for the welfare loss they will suffer as a consequence of the Eastern Enlargement and the MTR reform of the CAP. It is interesting to note that although the MTR reform reduces some of the distortionary effects of the CAP, the effect of the MTR reform itself accounts for over half of the African welfare loss (contributing to this welfare loss by US$272 m.), whereas an Eastern Enlargement under the old and more distortionary CAP produces a welfare loss of US$211 m. for Africa. However, looking at the decomposition of these changes in total EV, we see that the majority of the African welfare losses in both cases stems from adverse terms-of-trade effects. These terms-of-trade effects result from a combination of changes in world market prices and changes in trade volumes following the Eastern Enlargement and the MTR reform, which in turn translate into positive or negative welfare effects, depending on the region’s net trade positions. In the case of North Africa the deterioration in the terms of trade are, of course, largely due to increases in the region’s import prices for cereals, bovine meat products and dairy products, while in the Rest of Africa almost one third of the trade loss is linked to trade of other crops. However, it must be noted that, as shown in the second column of Table 3.4, these changes in African welfare are small in terms of changes in baseline welfare (a decline of
0.06 per cent). Of the individual countries shown, Malawi and Zimbabwe are faced with the largest relative reductions in welfare, but these still amount only to
0.22 and
0.21 per cent, respectively. Furthermore, the welfare gain to the EU following the Eastern Enlargement and the MTR reform clearly leaves ample scope for compensating the African countries for their welfare losses.

Conclusions The fundamental question underlying the analysis in this chapter has been whether the Eastern Enlargement of the EU and the MTR reform of the

MTR enlargement scenario

Contributions from Enlarg.

of which Percentage change Total EV

Contributions from MTR

of which

Allocative Terms of efficiency trade

Allocative Total EV efficiency

of which

Terms of trade

Total EV

Allocative efficiency

Terms of trade

EU-25 0.14 North Africa
0.06 Botswana 0.01 Rest of South African Customs Union
0.04

13099
182 0
64

11476
54 0
35

1793
127 0
31

7278
126
3
22

5832
52 0
6

1611
72
3
18

5821
57 3
42

5644
2 1
29

182
55 3
14

South African Customs Union Malawi Mozambique Tanzania Zambia Zimbabwe Other Southern Africa Uganda Rest of sub-Saharan Africa


0.04
0.22 0.02
0.13
0.02
0.21
0.17
0.03
0.07


64
8 1
17
1
17
38
4
153


34
2 0
5 0
3
11
1
30


32
7 1
10
1
14
34 –3
108


25 2 1
12
1 1
33 3
22


6 0 0
3
1 1
10 0 2


21 1 1
6
1 0
31 3
18


39
10 0
5 0
18
4
7
132


28
2 0
2 0
3
1
1
31


11
8 1
3 0
14
3
6
91

Rest of Africa


0.08


237


50


175


61


11


50


176


40


125

Africa


0.06


484


139


334


211


68


143


272


70


190

ROW


0.01


2934


1502


1502


2550


1110


1509


384


392

8

0.03

9681

9835


42

4516

4654


41

5165

5181


1

Total

CAP in an Enlarged Europe

Table 3.4. Change in Economic Welfare million 1997 US$

Note: Economic welfare is measured as the money metric value of the Equivalent Variation (EV); ROW – Rest of the World.

75

76

B. Gersfelt and H.G. Jensen

CAP would be beneficial or detrimental to the African economies. The results indicate negative impacts on African welfare, but the analysis also shows that the overall effects on the African economies are generally small. The Eastern Enlargement and the MTR reform thus appear to be primarily European matters, with only minor effects for economies of developing countries such as those in Africa. The analysis also showed that in many cases the effects of the MTR reform dominate the effects of the Eastern Enlargement. This may, however, not be that surprising considering that in our analysis the MTR reform effect is defined as the difference between enlarging the EU under the old Agenda 2000 CAP and enlarging it under a new MTR reformed CAP. Thus the MTR reform effect does not only capture the liberalization gains from undertaking reforms in the old EU-15, it also captures the effect of going from Agenda 2000 to the new MTR reformed CAP in the new member countries. Of course Agenda 2000 is not actually being introduced in the new member countries, so in this sense our MTR effect may overestimate the impact of the MTR reform. On the other hand, distortionary agricultural policy regimes had been established in many of the new member countries prior to the enlargement, meaning that the introduction of the MTRreformed CAP in these countries will entail a significant degree of liberalization. In the course of the former and the current WTO negotiation rounds there has been constant pressure on the highly distortionary agricultural policy regimes in the developed countries. Part of the argument has emphasized how these policy regimes are detrimental to the developing world. It may, therefore, seem surprising at first that the MTR reform will apparently not be beneficial to the African countries. However, this may actually not be that surprising when we consider that reducing the incentives for overproduction in the EU of e.g. cereals, bovine animals and milk, will ceteris paribus benefit the other major producers of these commodities, such as the USA, Australia and New Zealand, while developing countries typically do not have comparative advantages in the production of these commodities. Also, these simulations have not modified current high levels of protection for agricultural products in the EU. Elimination of distorting subsidies without changing other market access distortions may lead to complex and paradoxical welfare results. In any case, the European leaders have decided in favour of both the Eastern Enlargement of the EU and the MTR reform of the CAP, and the present chapter has explored the consequences of these developments. While these reforms entail major welfare gains for the enlarged EU, prospects for African countries are not encouraging, although the impacts are generally small. What remains to be seen now is whether the continuing WTO negotiations will succeed in furthering additional reforms, including market access, a policy change not considered in these simulations – hopefully also to the benefit of the developing world.

CAP in an Enlarged Europe

77

Notes 1

The EU’s Everything But Arms initiative (EBA), which entails a non-reciprocal removal of all EU restrictions on imports (except import of arms) from Least Developed Countries (LDCs) (see Yu and Jensen, 2003), will also apply in the new member countries after the enlargement. Thus LDCs’ access to accession country markets will improve following the Eastern Enlargement. 2 Except in the case of export subsidies to sugar, where special rules apply (see Jensen and Frandsen, 2003a). 3 Minimum producer prices (intervention prices) are guaranteed to producers for certain agricultural produce in the EU, through a combination of sales at floor prices to a buffer stock agency and measures taken at the border. Lowering of the intervention price reduces the floor price at which stock agency starts purchasing. Direct support payments are given to farmers as compensation for reductions in intervention prices. 4 The theoretical concepts of coupled and decoupled support refer to how closely the amount of support is linked to the size of current production. Fully coupled support entails a direct link between the size of current production and the amount of support (the typical example being price support), whereas fully decoupled support means that the amount of support is completely independent of current production (the typical example being direct support payments, which solely depend on historical production in a fixed reference period, and hence do not require any current production) (see also Frandsen et al., 2003). 5 The exact restrictions on the production of fruit and vegetables and table potatoes depend on whether the single farm payment is implemented according to the Fischler model or the regional model. Under the former model the farmer cannot receive the single farm payment on land used for growing these crops, whereas under the latter model the farmer may receive the single farms payment on land used for the production of these crops, provided that the total amount of land used for production of fruit and vegetables and table potatoes does not exceed the amount of land used for production of these crops in the reference period. Whether this regional model will constrain the production of fruit and vegetables and table potatoes in reality depends on how the restrictions are administered and how the yields for these crops develop (see Jensen and Frandsen, 2003c). 6 They can do that in different ways: (i) Retain 25 per cent of the hectare premium or alternatively up to 40 per cent of the supplementary durum wheat aid; (ii) Retain up to 50 per cent of the sheep and goat premiums; (iii) Retain up to 100 per cent of the suckler cow premium and up to 40 per cent of the slaughter premium, or instead retain either up to 100 per cent of the slaughter premium or alternatively up to 75 per cent of the special male premium component; and (iv) Make additional payments for the purposes of encouraging specific types of farming which are important for the protection or enhancement of the environment and for improving the quality and marketing of agricultural products. 7 References for the GTAP model and database are Hertel (1998) and Dimaranan and McDougall (2002). The model is solved using GEMPACK (Harrison and Pearson, 1996). 8 Hence, the present analysis abstracts from features such as imperfect competition and increasing returns to scale, which may, however, be important in certain sectors. 9 The commodity group beverages and tobacco is not considered an agricultural commodity in this study (see Table 3.1). 10 The EBA initiative is modelled in the baseline period by reducing all EU tariff rates to zero on imports from Malawi, Mozambique, Tanzania, Zambia, Uganda and the two GTAP aggregate regions Other Southern Africa and the Rest of sub-Saharan Africa. For an evaluation of the EBA initiative see Chapter 13, by Yu and Jensen.

78

B. Gersfelt and H.G. Jensen 11

In the present analysis, the distribution of direct payments within each class of payment (< 5000, > 5000 €), aggregated to country level for the old EU-15 members countries, is based on working document No. 12 from the Council Working Party (Council of the European Union, 2003). From these data, aggregate national rates of reduction have been calculated and used to modulate direct payments. 12 As in all quantitative studies, the results naturally depend on the data and the assumptions applied in the model, including the nature of the baseline and scenarios simulated, the value of key parameters and elasticities, and the level of commodity and regional aggregation. These caveats should of course be kept in mind when considering the estimated effects on trade, production and welfare. 13 In the GTAP database, other crops is a residual category of crops not classified elsewhere. Details of crops included in this category can be found on the GTAP home page www.gtap.agecon.purdue.edu under technical papers. 14 Agricultural exports and imports are comprised of the commodities listed as primary agricultural commodities and secondary agricultural commodities in Table 3.1. 15 However, the results for other processed foods should not be overemphasized, as they are largely driven by increases in EU-25 production of processed rice, which are exaggerated. 16 Apart from the EBA preferential trading framework, the EU also grants preferential access to a number of developing countries under the Generalized System of Preferences (GSP) and the Cotonou agreement (which replaced the Lomé conventions). For further information on these preferential trading schemes and their relation to the EBA initiative see Chapter 13, by Yu and Jensen. 17 We have not included the Cotonou agreement in our analysis, as the trade component of the Cotonou agreement is merely a framework for the EU and African, Carribean and Pacific countries (ACP) to renegotiate the so-called Regional Economic Partnerships Agreements (REPAs), which will gradually come into force after 2008 at the latest (see Chapter 13, by Yu and Jensen). It is currently not clear what these REPAs will entail and therefore not possible to assess what effect they might have in the year 2013, in which our scenarios take place. 18 It should also be noted that while the Eastern Enlargement results in an erosion of the special preferences enjoyed by LDCs in the old EU-15 market, it also entails an extension of the EBA to the new EU member countries’ markets, which should make it considerably easier for LDCs to export to these countries. The question is then whether the extended geographical coverage of the EBA can make up for the preference erosion in the old EU-15 markets. 19 The value of bovine meat products in North Africa is estimated to be US$13 m in 2013, which increases to 16 million after the enlargement of the EU. Compared to the value of primary agricultural production of bovine animals, which amounts to US$10178 m in 2013, the secondary industry of bovine meat production (commercially slaughtered animals) is a small industry in North Africa. Therefore, the relatively large percentage increase in bovine meat production in North Africa reflects a large percentage increase in a small industry. 20 For further information on EV and its decomposition, see Huff and Hertel (2000).

References Baker, A.D. (2002) Agriculture in the EU’s Eastern Enlargement – current status for the ACs. Danish Research Institute of Food Economics, Report no. 144, Copenhagen.

Brenton, P. (2003) Integrating the Least Developed Countries into the World Trading System: The Current Impact of EU Preferences under Everything But Arms.

CAP in an Enlarged Europe World Bank Policy Research Working Paper 3018, April 2003, Washington DC. Council of the European Union Council Working Party (2003) ‘Horizontal Regulation: Direct Support Schemes’. Working document No. 12, ‘Determination of the National Ceilings for the Additional Amount of Aid’ (Art. 11) DS 73/03, Brussels, 24 February 2003. De Melo, J. and Robinson, S. (1989) Product differentiation and the treatment of foreign trade in computable general equilibrium models of small economies. Journal of International Economics 27, 489–497. De Melo, J. and Tarr, D. (1992) A General Equilibrium Analysis of U.S. Foreign Trade Policy. MIT Press, Cambridge, Massachusetts. Dimaranan, B.V. and McDougall, R.A. (2002) Global Trade, Assistance, and Production. The GTAP 5 Data Base, Center for Global Trade Analysis, Purdue University, West Lafayette, Indiana. European Commission (2002) Fact Sheet – Enlargement and agriculture: A fair and tailor-made package which benefits farmers in accession countries. DN: Memo/02/301, 20/12/2002. http://europa.eu.int/rapid/start/ cgi/guesten.ksh?p_action.gettxt=gt&doc= MEMO/02/301|0|RAPID&lg=EN&display= Frandsen, S.E., Gersfelt, B. and Jensen, H.G. (2003) The impacts of redesigning European agricultural support. Review of Urban & Regional Development Studies 15(2), 106–131. Harrison, W.J. and Pearson, K.R. (1996)

79 Computing solutions for large general equilibrium models using GEMPACK. Computational Economics 9, 83–127. Hertel, T.W. (ed.) (1998) Global Trade Analysis, Modelling and Applications. Cambridge University Press, UK. Huff, K.M. and Hertel, T.W. (2000) Decomposing Welfare Changes in the GTAP Model. GTAP Technical Paper No. 5, West Lafayette, Indiana. Jensen, H.G. and Frandsen, S.E. (2003a) Implications of EU Accession of Ten New Members – The Copenhagen Agreement. Danish Research Institute of Food Economics Working Paper 01/03, Frederiksberg, Denmark. Jensen, H.G. and Frandsen, S.E. (2003b) Impacts of the Eastern European Accession and the 2003 reform of the CAP – Consequences for Individual Member Countries. Danish Research Institute of Food Economics, Working Paper 11/03, Frederiksberg, Denmark. Jensen, T.V. and Frandsen, S.E. (2003c) Reformen af den fælles landbrugspolitik – danske og europæiske konsekvenser. In: Frandsen, S. (ed.) Landbrugets Økonomi – Efteråret. Danish Research Institute of Food Economics, Frederiksberg, Denmark. Yu, W. and Jensen, V. (2003) Tariff Preferences, WTO Negotiations and the LDCs – The case of the ‘Everything But Arms’ Initiative. Danish Research Institute of Food Economics Working Paper 04/03, Frederiksberg, Denmark.

4

US Agricultural Policy: the 2002 Farm Bill and WTO Doha Round Proposals1 DAVID ORDEN Markets, Trade and Institutions Division, International Food Policy Research Institute, 2033 K Street NW, Washington DC 2006-1002, USA

Introduction In May 2002 the US Congress enacted and the President signed into law a new 6-year domestic farm bill, the Farm Security and Rural Investment Act (FSRIA). The new law replaced what was known as the ‘freedom to farm’ legislation of 1996 that received attention when it was enacted for potentially marking the end of US farm subsidies. If Congress had adhered to the 1996 law (formally titled the Federal Agriculture Reform and Improvement (FAIR) Act) in the late 1990s both the level and the year-toyear variability of previous farm support outlays would have been reduced. But when a 3-year run of high crop prices collapsed in 1998, implementing the FAIR Act cost more for price support payments than anticipated and Congress authorized additional support payments on an ‘emergency’ basis. The 2002 FSRIA continued or expanded the FAIR Act support programmes that provide producer price guarantees and fixed direct payments for wheat, feed grains, soybean and other oilseeds, rice and cotton. It also restored a third tier of counter-cyclical support for a large portion of farm output in place of the emergency payments. Passage of the new US farm bill was met with derision by domestic policy critics and a barrage of international condemnation. Nobel laureate Joseph Stiglitz disparaged the new farm support law as ‘the worst form of political hypocrisy’, while Malloch Brown, head of the United Nations Development Program, accused US policy of ‘holding down the prosperity of poor people in Africa and elsewhere for very narrow, selfish interests’. In reply, the US House Agriculture Committee offered a strident defence of US farm policy, arguing it was ‘important to national security, ensuring a safe, abundant, and affordable domestic food supply’. A document, posted on the Committee’s web page, made the claim that ‘Critics of US farm policy 80

© CAB International 2006. WTO Negotiations and Agricultural Trade Liberalization: the Effects of Developed Countries’ Policies on Developing Countries (eds E. Diaz-Bonilla et al.)

US Agricultural Policy

81

would cede our food production to unstable places like the Third World’, and asked ‘but in these times does any American want to depend on the Third World for a safe and abundant supply of food and fibre?’. These sharply worded and disparate views of the US farm bill are indicative of the festering global conflict over agricultural policies and trade. Yet both severe critics of the FSRIA and its staunch defenders overstate their cases in a situation that is complex in many dimensions. The 2002 US farm bill takes few, if any, constructive unilateral steps toward reduction of subsidies. Nor does it expand the worst subsidy policies as abhorrently as sometimes implied. Shortly after the FSRIA was signed into law, the USA submitted a proposal on agriculture for the World Trade Organization (WTO) Doha Round negotiations. The initial US proposal was out of step with the extensive support provided to farmers by its new domestic legislation. One provision of the US WTO proposal called for multilateral reductions of all tariffs on agricultural products to 25 per cent or less over a 5-year period. Likewise, the initial US WTO proposal called for consolidation of agricultural support measures into just two categories: exempt support having at most minimal trade-distorting effects on production; and nonexempt support that affects production and would be subject to WTO commitments with a ceiling after 5 years of 5 per cent of the value of aggregate national agricultural output. These were relatively strong proposals for multilateral agricultural policy disciplines. Unfortunately, the misalignment of domestic farm policy and the early US Doha proposal soon resulted in subsequent weakening of the US negotiating proposal.2 In August 2003, the USA joined the European Union in putting forward a proposal deemed by key developing countries as offering too little subsidy restraint. This set the stage for a mid-term collapse of the Doha Round negotiations on agriculture. To provide some understanding of the origins of this impasse, this chapter provides a brief description of the US farm sector and evolution of US policy in the 1996 FAIR Act. It then focuses on provisions of the 2002 farm bill and the subsequent US position that emerged in the WTO negotiations through the July 2004 framework agreement.

Background on the US Farm Sector and Policies American agriculture today scarcely resembles the troubled sector of the Depression-era 1930s that led to farm support programmes. The modernization of American agriculture has created a tri-modal farm sector. At one end are the most efficient commercial farms producing the bulk of food and fibre; at the other end are various small farms that account for most of the enumerated units but produce only a small part of output. In the middle is a group of farmers caught in the dynamics of modernization – the mid-sized farms on which there have been substantial investments and on which there remain full-time employment opportunities, but which may

82

D. Orden

lack an adequate resource base to be competitive in face of continuing advances in technology and market integration. Reforms of farm policy have been undertaken as the production and income of farmers have undergone changes. The basic direction of policy reform has been a shift in policy instruments from acreage supply controls – combined with price supports above market-clearing levels – to less supply intervention and more direct income support, at least for crops that are exported. This policy evolution toward direct payments began in the mid 1960s when price support levels were lowered for maize, wheat and cotton to enhance US competitiveness, and farmers were offered direct payments as compensation. Support payments from the government increased from less that 6 per cent of farm income in the 1950s to over 20 per cent in the 1960s. A second move toward direct payments came in the mid-1980s, when price supports set in anticipation of inflation that did not materialize were reduced, with direct payments once again offered to farmers as compensation. Still further steps in the direction of replacing market interventions with direct payments were taken in the 1996 FAIR Act. Thus, the reform path Congress took in 1996 was the familiar one of a heavily compensated ‘cash-out’ of farm programmes.3 Even so, some critical changes in farm programme instruments were made.

Unilateral farm policy reform in the 1996 FAIR Act The 1996 FAIR Act initiated four unilateral changes in US farm policy compared to previous legislation. First, under the FAIR Act, supported farmers attained flexibility to plant whatever crops they chose (excepting most fruit and vegetables) on ‘base acreages’.4 Second, authority ended for the USDA to require annual acreage idling to limit crop supplies. Third, farmers received fixed income transfers, known as production flexibility contract (PFC) payments, that were based on past production and were independent of current market prices and farmers’ planting decisions. These fixed income transfers replaced earlier ‘deficiency payments’ that had required continued production of the crop for which payments were received. Fourth, the price guarantees made to crop producers for any amount of output through ‘loan rates’ were capped under the FAIR Act at nominal levels well below market prices prevailing at the time. By 1996, mechanisms had also been put fully in place for most crops that allowed farmers to receive a cash payment (a ‘marketing gain’ or ‘loan deficiency payment’ (LDP)) if market prices (determined for each county for wheat, feed grains and oilseeds and by a common ‘effective adjusted world price’ for rice or upland cotton) were below their loan rate levels. Farmers received these cash payments instead of forfeiting those crops into government-owned storage. Thus the loan rates continued to support prices for producers, but market prices were freed from the loan rate as a floor level and the government was extricated from cumbersome commodity stockpiling.

US Agricultural Policy

83

The changes to farm policy made in 1996 were new partial reforms along the cash-out lines of movement toward direct income transfers instead of land idling or government stock-holding to push prices above free market-clearing levels. Farmers responded to the increased flexibility the FAIR Act allowed with substantial movements away from the crops to which deficiency payments previously had been tied, particularly reducing wheat acreage and expanding planting of soybean. Despite its innovations, the extent to which the FAIR Act put farm policy on a less interventionist or less costly path was uncertain from the outset. The market-oriented policy innovations in the FAIR Act came at a time of high crop prices in 1995 and 1996. It is unlikely that farm policy would have abandoned annual acreage idling had market prices not surged upward. Since prices did rise, agricultural proponents in Congress were able to tout the end to acreage setasides and the introduction of fixed payments as deregulation of a large part of agriculture. Such ‘freedom to farm’ had been a rallying point for the Republican Party since the 1950s – the last time before 1995 that Republicans had controlled Congress and been in a position to set the farm policy agenda. Yet even Republican proponents of these agricultural policy changes knew full well that while the FAIR Act gave farmers more cropping flexibility it also increased support expenditures in the short term because deficiency payments under the old farm programme were falling as prices rose. Farmers liked the short-term outcomes of the FAIR Act of less regulation of their production and more direct payments. When challenged that the new farm policy nevertheless undermined longer-term support levels, proponent Pat Roberts (Republican, Kansas), then chairman of the House Agriculture Committee, opined that Congress itself was the long-term safety net. This has turned out to be the case.

Re-institutionalizing higher farm support in 2002 After spiking upward in 1995 and 1996, crop prices began to fall in 1997 and remained low through 2001. As prices fell, support expenditures built into the FAIR Act increased automatically because of the price guarantees provided by loan rates. The loan rate-related expenditures jumped up to $1.8 billion in calendar year 1998, then $6.8 billion in 1999, $7.5 billion in 2000 and $6.2 billion in 2001 (Table 4.1). Once prices fell sharply, the PFC payments and built-in increased expenditures for price guarantees under the FAIR Act provided less support to farmers than would have been available under earlier farm programmes. Critics of freedom to farm decried it as ‘freedom to fail’, with low prices, reduced support and absence of a strong farm ‘safety net’. A Congress closely divided on party lines couldn’t resist responding to the criticism, stepping in firstly with ‘emergency’ legislation and then with supplemental annual appropriations for additional direct payments (called ‘market loss assistance’ (MLA) payments), as well as with new disaster

84

D. Orden

Table 4.1. Government payments to farmers 1996–2000 (US$ million). (From USDA.) 1996

1997

1998

1999

2000

2001

2002 (F)

Pre-FAIR Programmes CRP, Other Env. AMTA Payments Marketing Loan payments/Gains Emergency Assistance


732 2,099 5,973 0


575 1,950 6,120 0


5 1,475 6,001 1,792

na 1,494 5,046 6,814

na 1,615 5,049 7,551

na 1,803 4,040 6,172

na 1,845 9,916 4,026

0

0

2,841

7,804

8,492

8,405

908

Total*

7,340

7,495

12,380

21,513

22,896

20,727

16,971

*Includes small miscellaneous not shown. Note: F = forecast, Env. = environmental.

relief and crop insurance subsidies. The effects on support policy were first to speed up delivery of scheduled fixed payments, then to increase their levels by 50 per cent, and finally to double the payments in 1999, 2000 and 2001. Subsequently, the 2002 farm bill locked in new expenditures in the event of low prices.

Support Provisions of the 2002 Farm Security and Rural Investment Act The final 2002 FSRIA was signed into law by the President on 13 May 2002. Although budget assessments for the legislation were based on its provisions remaining in effect for ten years up until 2011, the new law was only specified for six years (to 2007).5

Three-tiered price and income support for grains, oilseeds, rice and cotton The final FSRIA incorporates three tiers of support for grains, oilseeds, rice and cotton. Direct payments are continued at rates similar to those provided by PFC payments under the FAIR Act and are added for soybean and other oilseeds, which had not been included in 1996 (Table 4.2). Most loan rates were raised compared to the maximum levels under the FAIR Act for 2002 and 2003, but then the amounts by which the rates went up were reduced for 2004–2007 (Table 4.3). Loan rates were added for several commodities (dry peas, lentils, small chickpeas and groundnuts, as shown in Table 4.3, and for mohair, wool and honey). The FSRIA also fixed the loan rates in nominal terms, removing the discretion of the Secretary of Agriculture to lower the rates based on an average of past market prices that was provided in 1985 – when the government was still taking the supported crops into storage if market prices fell below loan rate levels. Once LDPs and marketing gains came into effect, market prices below loan rates no longer resulted in crops going into government storage, and the pressure to keep loan rates below market price levels was lessened.6

US Agricultural Policy

85

Table 4.2. Direct payments. Crop

1996 Farm Bill rate (in 2002)

2002 Farm Bill rate (2002–2007)

0.261 0.314 0.461 0.0572 2.05 0.202 0.022 na na na

0.028 0.35 0.52 0.0667 2.35 0.24 0.024 0.44 0.008 36.00

Maize ($/bu) Sorghum ($/bu) Wheat ($/bu) Upland cotton ($/lb) Rice ($/cwt) Barley ($/bu) Oats ($/bu) Soybeans ($/bu) Minor oilseeds ($/lb) Groundnuts ($/ton)

Table 4.3. Loan rates. Crop Maize ($/bu) Sorghum ($/bu) Wheat ($/bu) Upland cotton ($/lb) Rice ($/cwt) Barley ($/bu) Oats ($/bu) Soybeans ($/bu) Minor oilseeds ($/lb) Groundnuts ($/ton) Dry peas ($/cwt) Lentils ($/cwt) Small chickpeas ($/cwt)

1996 Farm Bill 2001 rate

2002 Farm Bill 2002–2003 rate

2002 Farm Bill 2004–2007 rate

1.89 1.71 2.58 0.5192 6.50 1.65 1.21 5.26 0.093 na na na na

1.98 1.98 2.80 0.52 6.50 1.88 1.35 5.00 0.096 355.0 6.33 11.94 7.56

1.95 1.95 2.75 0.52 6.50 1.85 1.33 5.00 0.093 355.0 6.22 11.72 7.43

The target prices in the FSRIA for the new counter-cyclical payments are shown in Table 4.4. In contrast to loan rates, the target prices for most crops were increased after 2003. Counter-cyclical payments are made when the sum of the market price (or loan rate if the market price is lower) and the fixed direct payment is less than the target price. In the final FSRIA, farmers retain flexibility to plant a range of crops – thus they do not necessarily produce the crops for which they receive fixed and counter-cyclical payments. Both the direct fixed payments and counter-cyclical payments are made on 85 per cent of base acreage and ‘payment yield’ determined under the bill. Each participant has to make a once-only decision about bases and yields that will then determine their payment eligibility for the duration of the FSRIA. The rules in the FSRIA for determining base acreage are the same for both the fixed and counter-cyclical payments, but rules for setting programme yields differ for the two types of payment.

86

D. Orden Table 4.4. Target prices. Crop

2002–2003

2004–2007

Maize ($/bu) Sorghum ($/bu) Wheat ($/bu) Upland cotton ($/lb) Rice ($/cwt) Barley ($/bu) Oats ($/bu) Soybeans ($/bu) Minor oilseeds ($/lb) Groundnuts ($/ton)

2.60 2.54 3.86 0.724 10.50 2.21 1.40 5.80 0.0980 495.0

2.63 2.57 3.92 0.724 10.50 2.24 1.44 5.80 0.1010 495.0

Programme payment yields for direct payments are unchanged for those crops previously covered under PFC payments. However, those farmers who update their base acreage are also given options to update yields for the counter-cyclical payments. This distinction between the two support programme in part reflects WTO considerations. The fixed payments had been reported to the WTO by the USA as Green Box payments that did not affect production. By not allowing yield updating, the USA reduced the likelihood of a challenge to the classification of these payments, even though updating of the base acreage was allowed. The counter-cyclical payments, in contrast, were likely to be reported as an Amber Box policy, similarly to the MLA payments. Even thought they were made on a fixed acreage and yield once a producer enrolls in the support programme, and do not require production of specific crops, the counter-cyclical payments are explicitly linked to market prices. Since the counter-cyclical payments were expected at the time to be reported as WTO Amber Box, no claim of being exempt from commitments was being made, and yield updating did not pose the threat of a challenge to their classification. Of course, because the payments will not necessarily be made on crops actually being produced, the USA was also likely to report these payments as not commodity specific, which made the WTO constraint on their use less binding.7 The final FSRIA includes only modest payment limitations for individual producers. Annual payment limits on direct payments, countercyclical payments and marketing loan payments (LDPs and marketing gains) are $40,000, $65,000 and $75,000, respectively. A ‘three-entity’ rule is retained that allows any individual to receive a full payment directly and up to a half payment from two additional entities. Thus, the maximum payment that an individual can receive is $360,000 per year. Producers with average adjusted gross income over three years of $2.5 million or more are not eligible for payments unless at least three-fourths of their adjusted gross income is from agriculture. For the payments related to loan rates, the limitations on individual eligibility are undermined by special ‘commodity certificates’ that enable producers who are facing payment limits to continue to benefit from repayment rates below the loan rates.

US Agricultural Policy

87

Divergent programmes for sugar, dairy and groundnuts The FSRIA continues special programmes for sugar, dairy and groundnuts. These commodities have long been protected by import restrictions and high tariffs. Domestic consumer prices have been sustained above world market prices by also restricting domestic supplies, or by government stockpiling purchases, when necessary. Thus, the support programmes for these three commodities had undergone less of the conversion to direct payments and reduced market intervention by 2002 than had occurred for the main crops. Under the FSRIA, the programmes for sugar, dairy and groundnuts became more divergent. Sugar The USA is a large net sugar importer. In 1996, the FAIR Act continued the traditional sugar programme, with domestic sugar price supports fixed nominally at 18 cents per pound for raw cane sugar (about one-half of domestic sugar production) and 22.9 cents per pound for refined beet sugar (the other half of production). Sugar could be forfeited at these rates to USDA’s Commodity Credit Corporation (CCC) under ‘non-recourse’ loans (for which the commodity collateral is accepted in lieu of repayment). Thus, no basic liberalization of the sugar market was achieved, and the loan rates continued to provide a floor under domestic market prices. As prices for crops fell after 1997, a critical policy point for sugar emerged. By 2000 domestic sugar production, plus minimum imports to which the USA was committed, was about to exceed demand for domestic consumption and private stock-building at the supported domestic prices. To sustain those prices, the CCC accumulated sugar stocks and the USDA offered a sugar ‘plough-down’, in which it exchanged stockpiled sugar for destruction of some of the planted new sugar beet crop. The alternative of letting the domestic sugar price fall was rejected at this time. To avoid plowing down a growing crop in the subsequent year, a payment-in-kind (PIK) programme was initiated to trade CCC-stockpiled sugar for reduced beet planting. Supply pressure on the sugar market then eased in 2001, lessening political pressure for reform. In the 2002 FSRIA, domestic producers succeeded in tightening the provisions of the sugar support policies. The loan rates were continued at the same levels as the FAIR Act and other adjustments made the sugar programme more lucrative for producers. The new farm bill reinstated an earlier stipulation that the sugar programme be operated to the extent possible at no net cost to the government. To make these new sugar provisions operational, authorization for a PIK was continued and authority was restored for USDA to control supply through domestic marketing allotments, but only as long as imports were below 1,360,000 metric tons. The combination of the no-net-cost provision and the constraint on use of domestic marketing allotments – if imports exceeded the level set in the legislation – was designed, in the words of the US

88

D. Orden

producers, to ensure that the USDA and US trade representatives stood ‘shoulder to shoulder’ with the domestic industry in opposing loosening of import restrictions. Together these provisions tie the hands of policy administrators: imports above 1,320,000 metric tons can not be offset by domestic marketing allotments to sustain the supported price, while allowing imports to exceed this level would induce violation of the no-net-cost provision if CCC stockpiling were to result. Thus, under the FSRIA the sugar programme has to continue to be administered with tight import restraints, which sets the farm bill firmly against sugar trade liberalization. Dairy For dairy products, import restrictions under TRQs remain the primary instrument for sustaining domestic prices above world levels. Related dairy provisions of the domestic farm bill are among the most complex among farm programmes. The 2002 FSRIA extends the two main dairy programmes: purchases by the CCC to support the price of milk used for various processed (manufactured) products and federal milk marketing orders that regulate markets for the fluid milk consumed directly. Under the FAIR Act, the dairy price support programme was scheduled to end on 31 December 1999, but it was extended in subsequent legislation. Under the 2002 FSRIA, the milk price support programme is renewed again. Support purchase prices are set to ensure that the market price of milk used in processing averages at least $9.90 per hundredweight. To provide this price support, the CCC is authorized to buy necessary quantities of butter, Cheddar cheese or non-fat dried milk. The Secretary of Agriculture retains the authority to adjust product purchase prices as deemed necessary. Fluid milk markets are regulated by federal and state milk marketing orders retained under the FSRIA. Milk marketing orders define the relationship between prices of fluid and manufactured dairy products and maintain a regulated geographic price structure. One modest ‘cash-out’ innovation under the FSRIA involved a new national dairy payment programme. The Dairy Market Loss Payments (DMLP) programme provided counter-cyclical payments for dairy producers until September 2005. These payments were made to dairy farmers on a monthly basis if the market price of fluid milk (called Class I) in Boston (Federal Marketing Order 1) was less than $16.94 per hundredweight. Payments were limited to 2.4 million pounds of milk per year per operation, which corresponds to the production from a relatively small dairy herd of about 135 cows. With this limit, about 50 per cent of total national milk production was likely to be eligible for the direct payments, but only about 30 per cent of the total production was from the smaller operations that produce less than the 2.4 million pound limit. For these small producers, the counter-cyclical payments created an incentive to expand production at the margin because the per-unit price they received was supported at the target price level. For the larger farms producing about 70 per cent of the milk in the USA, the direct payments programme was

US Agricultural Policy

89

essentially decoupled from production – it provided a variable payment on a fixed output that was inversely related to the price of milk. This is similar to the counter-cyclical crop support programme in the FSRIA. Groundnuts Groundnuts are a minor crop but evaluation of the changes in policy in 2002 is informative about the economic and political pressures determining farm programmes. Under the 2002 FSRIA a regime of domestic price supports well above world levels for edible groundnuts, combined with long-established quotas on the domestic production eligible for sale in the US market, was scrapped in favour of direct cash payments – a substantial cash-out reform. Under the traditional support programme, domestic quota holders received preferential prices for groundnuts supplied to the domestic market for direct edible use compared to prices received for groundnuts (known as ‘additionals’) that went into processing (crushing into oil and meal) or were exported. Access to the domestic edible market by foreign competitors was restricted by TRQs. Thus, the traditional groundnut programme created an income stream from higher prices reserved exclusively for those domestic and foreign farmers who had privileged access. Even domestic farmers without quotas were barred from producing groundnuts for the domestic edible market but could produce groundnuts as additionals. The 1996 FAIR Act had included some policy changes in the groundnut programme. The loan rate for quota groundnuts for the domestic edible market was lowered from $678/ton to $610/ton, a minimum national quota was also eliminated, and geographical production restrictions were partially relaxed.8 Elimination of the minimum national quota allowed USDA to set an annual quota poundage eligible for the domestic market based on demand estimates. The annual effective quota poundage was reduced from 1.47 million tons for the 1995 crop year to 1.15 million tons in 1996, and averaged 1.24 million tons during 1996–2000. Despite the reduced quota, domestic groundnut production remained nearly constant. The average national production for 1996–2000 was 1.82 million tons, or 99 per cent of the average of 1.85 million tons for the years 1993–1995. As a result, under the FSRIA groundnut producers were selling a relatively smaller proportion of their output at a lower quota support price for domestic consumption, and a relatively higher proportion of their groundnuts at much lower prices in the additionals market. One reason for the declining effective quota for the domestic edible market was the international trade agreements – to which the USA committed in the 1990s – that increased foreign access to the US groundnut market. Foreign producer access to the US domestic market for groundnuts increased from less than 4 per cent of consumption prior to the 1993/1994 marketing year to over 10 per cent by the 1999/2000 marketing year, due to market access provisions of the WTO and NAFTA. Moreover, under NAFTA the tariff rate for groundnuts is scheduled to decline to zero for Mexico in 2008, so imports are likely to rise. As long as the price in the US

90

D. Orden

domestic market remained above the price in world markets, as it had under the groundnut quota programme, other foreign producers also had incentives to seek additional access in trade negotiations. The 2002 FSRIA, however, made a fundamental change to the groundnut programme. Under the FSRIA groundnuts receive support through policies similar to other crops. The quota-based dual market structure is replaced with a support programme of direct payments that includes the basic three components: loan rates and related payments, fixed direct payments and counter-cyclical payments. In addition, groundnut quota holders are compensated for their loss of quota rights. The new groundnut programme is quite lucrative for both former quota holders and for producers of groundnuts that were once sold as additionals. The cash-out has an estimated cost of $4 billion over 10 years. Under the FSRIA, any producer of groundnuts is eligible for a loan rate of $355/ton on all current production. Those who qualify as an ‘historic producer’ of quota or additional groundnuts are also guaranteed a direct fixed payment of $36/ton and a target price under the counter-cyclical payment programme of $495/ton for the output from 85 per cent of historic groundnut acres and recent yields. Thus, for a traditional producer who continues to grow groundnuts, the minimum average revenue would be $474/ton on a level of production equal to recent output ((0.85)  ($495) + (0.15)  $355 = $474). The traditional groundnut producers also attain planting flexibility. They can receive the fixed payments ($36/ton) and the counter-cyclical payments while growing another crop if that is deemed more profitable. If groundnuts are grown, the new guaranteed revenue is much higher than that received in the past by additional producers, who had only been eligible for a loan rate of less than $200/ton. And for 5 years, the quota holders attain an additional $220/ton. Thus, for 5 years the total guaranteed revenue was $694/ton for a quota owner, compared to $610 under the FAIR Act. After 5 years, guaranteed revenue for a quota holder falls below the previous guarantee, but the quota buy-out of $220/ton for 5 years compares favourably with market prices for many sales of quota rights before the 2002 farm bill was passed. Apparently, these market prices included a discount for the possibility that the groundnut quota programme would not last forever. There are a number of additional political economy aspects to the cashout adopted for groundnuts in the USA in 2002. Rising imports and potential trade liberalization that increased foreign access to the domestic edible market were used as arguments to motivate a policy change as being necessary to ‘preserve the domestic industry’. The preservation argument was central – the intent of the new groundnut support policy is to sustain the domestic industry, not to cause its demise. The future of the domestic industry can not be guaranteed under the FSRIA because farmers can switch from groundnut production under planting flexibility, whereas under the earlier quota system groundnuts had to be grown to attain the high domestic price. But domestic producers are well compensated for lower market prices under the FSRIA, and incentives for domestic groundnut production are improved for both the traditional growers of additionals

US Agricultural Policy

91

and new entrants. In contrast, one group that is disadvantaged by the change in US groundnut policy is foreign producers who had attained TRQ market access. Unlike domestic producers, the foreign producers do not receive any payments as compensation for the lower US domestic groundnut prices. Under the FSRIA, with lower prices access to the US market is of less value to foreign producers, but the USA is better positioned to liberalize groundnut trade. This positioning is quite in contrast to the sugar policies adopted in 2002. It is noteworthy that US sugar producers did not endorse a cash-out reform similar to groundnuts in 2002, instead opting to tighten restrictions under their traditional price support programme. The sugar producers faced the same farm bill budget as groundnut producers, and could have sought to have some of the money Congress made available for agricultural support used for sugar payments. Sugar potentially faces even more pressure from imports (tariffs on sugar imports from Mexico also are eliminated in 2008 under NAFTA and world sugar prices remain far below the US levels). Sugar producers are widely held to be a more powerful lobby than groundnut producers. The sugar industry didn’t seek new cash-out payments under these circumstances; apparently the industry expects to hold on to its current support programme for some time. Yet trade liberalization for agriculture is unlikely to occur without being accompanied by some sort of cash-out for sugar producers in the USA. Part of the reason the sugar industry did not endorse a cash-out in 2002 lies in the domestic structure of the industry. Cane sugar is characterized by large production units, in Florida in particular, making payment limits per operation a political obstacle to the adoption of direct support. A second reason lies in the prospective short-term cost of a sugar cash-out. For each penny of payments per pound of sugar under a marketing loan programme, the cost is around $180 million, assuming payments on recent levels of output. The PIK programme reduced sugar beet acreage by about 6 per cent in 2001, which, all else being equal, reduced total domestic sugar production by about 3 per cent. If, instead of constraining supply market prices had been allowed to fall below the loan rate with compensating cash payments, the programme cost would have risen from $200 million to as high as $1 billion per year, depending on the price responsiveness (elasticity) of demand. Marketing allotments and PIK programmes were anticipated as being in use for at least several years in 2002, implying that a cash-out would prove costly over this time period. Longer term, the cost would depend on uncertain supply and demand conditions, including future international trade agreements, as well as on the supply and demand responses to lower prices.

Conservation programmes Conservation and environmental programmes play an important role in agricultural production decisions. Through these programmes, producers receive cost-share, rental and other direct payments in return for

92

D. Orden

using specified farming practices or for setting aside land to aid conservation. The FSRIA continues and, in most cases, expands various conservation/environmental programmes. The programmes that retire environmentally sensitive land from crop production are extended, but most new expenditures were targeted for conservation measures for livestock operations and land that stays in production. Idling of farmland under the Conservation Reserve Program (CRP) has been the primary conservation/environmental programme in effect since 1985. The final FSRIA increased the land-idling authority of the CRP to 39.2 million acres, compared to 36.4 million under the FAIR Act. This increase in CRP acreage will add marginally to its output-reducing effect. The Environmental Quality Incentives Program (EQIP), which provides technical assistance, cost sharing and incentive payments to assist livestock and crop producers with conservation and environmental improvements, is expanded under the FSRIA. Cost sharing (up to 75 per cent) or incentive payments can be provided for a wide range of practices, including nutrient management, livestock waste handling, conservation tillage, terraces and filter strips. EQIP is unique in its relative focus on livestock producers. Under the FSRIA, a new Conservation Security Program (CSP) was also initiated. The CSP is focused on land-based practices and specifically excludes livestock waste-handling facilities. Producers will develop and submit a conservation plan to USDA that identifies the resources and designated land to be conserved. The plan can include conservation practices that fall within one of three tiers provided in the programme. Producers entering into first-tier conservation security contracts will receive a base payment for conducting the practices designated in the conservation plan. Producers may also be eligible for bonus payments for implementing additional (tiers two and three) conservation measures. The new mix of conservation support programmes under the FSRIA calls attention to the policy discretion involved in US programmes regarding acreage idling for environmental purposes. While the USA has maintained the CRP and related long-term land-idling since 1985, it is not under any international obligation to do so. Historically, the USA has enacted conservation land-idling as a supply control measure during times of low prices (the 1930s, the 1960s and again in 1985) and has let these programmes expire when market demand is relatively strong, as in the 1970s. Competitors in world markets don’t object to land-idling in the USA, which reduces US production and gives the foreign producers a competitive advantage, but the CRP has been criticized for unnecessarily restricting output and pushing world prices for basic grains higher than otherwise. Were the USA to shift more fully toward support for use of environmental practices on land that continues in production in the future, along the lines of the CSP, output could expand, but competitors in world markets would have little basis for objections under the WTO or other trade agreements. The conservation programmes of the FSRIA also bring attention to the effects of domestic environmental regulations on agricultural competitiveness.

US Agricultural Policy

93

Should EQIP or CSP payments be considered as production subsidies? Once domestic regulations are enacted, requiring certain environmental performance, producers are obliged to comply. The EQIP expenditures reduce compliance costs of producers. Under an alternative approach, these could be viewed as costs that should be borne by producers (the polluter pays) that might affect agricultural production levels. Thus, the EQIP expenditures can be considered production subsidies, but under WTO rules any subsidies that offset (but do not exceed) environmental costs of measures undertaken by producers are eligible for classification in the Green Box and are exempt from limit commitments. Likewise, subsidies under the CSP are, in principle, offsetting costs related to maintaining environmental quality, and thus qualify as being in the WTO Green Box, even if adoption of the supported practices is not required by domestic regulations.

Trade provisions The FSRIA continues and modifies programmes designed to develop and expand commercial outlets for US commodities in world markets and to provide international food assistance. Increased emphasis is placed on high-value and value-added products and export programmes oriented toward development of commercial markets. The subsidizing Export Enhancement Program (EEP) is re-authorized in the FSRIA, although this programme has not been used for crops in recent years because of the loan rate payments, which have allowed market prices for crops to fall as low as required in order to make US products competitive in world markets. For dairy products, US domestic prices can exceed world levels. The Dairy Export Incentive Program (DEIP) is extended and pays cash subsidies that allow dairy product exporters to buy US products and sell them abroad when international prices are below domestic prices. Removing products from the domestic market with the DEIP helps sustain domestic price levels and thus plays a role in the milk price support programme. The DEIP quantities and dollar amounts are subject to WTO limit commitments. There was also a question of whether total payments for non-exempt crop support under the FSRIA would violate the US Amber Box limit commitment in the WTO Uruguay Round Agreement on Agriculture (URAA). The FSRIA requires the Secretary of Agriculture ‘to the maximum extent practicable, to adjust domestic commodity program expenditures to avoid exceeding allowable WTO domestic support ceilings’. Although exceeding the US commitment is possible with very low prices, it is unlikely. The USA may continue to use the de minimis exemption for noncommodity-specific support to reduce the probability that the new FSRIA counter-cyclical payments will count against its WTO constraint. A number of countries are finding other clever tricks that can be used to meet the letter of their WTO commitments without substantially changing the support provided to agriculture.

94

D. Orden

Finally, the FSRIA introduces several new technical trade barriers. Country of origin labelling was mandated by 2004 for meats and fresh fruit and vegetables. Increased importation for fattening in the USA of feeder cattle from Mexico and feeder pigs from Canada provided much of the impetus for some agricultural producers to lobby for these new rules, but their implementation remained delayed through 2005. In order for livestock to be labelled as a product of the USA, the animals have to be born, raised and processed domestically. The country of origin labelling requirement would impose implementation costs that could reduce access of foreign products to the US market or lead to price discounts for foreign producers. In the case of a few specialty products (catfish and ginseng) the FSRIA includes definitions that preclude foreign products from the domestic market.

The US in the Doha Round: Shifting Away from Reforms After the 2002 Farm Bill The URAA committed WTO members to further agriculture negotiations starting in 2000. These sectoral negotiations had not progressed very far when they were incorporated into the Doha Round. The Doha negotiations thus offer a test of the hypothesis that the Uruguay Round set the stage for more substantial subsequent agricultural trade liberalization, and possibly the opportunity to strengthen the disciplines on agricultural support and protection. Agriculture emerged high on the agenda as the Doha Round was launched in November 2001. Adoption shortly thereafter of the costly 2002 US farm bill then drove a wedge between the USA and the Cairns Group of agricultural exporters, and even allowed the EU to fault the USA for increasing subsidies. Despite the setbacks to reform in the 2002 US farm bill and these political considerations, shortly after the farm bill was signed into law the USA submitted a strong initial proposal on agriculture for the Doha Round negotiations. The initial US proposal called both for changes in the instruments through which reduction commitments are measured and for significant new disciplines on the levels of agricultural tariffs and subsidies. The USA proposed that the linear tariff reduction rules of 1995–2000 be replaced by the use of a ‘Swiss formula’ for future cuts that would bring high tariffs down more rapidly than low tariffs. A maximum tariff rate of 25 per cent would apply after 5 years. The initial US proposal also called for the Blue Box exemption from WTO limitation commitments to be abolished. Any former Blue Box expenditures were to be subjected, along with other nonexempt support, to an aggregate ceiling of 5 per cent of the total value of agricultural production. The US proposal called for a complete elimination of direct export subsidies and export taxes and phased reductions in controls on agricultural trade by state enterprises. Special safeguard provisions regarding TRQ commodities would be eliminated and

US Agricultural Policy

95

clarifications made about the use of export credits (and other forms of implicit trade subsidization) and about how TRQs were administered. The US proposal acknowledged that special and differential treatment would apply to developing countries. These criteria were spelled out in some cases and left vague in others. Overall, the early US Doha Round proposal, like the US ‘zero option’ in the Uruguay Round, called for substantial reductions of trade barriers in world agriculture. Yet in several respects, the US proposal was out of step with the extensive support provided to farmers by the 2002 legislation. In other respects, the US proposal was designed not only to bring substantial trade liberalization but to do so in ways that were particularly easy on its own farm programmes. If other countries ‘called the US’ bluff’, the US has relatively few high protective tariffs that would have to reduced sharply with the Swiss formula. It also seemed easy for the US to call for eliminating Blue Box exemption, since the US unilaterally gave up support programmes tied to annual land-idling when it adopted ‘freedom to farm’ policies eliminating such restrictions (but, as described above, not eliminating paid longer-term land-idling under conservation programmes) in its 1996 farm bill. The burdens of reform called for on export subsidies and state trading enterprises by the US proposal would mostly fall elsewhere. Still, enacting the initial US proposal would have imposed some significant restrictions on it own farm programmes. Bringing down US tariffs on sugar and dairy products would require changes in entrenched domestic support programmes. The proposed limit for non-exempt domestic support of 5 per cent of the aggregate value of total agricultural production would leave the USA with a cap of about $10 billion, just half of its current WTO limitation (not counting de minimis, which the US proposal did not address). Meeting that cap would entail reducing support, unless clever new schemes to shield expenditures were concocted. Thus, at face value, the initial US Doha Round proposal could have disciplined the agricultural subsidies and protection of the USA compared to current levels, as well as those of other WTO member countries. As in the Uruguay Round, most aspects of the US negotiating proposal were looked at favourably by the Cairns Group of developed and developing country agricultural exporters, while the EU and other countries with protected agriculture found it as unacceptable as the earlier zero option had been. The EU put forth its own Doha Round position, calling for a repeat of the linear Uruguay Round tariff cuts (15 per cent minimum and 36 per cent average); reduction of Amber Box expenditures (by 55 per cent), while maintaining the Blue Box exemption; partial elimination of export subsidies (a 45 per cent cut); and provisions related to various concerns such as animal welfare and added protection of geographical indications of product origin under international law. Japan and a few other countries mostly opposed agricultural trade liberalization. Brazil, India and China called for substantial cuts or elimination of domestic support by developed countries, elimination of export subsidies,

96

D. Orden

large cuts in tariffs at least by the developed world (an ‘offensive strategy’), and for retaining or expanding various forms of special and differential treatment for developing countries (a ‘defensive strategy’). Efforts to forge consensus by striking various compromises among these positions failed to achieve an agreement in spring 2003, as called for in the Doha negotiating schedule. With the positions of governments far apart, as a September 2003 WTO ministerial meeting in Cancún approached, in August the USA and the EU negotiated a brief framework document; convergences between the USA and the EU, at least in terms of a basic negotiating framework, came earlier in the Doha Round than it had in the Uruguay Round. The US–EU document became the basis for the intended Doha Round mid-term ministerial text, but it met with the strong resistance of the new ‘G-20’ group of developing countries when presented in Cancún. The US–EU proposal struck some awkward compromises and was short on details – no levels were specified for the various measures proposed in the negotiating framework, leaving the document at best a template for much contention in the future. For example, the proposal called for reducing Amber Box support policies ‘in the range of [] per cent – [] per cent’, without providing numbers. One of the biggest changes from the initial US proposal came not only in retention of the Blue Box, but in the weakening of the eligibility criteria for its use by excluding the requirement that these payments be associated with production limiting programmes. This about-turn in approach represented catching up of the US negotiating stance with changes made in the 2002 farm bill; its renewed counter-cyclical support payments would fit in the redefined Blue Box but not under the original Uruguay Round criteria. More positively, the US–EU proposal called for some reduction in de minimis payment limits, a limit of 5 per cent of production value for the Blue Box, and a net limit on Amber Box, Blue Box and de minimis of significantly less than their sum in 2004. On export subsidies, the US-EU proposal called not for full elimination, but only for elimination for unspecified products of interest to developing countries over an indeterminate period, together with unstipulated cuts to expenditures for other commodities. Essentially, this backed off from the stronger initial US proposal at the insistence of the EU. Similar disciplines were to be applied to any trade-distorting elements of governmentsubsidized export credit programmes. Tariff reductions were to fall into four categories of unspecified relative sizes: those to be reduced to zero, those to be reduced linearly, those to be subject to a Swiss formula and those for a few highly protected commodities that were to be allowed to remain above some maximum level, but only if additional market access were negotiated through TRQs or some other mechanism. The US-EU proposal acknowledged the principle of special and differential treatment for developing countries, allowing lower tariff reductions and longer implementation periods, a new special safeguard for import-sensitive commodities, and promising duty-free access for ‘at least [] per cent of imports’ from those countries (again unspecified).

US Agricultural Policy

97

Whereas, in the Uruguay Round, the Cairns Group would have voiced objection to a proposal along the US–EU lines, in the Doha Round it was the new block of developing countries, led by Brazil, India, China and South Africa, which orchestrated dissent. The ‘G-20’ proposal put forward as an alternative in Cancún called for product-specific reductions in AMS Amber Box support, elimination of the Blue Box, a cap on direct income support payments in the Green Box and cuts to de minimis support expenditures. It called for reduction of all tariffs of developed countries to some maximum level, expansion of market access under TRQs, an end to special safeguards for developed countries and elimination of all export subsidies. Several of the strongest of these proposals were similar to those put forward by the USA in its initial proposal, but subsequently abandoned. The G-20 also called for continuation of special and differential treatment and a new special safeguard for developing countries. The collapse of the Cancún conference left it uncertain how far apart the negotiations on agriculture remained. One of the most visible conflicts in Cancún about agricultural trade took place not over the general provisions of the negotiating framework, but over a specific proposal by four low-income African countries (Benin, Burkina Faso, Chad and Mali) to eliminate subsidies and protection for cotton, one of their principal export crops. Building on the concept of products ‘of interest to developing countries’, these African nations made the argument that cotton was of critical importance to their development and to the livelihood of two million poor farmers, and thus should be subject to liberalization and elimination of subsidies, even in advance of other agricultural trade reform. Until this was accomplished for cotton, the African countries argued that a mechanism should be arranged to provide cash compensation to them for the damage done to their export incomes. A separate case by Brazil against US cotton subsidies was also under review in the WTO dispute settlement process.9 The African cotton case was particularly poignant in 2003 because world prices over the previous few years had been below productionstimulating (Amber Box) US loan rates, so the most egregious form of pricesupporting subsidies in a developed country had been in force. But the structure of the US cotton programme, with both guaranteed producer prices and direct and counter-cyclical payments somewhat decoupled from production, differs little from the domestic programmes for other supported crops. Barring US trade negotiators being authorized to take on a contentious internal political battle, they would not be able to accede to such specific liberalization for one crop. The USA is likewise unwilling to open the door to consideration of explicit international compensation payments for damage done by subsidy policies. Not only would international transfers be involved, but compensation for policies that continued in effect differs from the compensation accompanying changes in policy instruments, as has characterized the cash-out of domestic US farm support programmes.

98

D. Orden

The July 2004 framework agreement on agriculture The divergent positions that prevented agreement in Cancún continue to demonstrate how difficult it is to achieve multilateral trade liberalization and subsidy reduction for agriculture. In subsequent months negotiations intensified, with the goal of achieving the planned framework agreement by July 2004, before the US presidential elections. In these efforts, both the USA and the EU are constrained by their domestic farm programme commitments and other developed countries (Japan and other members of a group known as the G-10, for example) retain even higher farm support levels. The USA has long been committed to the concept that farm support can be divided between trade-distorting policy instruments and those that provide income transfers without distorting trade. Even the US zero option proposal in the Uruguay Round addressed only elimination of tradedistorting policies, leaving room for unlimited expenditures that were arguably decoupled from production. The WTO panel ruling in the case of Brazil against US cotton subsidies notwithstanding, the USA is not prepared to have its latitude for making domestic support payments constrained very much by a Doha agreement. But in January 2004, the USA returned in part to its earlier position, arguing in particular for a firm date for ending all export subsidies. The EU has moved more slowly than the USA away from subsidy payments directly tied to farm production and toward income support with less explicit linkages to output. EU partial reforms of the CAP in this direction in July 2003 (see Chapter 2) gave it additional flexibility in its WTO positions, but not enough to accede easily to full elimination of export subsidies. Under pressure to do so, the EU called for export subsidy elimination to be accompanied by strong disciplines on other export assistance, including credit programmes/guarantees and food aid (both used primarily by the USA), and activities of exporting state trading enterprises. In the area of tariffs and market access, there remain difficult conflicts between ambitious goals for market opening and continued provision of high levels of protection to various ‘special’ or ‘sensitive’ products as designated by importing countries. A multi-tier approach to tariff reductions leaves latitude for substantial tariff differences among commodities and for numerous cases in which any tariff reductions agreed upon do very little to open market access. On top of this, the developing countries, through groups known as the G-33 and G-90, are pressing the case for their own special and differential treatment, including new special safeguards and limited tariff reductions. The net result of these proposals may well turn out looking much more like Swiss cheese – disciplines full of holes for high protection – than like the liberalized regime that would result from application of a tariff-cutting Swiss formula. There is a divergence of views here between advocates of liberalization of agricultural trade and many developing countries and development advocates. Trade liberalization proponents are wary of tariff exceptions, including special

US Agricultural Policy

99

and differential treatment, that leave high levels of trade protection in place. Yet many development advocates and governments of developing countries are adamant that they be granted room to retain trade barriers. By July 2004, after nearly 3 years of Doha negotiations, it proved possible only to reach a very broad WTO framework agreement, leaving many agricultural issues unresolved. The agreement reached on 30 July (Annex A for agriculture) specifies that ‘substantial improvements in market access will be achieved for all products’, but also that there will be ‘flexibilities for sensitive products’. No specifics about the structure of commitments to cut tariffs were agreed upon, except an indication that cuts will be from bound rates, that a tiered approach will be used and that some expansion will be required for TRQs. Developing countries are to receive special and differential treatment on market access commitments in terms of lesser cuts, the ability to designate some commodities as new ‘special products’, and establishment of a new special safeguard mechanism. LDCs are not required to undertake reduction commitments, but cotton subsidy and market access issues that had been highlighted will be addressed ‘ambitiously, expeditiously and specifically’ only within the agricultural negotiations, not in a separate initiative as some of the LDCs had sought. The draft framework agreement calls for elimination ‘by a credible end date’ of export subsidies and ‘parallel’ elimination of ‘all export measures with equivalent effect’. The agreement calls for elimination of food aid that is not ‘in conformity with operationally effective disciplines to be agreed’, with the objective to ‘prevent commercial displacement’. The July, 2004 framework agreement calls for an overall reduction in the sum of ‘all trade-distorting domestic support’, defined to include total AMS, plus allowed de minimus, plus the Blue Box, as well as for specific reduction commitments on the three separate components. Total AMS subsidies are to be reduced, with members with the highest levels making ‘greater reductions’. Product-specific AMSs are to be capped according to ‘a methodology to be agreed’ and reductions in total AMS are to ‘result in reductions of some product-specific support’, whereas in the URAA it is only the total AMS that is bound and reduced. Allowed de minimis is to be reduced by a percentage ‘to be agreed’ in further negotiations. The Blue Box is to be limited not to exceed 5 per cent of the total value of agricultural production of a country during a historical period to be established in the negotiations. The definition of the Blue Box is extended to accommodate new payments in the 2002 US farm bill, by allowing into the Blue Box farm support payments ‘that do not require production’, as well as the payments originally included when made ‘under production-limiting programmes’. No constraints were placed on Green Box expenditures, but the eligibility criteria are to be reviewed for consistency with the included measures having ‘at most minimally trade-distorting effect’. The July 2004 framework agreement provides some limited structure to Doha Round negotiations to follow in 2005, and probably for longer. The framework leaves many key decisions to be made that would determine how ambitious the final Doha Round agriculture agreement will be in

100

D. Orden

opening market access and reducing subsidies. Failure to have agreed to more specificity among such diverse groups as the USA, EU, G-10, G-20, G-33 and G-90 after nearly 3 years of negotiations suggests only limited trade liberalization will emerge from the Doha Round. Inclusion of new protection mechanisms for developing countries reinforces this perception.

Conclusions The 2002 US farm bill has been widely criticized for increasing subsidies with detrimental effects on competing agricultural producers abroad and for undermining US leadership in achieving liberalized world agricultural trade. A careful assessment shows that the 2002 FSRIA has effects that are nuanced in at least four respects: it raises expenditures compared to 1996 legislation, but not compared to actual 1998–2001 outlays; it maintains planting flexibility, but extends support to new crops and undermines some of the decoupling of subsidy payments from production and market prices that had occurred; it violates the spirit of US trade liberalization rhetoric, but probably not the letter of US WTO commitments; and it continues the policies of wealthy countries that collectively distort agricultural production and world prices, but only marginally worsen the net effects of those policies. That said, one of the unfortunate consequences of the 2002 farm bill has been a retreat by the USA from a strong reform proposal in the WTO Doha negotiations. The initial July, 2002 US proposal on agriculture called for significant multilateral restraint on subsidies and protection, none of which was undertaken on a unilateral basis in the 2002 US farm bill. This divergence frustrated proponents of further agricultural trade liberalization, who would have preferred sharp unilateral reform action by the USA in 2002 as a clarion call for similar reforms abroad. Movement away from reform in the domestic farm bill was then followed by enunciation by the USA of a common negotiating position with the EU that embodies much less reform than the initial US position. The cohesion between US domestic policy and its international negotiating position does not preclude progress on agriculture as the multilateral negotiations proceed, but it is an obstacle. The Uruguay Round began under circumstances of another expensive US farm bill of 1985. Despite that expensive bill, the USA started with, and for a long time stuck with, a call for substantial multilateral reform. By the conclusion of those negotiations, with limited progress eventually made, the 1985 farm bill could hardly be considered a key obstacle to the Agreement on Agriculture that was reached. It is to be hoped that substantial additional progress is made on agriculture in the Doha Round. The expensive 2002 US farm bill that precedes the international negotiations, while unfortunate, may not prove the limiting determinant of reforms achieved in a new multilateral agreement for agriculture. But this time it will not be the USA that makes the argument for pursuing those reforms most aggressively.

US Agricultural Policy

101

Notes 1

A longer version of this analysis of the 2002 US farm bill can be found in Orden (2003). 2 The Australian Agriculture Minister, Warren Truss, presciently anticipated this outcome when he observed after the 2002 farm bill was enacted that its worst feature might turn out to be ‘the way it limits the capacity of the US to take a strong leadership role in trade negotiations’. 3 Orden, Paarlberg and Roe (1999) describe four alternative approaches that can be pursued to reform farm policy interventions. The four approaches differ in speed of implementation and whether compensation is made to those benefiting from the intervention programmes (a fast, compensated ‘buyout’ or uncompensated ‘cut-out’, versus a slower compensated ‘cash-out’ or uncompensated ‘squeeze-out’). Of the four, they argue that only a slow reform with substantial compensatory payments (the cashout) has proved politically viable in the USA. 4 The term ‘base acreage’ refers to the acreage on which payment eligibility is determined; ‘deficiency payments’ refer to subsidies provided on most but not all of farm output when market prices were below a legislated ‘target price’, and ‘loan rates’ refer to price guarantees for all output of the covered commodities. The 1990 farm bill had provided limited flexibility under which farmers could shift an optional 10 per cent of their base acreage among crops without that land permanently losing payments eligibility, but eligibility for deficiency payments was suspended on that acreage during years when alternative crops were grown. 5 For detailed side-by-side comparison of the 1996 FAIR Act and the 2002 FSRIA see ERS/USDA (2002). An additional summary of the main provisions of the FSRIA related to support programmes, conservation and trade is provided by Westcott et al. (2002). 6 Although the Secretary of Agriculture was authorized to do so, loan rates had not been changed under the FAIR Act. Such formula-based rates would have been lower than the maxima specified in the law after market prices dropped sharply from 1998 to 2001. 7 This is because the payments did not have to count against the Amber Box constraint if they remained under a de minimis limit of 5 per cent of the value of agricultural production. 8 The tight restriction on groundnut production for the domestic edible market had been relaxed only slightly from the original supply control programme of the 1930s to 2001. Until 1996, quota groundnuts had to be grown in the county and state in which the quota had originally been assigned. Under the FAIR Act some quota could be transferred (leased or permanently sold) across county lines within a state. This reform was phased into effect, until a maximum 40 per cent of the state’s quota was eligible to move across county lines. This policy change allowed part of quota groundnut production to move from higher-cost production areas to lower-cost areas. The largest shift of production has occurred in Texas, where nearly all of the allowed quota transfer has occurred from Central Texas to West Texas. Fewer transfers of quota across county lines occurred in other states. 9 In the cotton dispute Brazil challenged numerous aspects of US cotton policies: the fixed, counter-cyclical and loan rate support programmes but also uses of crop insurance, export credits and special payments to processors for use of domestic cotton. A panel ruling made available to the public in September 2004 ruled in favour of Brazil on several counts. The decision was appealed against by the USA, and went to the WTO’s Appellate Body, which found that several sections of the cotton programme and related policy provisions violated WTO rules. In March 2005, the Dispute Settlement Body accepted the panel report with the changes suggested by the Appellate Body.

102

D. Orden

References ERS, USDA (2002) The 2002 Farm Bill: Side by Side Comparison with 1996–2001 Farm Legislation. Available at www.ers.usda.gov House Agriculture Committee (2002) The Facts on US Farm Policy. Accessed September 2002 at www.agriculture.house.gov Orden, D. (2003) US Agricultural Policy: The 2002 Farm Bill and WTO Doha Rounds Proposal. TMD Discussion Paper 109, International Food Policy Research Institute, Washington DC.

Orden, D., Paarlberg, R. and Roe, T. (1999) Policy Reform in American Agriculture: Analysis and Prognosis. The University of Chicago Press, Chicago. Westcott, P.C., Young, C.E. and Price, J.M. (2002) The 2002 Farm Act: Provisions and Implications for Commodity Markets. Agriculture Information Bulletin Number 778, Economic Research Service, US Department of Agriculture. Available at www.ers.usda.gov

5

The Effects of Domestic Agricultural Reforms and Market Access on Trade and Production in Less Developed Countries SHERMAN ROBINSON* AND KAREN THIERFELDER† *Department

of Economics, School of Social Sciences, University of Sussex, Falmer, Brighton BN1 95N, UK; †Economics Department, US Naval Academy, 589 McNair Road, Annapolis, MD 21402, USA

Introduction In this chapter, we analyse the impact of OECD policies, particularly US farm programmes, on the developing countries. We separately consider: (i) the impact of OECD domestic support programmes on world markets through their impact on OECD agricultural exports; and (ii) OECD protection of domestic agricultural markets that affect potential agricultural exports of developing countries. In addition, high agricultural tariffs in developing countries restrict ‘south–south’ trade as well as ‘north–south’ trade. There are potential trade gains from global agricultural liberalization as well as from reform of OECD policies alone. While all agree that agricultural support in OECD countries is large, a contentious issue is the degree to which these programmes affect farmer incentives, production and trade – the extent to which programmes are ‘coupled’ or ‘decoupled’. It should not matter to developing countries how much OECD countries support farmers, but whether that support affects their production and trade. The operation of OECD policies matters more than their size, and in analysing their impact it is important to consider the institutional details of the policy regimes. Furthermore, any domestic support programmes that seek to support prices received by farmers must be accompanied by policies that insulate domestic markets from world markets in order to prevent commodity arbitrage between the markets. Issues of market access are therefore inextricably linked to the nature of domestic support programmes – one cannot argue for removal of border protection independently of reforming domestic price support programmes.1 © CAB International 2006. WTO Negotiations and Agricultural Trade Liberalization: the Effects of Developed Countries’ Policies on Developing Countries (eds E. Diaz-Bonilla et al.)

103

104

S. Robinson and K. Thierfelder

Market access is a complex issue for developing countries because of the emergence in the past 20 years of overlapping networks of regional trade agreements, both north–south and south–south, and various preferential trade arrangements initiated by OECD countries to provide some market access for the poorest countries (e.g. the US General System of Preferences (GSP), the Caribbean Basin Initiative (CBI), the Andean Pact, Canada’s Generalized Preferential Tariffs (GPT) and the EU Lome/Cotonou agreements.2 For developing countries, there is a potential trade-off between having preferential access to protected markets versus global liberalization which provides unlimited access to unprotected markets. A playing field tilted slightly in your favour may be better than a level playing field. While theory can provide some guidance in judging the potential impacts to developing countries of different reform scenarios, empirical analysis is required to untangle the complex threads of linked policies. We analyse these issues using an empirical global trade model, with explicit treatment of OECD domestic agricultural support programmes. The next section provides a summary presentation of the model. We then analyse various reform scenarios and conclude.

Modelling Trade and Agricultural Programmes The framework of analysis is a general equilibrium model with a multiregion and multi-sector specification. The base year is 1997 and most of the data come from the database of the Global Trade Analysis Project (GTAP), version 5 (Dimaranan and McDougall, 2002). The structure of this class of static world CGE model is described in greater detail elsewhere (see for instance Lewis, Robinson and Thierfelder, 2003 and Diao et al., 2002). The model focuses on the real side of the economy, including domestic production, consumption, real income and GDP within each country/region, and international trade flows across countries/regions. There are 13 developing countries or country groups (including transition economies and central European countries) in the model (see Table 5.1)3. There are also six other industrialized countries or regions: the USA (with Canada), the EU (with EFTA), Japan (with Korea), Australia/New Zealand and Other Developed. The developing countries have been divided into two categories: food insecure and other.4 The model includes 15 agricultural and processed food products and the rest of activities in the economy are aggregated into three broad sectors: resources, manufacturing and services. Those products are either sold to the domestic markets or exported to the other countries/regions. On the demand side, there are four different types of demand: consumer demand, government demand investment demand and intermediate demand. All the demands are met either by domestic supply or imports. Imports and exports are differentiated by country of origin and destination, and the model determines all bilateral trade flows. There are five factors of production: skilled labour, unskilled labour, capital, land and natural resources.

Domestic Agricultural Reforms and Trade in Less Developed Countries

105

Table 5.1. Countries and regions in the static world CGE model. Regions from GTAP data USA Canada Mexico Central America and Caribbean Andean Pact Argentina Brazil Chile Rest of South America European Union (EU)

Central Europe Japan Australia and New Zealand China and Hong Kong India Southern Africa

Food-insecure LDCs Other LDCs

High-income Rest of World

Aggregation of GTAP regions/countries

Colombia, Peru, Venezuela, Rest of Andean

Uruguay, Rest of South America Austria, Belgium, Denmark, Finland, France, Germany, UK, Greece, Ireland, Italy, Luxembourg, Netherlands, Portugal, Spain, Sweden, Switzerland Hungary, Poland, Rest of Central European Association

Botswana, Rest of SACU (Namibia and South Africa), Malawi, Mozambique, Tanzania, Zambia, Zimbabwe, Other Southern Africa Philippines, Bangladesh, Sri Lanka, Rest of South Asia Indonesia, Malaysia, Thailand, Vietnam, Former Soviet Union, Turkey, Rest of Middle East, Morocco, Rest of North Africa, Rest of World Korea, Taiwan, Switzerland, Rest of EFTA

The main institutions or economic actors in the model include: a single private household in each country that receives income from its ownership of factors of production, saves a constant proportion of disposable income, pays taxes and buys consumption goods; the government, which spends all its tax revenues on consumption or lump-sum transfers to households; a capital account, which collects savings and buys investment goods; and producers, one representative firm for each sector, which produces output, buys intermediate goods and pays factors of productions. In making production decisions, firms maximize profits given their technology and input costs, selling their output on domestic or foreign markets (exports). Domestically produced goods in a given sector sold on the domestic market are assumed to be imperfect substitutes for goods from different countries. Consumers demand a constant elasticity of substitution (CES) aggregate of domestic goods and imports from all sources. Similarly, firms produce a good which is differentiated by country of destination. The export ‘transformation’ is given by a constant elasticity of transformation (CET) function. Domestic and world markets for goods and services equilibrate through changes in endogenously determined prices. Prices in particular countries

106

S. Robinson and K. Thierfelder

are affected by real exchange rates, different levels of border protection and, if applicable, consumption, production and export subsidies. Factor markets also equilibrate through the interaction of demand, supply and prices. Labor markets differ in developed and developing countries. Following a long tradition in development analysis (see for instance Lewis, 1954), labour markets in developing countries are assumed to be characterized by the existence of significant unemployment and inflexible wages. In developing countries, any shock that would induce a rise in the real wage will, instead, generate an increase in aggregate labour demand. Developed countries are assumed to have fixed aggregate labour supplies, and wages change to clear labour markets. Different studies have shown that trade liberalization may positively affect country productivity through different channels such as: (i) learningby-doing, access to new knowledge and scale effects associated with increased exports; (ii) technological spillovers due to greater availability through imports of better capital and intermediate goods for production; and (iii) increases in competition in previously protected domestic markets due to increased international trade.5 The model includes an endogenously determined total factor productivity (TFP) variable for each sector’s valueadded function. Within each country, sectoral TFP depends on the ratio of total trade (exports plus imports) to GDP. For developing countries, increased trade leads to productivity gains. The US dollar is chosen as the world numeraire for the model, with the US nominal exchange rate fixed at one and world prices expressed in US dollars. Every country has its own numeraire price index, which is defined as the aggregate consumer price index. Country/region trade balances are assumed to be fixed, and the real exchange rate in each country/region adjusts endogenously, changing the ratio of domestic prices of traded and non-traded goods to achieve levels of aggregate exports and imports to equilibrate the specified trade balance.6 The model has investment-driven savings and the household savings rate adjusts. There are fixed absorption shares for investment demand and for government demand. The model incorporates most-favoured-nation (MFN) bound and applied agricultural tariffs from the Agriculture and Market Access Database (AMAD).7 These rates include the ad valorem equivalents of tariffrate quotas.8 Non-agricultural tariff data are from GTAP version 5.1. Regional trade agreements and non-reciprocal preferences that reduce MFN tariffs for selected bilateral trade flows have become increasingly important in the global trading system, but the availability of preferential tariff data is uneven. Our model includes US and Canadian agricultural trade non-reciprocal preferences provided to developing countries. We also include NAFTA preferences. Chilean bilateral trade agreements, the BrazilMexico FTA and MERCOSUR are modelled in a stylized way, with zero bilateral tariffs among members. EU preferential tariffs apply to about twothirds of its agricultural imports (Hasha, 2001), but a comprehensive database on its preferences is not available. Table 5.2 shows the import-

Domestic Agricultural Reforms and Trade in Less Developed Countries

107

Table 5.2. Import-weighted, agricultural tariff rates for the USA, EU and Japan, in the CGE model (%). (From Agricultural and Market Access Database (AMAD); Global Trade Analysis Project (GTAP), version 5.1; Harmonized Tariff Schedule of the United States Annotated, 2000.)

Rice Wheat Other grains Horticulture Oilseeds Sugar Other crops Livestock Milk Resources Meats Oils and fats Dairy products Processed rice Processed sugar Other foods Manufacturing Services

USA

EU

Japan

4.1 0.1 0.2 0.9 0.8 0.0 2.0 0.3 0.0 0.3 2.2 2.2 36.9 2.4 16.1 5.5 2.6 0.9

64.9 61.4 38.6 14.5 0.0 251.6 2.7 9.3 0.0 0.1 62.7 11.4 87.7 87.4 76.4 24.7 4.3 0.8

325.0 249.2 20.2 44.9 76.4 0.0 18.7 28.2 0.0
1.1 49.0 6.6 287.0 325.0 116.1 33.9 2.3 2.4

weighted, average agricultural tariffs in the model for the USA, the EU and Japan, inclusive of available bilateral trade preference data. We incorporate ad valorem export subsidy rates for 1998 from USDA (2001) for the USA and the EU (Table 5.3). Other countries that use export subsidies, including Korea, Switzerland and Norway, have been aggregated into large regional groupings in our model, and their subsidies are not accounted for. By 2000, subsidy expenditure outlays and subsidized quantities were below the allowed limits set by the Agreement on Agriculture (AoA), with the exception of US dairy subsidy volume. In 1998, EU sugar subsidies exceeded both value and volume commitments, but in 2000 these subsidies were under 70 per cent of WTO limits.

Agricultural Domestic Support We model the domestic agricultural support programmes of the USA, the EU and Japan using data from the 2000 Producer Support Estimate (PSE) database of the Organization for Economic Cooperation and Development (OECD) (2001). We do not use the market price support component of the PSE, but instead explicitly include agricultural tariffs and export subsidies. We use only the budgetary expenditure component of the PSE. Budget expenditures for the three countries/regions totalled about $165 billion in

108

Table 5.3. Export subsidy rates in the CGE model, and 2000 outlays versus commitments. (From USDA (2001), and US and EU export subsidy notifications to the WTO.) Ad valorem export subsidy rate in CGE model

Actual export subsidy quantity, 2000

Percentage of 2000 volume commitment filled

Percentage

$US million

Percentage

1000 metric tons

Percentage

European Union Rice Wheat Other grains Horticulture Meats Dairy Processed sugar

13.8 9.1 34.2 0.8 27.1 24.2 54.4

35.2 118.1 208.7 29.4 516.6 1112.1 357.2

88 8 18 51 31 44 66

132.3 10203.7 7080.1 738.4 863.9 1581.7 882.2

99 71 65 98 47 77 69

USA Poultry meat Dairy

0.0 18.6

6.8 8.5

47 10

11.5 71.2

41 100

S. Robinson and K. Thierfelder

Actual outlay, 2000

Percentage of 2000 outlay commitment filled

Domestic Agricultural Reforms and Trade in Less Developed Countries

109

2000. About one-third of that amount comprised expenditures on input or output subsidies, or direct payments to producers; the remainder comprised general services expenditures on programmes such as research and development, and on pest and disease control. Following USDA (2001), we decompose PSE budget expenditures into WTO ‘colours’ – the Amber, Green and Blue Boxes that have different treatment under the AoA. We do this by linking farm expenditures in the PSE database to countries’ most recent WTO notifications on their aggregate measures of support (AMS).9 Different countries may notify similar programmes under different WTO colours. For instance, many developing countries notify some input subsidies as Green Box because they are for development purposes, while other countries notify similar programmes as Amber. In the model, the policies have the same economic impact. We include de minimus expenditures, although the WTO AMS notification excludes support that does not exceed 5 per cent of the member’s total value of production (10 per cent for developing countries). For example, we include US irrigation subsidies, which are considered de minimus. See Table 5.4 for a description of how the OECD PSE data are mapped into WTO colours.

Modelling Exogenous and Endogenous Types of Agricultural Programmes We model five types of agricultural programmes. Three are exogenous: per unit output subsidies, household income transfer payments and other minimally distorting payments. Two types are endogenous: variable output subsidies and price support payments. In Table 5.4, we describe how the five policy types operate in the CGE model, and describe the programmes in the USA, the EU and Japan by policy type. Output subsidies are fixed ad valorem subsidies per unit of output. They directly change relative prices and shift resources toward production of the subsidized goods. Since the production technology in the model uses fixed input–output coefficients for intermediate inputs, a subsidy to intermediate goods operates like an output subsidy, and we treat them identically. Fixed, per-unit output or input subsidies represent a relatively small share of subsidy expenditures. In 2000, about one-third of EU subsidies in our database, excluding general services, were output or input subsidies; 9 per cent of US subsidies were in this category, and no expenditures by Japan were modelled in this way. Household income transfer payments are fixed, lump-sum payments that are not linked to the current level of production or to current prices. Income transfers in the model include US production flexibility contracts and EU setaside payments. We link the EU setaside payments to a 10 per cent setaside of grains and oilseeds land area. The effect of income transfers on agricultural production is a controversial issue for policy makers and economists. The growing body of theoretical literature on these decoupled payments has described a number

Variable output subsidy

Input subsidy

Farm household income transfer

Price support payment

Other

Output subsidy, fixed per unit of output

Input subsidy, fixed per Variable unit of input per unit of output

Direct payment to household, fixed

Price subsidy, endogenous, fixed producer price

Minimally distorting

WTO treatment

Amber

Blue

Amber

Blue

Green

Amber

Export subsidy

Green

EU programmes

Production aid and subsidies to crops and livestock, interest subsidies

Production aid for peas, beans

Production aid and payments per ha for crops, production aid for livestock, fodder, silage

Compensatory payments and livestock premiums

Transitional Setaside payments to payments Sweden, payments for cessations, conservation, income aid

Intervention prices and export subsidies

Land management, organic farming, pest and disease control, conservation, farm improvement, agro-tourism, sub-national payments

Japan programmes

Interest and insurance subsidies

US programmes

Green

Blue

Amber

Price stabilization for eggs, horticulture

Crop insurance

Input credit subsidy

PFC

Credit on Market loss assets, assistance income tax concessions

Blue

Price stabilization with supply controls for rice, soybeans, milk

Loan deficiency payments and marketing loan gains

Extension, pest and disease control, infrastructure, rice diversion/ environmental payments Extension, conservation, irrigation, farmland protection, crop disaster

S. Robinson and K. Thierfelder

Output subsidy Policy type in CGE model

110

Table 5.4. Mapping agricultural subsidies from OECD PSE data into WTO ‘colours’ and CGE model. (From Authors’ calculations using approach described in USDA (2000), data from OECD (2001), country notifications on domestic support to the WTO, Japan (2000), WalterJørgensen and Jensen (2001), Hasha (1999), Dyck (2000)).

Programme costs in CGE model

Subsidy costs increase (decrease) when output increases (decreases)

Increase market Income transfers to household, with returns; per unit no links to production decisions or return varies prices with output

Subsidy costs fixed

Provides price floor for producers; when market prices are below floor, producers are insulated from market prices; when market prices rise above floor, farmers respond to market prices

Provides No production or domestic price effects intervention price floor for producers/ consumers; export subsidies maintain intervention price if world price falls below it

Subsidy costs increase (decrease) when market prices fall (rise), if market price is below floor price

Subsidy costs Subsidy costs increase fixed (decrease) when world prices fall (rise) relative to intervention price

Domestic Agricultural Reforms and Trade in Less Developed Countries

Production Increase marginal returns by raising revenue or effects in CGE lowering input costs, resulting in increased model output; increased supply reduces market prices, offsetting some of the benefit

111

112

S. Robinson and K. Thierfelder

of market features that could lead households to respond to transfers in ways that have direct effects on farm production. Rude (2000) describes the roles of risk aversion, increasing returns to scale and debt constraints in creating links between transfers and production decisions. Vercammen (2001, 2003) describes several potential links: rural labour market rigidities, bequest motives, a rising marginal tax rate, a wedge between borrowing and savings rates, and initial debt:asset ratios. OECD (2001) describes the possible insurance and wealth effects of decoupled payments, and their possible effects on exit, expectations and responsiveness. Decoupled payments can also lead to a wealth effect on the rural labour supply (Findeis, 2002). The emerging empirical literature relevant to decoupled income transfers in the USA suggests that they are unlikely to have significant production effects given the efficiency of US agricultural markets for land, labour and capital, and for managing risk (Gardner, 1992; Collender, 1999; Harwood et al., 1999). An analysis of the US experience with decoupled payments under the 1996 farm legislation found no evidence at the aggregate level that receipt of income transfers affected investment rates in machinery and equipment, although some households are probably credit constrained (USDA, 2003). However, the payments may affect production indirectly, by changing household choices about work, consumption, savings and investment. USDA (2003) found that farm households that received the payments had higher rates of consumption out of income than non-recipient farm households of comparable incomes. Dewbre and Mishra (2002) found that US households receiving decoupled payments reduced their on-farm hours, consistent with the expectation that changes in the household’s wealth would lead to adjustments in their labour–leisure choices. However, the effect is small. Roe et al. (2002) showed that if US agricultural capital markets are complete, direct payments have long-term effects on land asset values and rental rates, but no effect on agricultural production. Lin (2003) reports that US counter-cyclical payments, which are a hybrid programme in which payments on a fixed historical base area move inversely with market prices, have generated a ‘swap’ market. Financial institutions purchase the farmer’s right to the payment, absorbing the risk of payment reduction or loss. Alternatively, some farmers are ensuring a level of CCP benefit by participating in futures markets. Anecdotes such as this reinforce the view that US markets for rural capital and risk management are efficient, and that links between household income transfers and production decisions are likely to be weak.10 In our CGE model, we assume that factor markets in developed countries are efficient, so that income transfers have no direct incentive effects on production. Our model structure does not allow us to address other issues related to income transfers: the model has a single aggregate household; we do not incorporate a labour–leisure choice in the model; and we do not model intertemporal savings and consumption choices.11 These are areas of model development that are likely to become increasingly important for agricultural policy analysis.

Domestic Agricultural Reforms and Trade in Less Developed Countries

113

Other minimally distorting payments are fixed budget expenditures that have minimal links to production decisions and which satisfy WTO Green Box criteria. These programmes include payments for land conservation and management, pest and disease control, infrastructure and some disaster assistance. We account for these fiscal expenditures but assume they have no production effects. We allow for endogenous farm programmes, where applicable. Our approach follows a literature of endogenous policies in CGE models that began with a single-country model of US farm programmes by Kilkenny and Robinson (1990) and Kilkenny (1991). Burfisher et al. (1992) extended that analysis to a two-country CGE model with farm programmes in the USA and Mexico. Weyerbrock (2001) modelled intervention prices and endogenous export subsidies in a multi-country CGE model of the EU with farm programmes. In our CGE model, we consider the EU compensatory payments (Blue Box) to be a variable output subsidy. These subsidies are paid directly to producers of grains, oilseeds, protein crops and some animals. Introduced in 1993, the compensatory payments were designed to compensate EU producers for declining guaranteed producer price levels (Hasha, 2001; Walter-Jørgensen and Jensen, 2001). They influence production decisions because payment eligibility requires current production of supported products. Since the total payment is fixed, the rate per unit of output contracts (expands) when output expands (contracts). We allow endogenous price support payments to insulate producers from falling output prices by providing guaranteed floor prices. The domestic policies in the USA, the EU and Japan offer different levels of insulation, and face different AoA disciplines. The US marketing loan rates are an Amber Box policy that provides payments to farmers that increase as market prices decline below the loan rates. Japan’s income stabilization programme provides guaranteed producer prices for rice and other commodities (Japan, 2000). Their deficiency payments adjust with changes in market prices. Japan’s rice programme requires a reduction of rice acreage, which can be diverted into other uses and is eligible for other commodity programme benefits. Japan’s rice programme is a Blue Box expenditure, while other income stabilization payments have been notified as Amber Box support. In the EU, export subsidies are used to clear excess domestic supplies resulting from the EU’s fixed intervention prices for grain, oilseeds, livestock, dairy and processed sugar. An implication of modelling programmes endogenously is that there can be strong interaction effects when simulating the separate removal of each type of agricultural policy – tariffs, domestic support and export subsidies. In our model, removing tariffs alone while price-support programmes remain in place can lead to larger world price swings, since demand will increase but supply response within certain price ranges may be limited. Similarly, the removal of export subsidies in our model would imply the dismantling of some domestic price supports. The

114

S. Robinson and K. Thierfelder

separate reforms are not additive. Nevertheless, many analyses of agricultural policy reform describe policies as fixed ad valorem price wedges, decomposing their separate effects and generally finding that tariffs are the most distorting policy (e.g. Diao et al., 2001; Hoekman et al., 2002; Tokarick, 2003). We argue that our model captures these policies more realistically, by depicting them as linked elements of an integrated support system.

Agricultural Policy Reform Scenarios To describe the effects of OECD domestic and trade policies on global markets in general, and on developing countries in particular, we consider four scenarios. First, we consider the effects of unilateral reform in the USA – this eliminates all domestic agricultural support and liberalizes agricultural trade. Then we consider the effects of domestic support alone by removing domestic agricultural support in OECD countries. Third, OECD countries remove domestic support for agriculture and liberalize trade in agriculture. Finally, we consider complete liberalization in agriculture: we eliminate all OECD domestic support and all countries liberalize trade in agriculture.

Results To evaluate the concern that domestic support in developed countries distorts world markets and trade opportunities for developing countries, we consider changes in trade when domestic support is removed. The distorting effect of domestic support depends on the extent to which programmes affect incentives. We report results for a range of possible effects. In the most coupled case (CCC in DIRPAY) we treat US countercyclical payments as an income transfer. Next, we allow the counter-cyclical payments to operate like production subsidies which do affect incentives (CCC in DEFPAY). In the most extreme case, we assume all income transfers affect production incentives (all DIRPAY coupled). We find that agricultural exports from developing countries (e.g. the regions Southern Africa, Food-insecure LDCs and Other LDCs) expand, but by a small percentage over initial exports, with the largest export gain for other LDCs, whose exports increased by approximately 5 per cent (see Fig. 5.1). The USA and the EU experience a decline in agricultural exports when domestic support is removed. When the USA unilaterally reforms, the export gains to developing countries are smaller. Exports from other nonreforming developed countries also expand as they gain market share from the USA. Market access, in addition to domestic reform in developed countries, causes developing country exports to expand much further, with exports increased by 28 per cent for other LDCs and 30 per cent for food-insecure

Domestic Agricultural Reforms and Trade in Less Developed Countries

115

Other LDC

Countries/regions

FILDC

SAFR

JAPAN

EU

USA

–20

–10 US unilateral ag. reform

0

10 20 Percentage change

OECD domestic reform

OECD dom. ref. + OECD ag. tlib.

30

40

50

OECD dom. ref. + all ag. tlib.

Fig. 5.1. Percentage change in agricultural export volumes.

LDCs. For Southern Africa, the export gain is not as large, but the magnitude is the same – exports expand much more when developed countries liberalize agricultural trade, in addition to domestic reforms. Our results suggest that market access is more important to developing countries than developed country support for agriculture. We also find that developing countries benefit from liberalization in other developing countries. Agricultural exports expand further when all countries liberalize. The larger export gains when all developed countries liberalize reflect the initial trade dependence: Southern Africa, food-insecure LDCs and Other LDCs send 69.2, 67.4 and 63.1 per cent of exports, respectively, to the USA, the EU and Japan. We find that unilateral reforms by the USA reduce agricultural imports in all other regions, with the biggest percentage decline in developing countries (see Fig. 5.2). This suggests that developing countries benefit from domestic support policies that reduce the price of agricultural products on world markets. The same result holds when all OECD countries, rather than the USA alone, liberalized domestic and trade policies in agriculture. Changes in trade volumes are consistent with changes in the world price of agriculture (Table 5.5). When all developed countries liberalize agriculture, the world price increases by 1.8 per cent – supply contracts when domestic support is removed.12 In the most extreme case, in which developed countries remove domestic support for agriculture and all countries liberalize trade, the combined effect of an increase in demand for agricultural imports and a decline in supply due to the elimination of domestic support results in a price increase of 7.1 per cent.

116

S. Robinson and K. Thierfelder

Other LDC

Countries/regions

FILDC SAFR JAPAN EU USA

–10

–5

0

5

10

15

20

25

30

35

40

Percentage change US unilateral ag. reform

OECD domestic reform

OECD dom. ref. + OECD ag. tlib.

OECD dom. ref. + all ag. tlib.

Fig. 5.2. Percentage change in agricultural import volumes.

Table 5.5. Percentage change in the world price of agriculture. CCC in DIRPAY

CCC in DEFPAY

All DIRPAY coupled

No LDC closure Domestic reform Dometic reform and DCs remove ag tm Domestic reform and all remove ag tm

1.543 6.321 8.822

1.976 6.768 9.294

2.705 7.524 10.087

LDC closure Domestic reform Dometic reform and DCs remove ag tm US unilateral reform Domestic reform and all remove ag tm

0.76 3.768 1.659 5.975

1.192 4.216 na 6.451

1.754 4.81 na 7.083

Note: for US unilateral reform, we considered only LDC closure and the scenario in which programmes are most coupled, CCC in DIRPAY.

The change in aggregate real absorption (see Table 5.6) indicates the effect of a policy change on welfare. We find that domestic reform, whether by the USA alone, or for all OECD countries, has a small, positive welfare effect on developing countries, with real absorption increasing from 0.09 to 0.57 per cent. The gains are much larger when the OECD countries also liberalize trade in agriculture, when they range from 2.0 to 4.7 per cent. When developed countries also liberalize trade in agriculture, the gains expand further, ranging from 2.8 to 5.6 per cent.

Domestic Agricultural Reforms and Trade in Less Developed Countries

117

Table 5.6. Percentage change in real absorption. OECD domestic reform in agriculture Southern Africa Food-insecure LDCs Other LDCs USA EU Japan

0.52 1.04 0.40 0.07 0.16
0.04

OECD domestic and trade reform in agriculture 2.90 4.71 2.00 0.15 0.18
0.23

US unilateral reform in agriculture

OECD domestic reform and global trade reform in agriculture

0.30 0.57 0.09 0.01 0.03
0.03

3.66 5.63 2.75 0.24 0.33
0.22

Note: we report results for LDC closure and the coupled version of US programmes (this gives the upper bound on the effects of agricultural programmes in the world market).

Conclusions Agricultural support policies in the OECD countries consist of different mixes of direct income support to farmers, input and output subsidies, price support programmes, supply restrictions and border protection (trade policy). Many of these programmes distort domestic and world markets, hurting both agricultural exporters in developing countries who are denied access to OECD markets and local farmers producing crops in competition with subsidized OECD exports. Within the OECD countries, the mix of policies is mutually supporting. For example, policies to restrict supply and imports are needed to limit the budgetary cost of output subsidies and price support programmes. It is impossible to change individual components of these agricultural support programmes in isolation – they must be tackled in a comprehensive effort. Such comprehensive reform would be greatly facilitated if OECD countries moved toward a system of decoupled income support for farmers, which would meet the needs of developing countries for market access and for protection from dumping in world markets. There is an active debate about the size of the impact of OECD agricultural support policies on world markets in general, and on developing countries in particular. We examine these questions using a multi-country, multi-commodity model of world trade that captures the complex structure of agricultural support programmes in the OECD countries. We analyse a variety of policy reform scenarios to determine what are the most empirically important elements of the OECD agricultural support programmes from the perspective of the developing countries. We consider separately the impact on developing countries that are ‘food insecure’ (poor, food-importing countries, with a large sector of economic activity in agriculture), who care about world food markets; and on agricultural exporters, who care about access to OECD markets.

118

S. Robinson and K. Thierfelder

The empirical work leads to a number of conclusions: ●

●

●

●

Eliminating only domestic programmes in OECD countries, without any change in trade policies, has little impact on trade with developing countries. Even assuming that virtually all programme costs in the OECD countries are distorting (‘coupled’ to production decisions), the impact is small – world agricultural prices increase slightly, agricultural imports in developing countries fall by 4 to 6 per cent, and their exports expand by 5 per cent or less. Adding agricultural trade liberalization in the OECD countries to domestic policy liberalization in agriculture has considerable effects on trade. Agricultural exports from the developing countries rise dramatically, while their agricultural imports decline modestly. From the perspective of developing countries, access to OECD markets is very important – trade policy is more important to them than OECD domestic support policies, although the two types of policy are linked. Trade among developing countries in agriculture faces more protection than trade between developing countries and the OECD countries, but the trade volumes are lower. Adding trade liberalization among the developing countries to OECD policy reform increases agricultural exports from developing countries by another 50 per cent. In other words, in a scenario of comprehensive agricultural policy reform in both OECD and developing countries, two-thirds of the gain in exports for the developing countries is due to OECD reform, while one-third is due to the removal of barriers against trade among the developing countries. Unilateral agricultural policy reform by either the USA or the EU is difficult, with the non-reforming region gaining market share and serious impacts on farm incomes in the reforming region. As evidenced during the Uruguay Round of GATT negotiations, there is an important role for global negotiations. All countries gain from a mutually supporting programme of global reform.

In sum, the developing countries have an important stake in facilitating the WTO Doha Development Round. It is impossible to negotiate agreements on market access in agriculture in bilateral negotiations, given the interdependence of domestic agricultural support and trade policies within and between the OECD countries. Yet the developing countries have an enormous interest in increasing access to OECD agricultural markets. Market access is much more important to the developing countries than domestic support programmes within the OECD countries, but the two are linked and the developing countries cannot gain better access unless the OECD countries achieve coordinated reform of their agricultural support programmes.

Notes 1

See Burfisher et al. (2002) for a discussion on links between domestic and border policies in the context of NAFTA.

Domestic Agricultural Reforms and Trade in Less Developed Countries

119

2

See Burfisher and Jones (1998), Burfisher et al. (2003) and Schiff and Winters (2003). The country/region and sectoral aggregations are given in Table 5.1. Other data on the economic structure and policy settings for different countries and regions can be requested from the authors. 4 The definition of ‘food insecure-countries’ is given in Diaz-Bonilla et al. (2000) and is based on a variety of measures; see also Chapter 8. 5 See the discussion of the links between trade, technology and productivity in Balassa (1989), Grossman and Helpman (1995) and Romer (1994). For CGE applications with productivity linked to trade see, for instance, de Melo and Robinson (1992), Burfisher et al. (2004) and Diaz-Bonilla et al. (2003). 6 See Devarajan et al. (1993) and Robinson (1991) for a discussion of the real exchange rate in trade-focused CGE models. 7 Many developing countries have agricultural bound rates that are substantially higher than their applied rates. In our trade liberalization scenarios we remove applied rates, since these are assumed to be binding on imports. 8 We do not explicitly account for Japan’s import mark-ups, which are a charge to domestic consumers for purchases from government importers, who import duty-free. Instead, we applied AMAD tariff equivalents to all imports. 9 The WTO AMS notifications of the five countries used for this component of the database were from these years: Canada (1998), USA (1998), EU (1998/1999), Mexico (1998) and Japan (1998). 10 Market inefficiencies in developing countries may create stronger links between income transfers and farm production. For example, decoupled PROCAMPO agricultural income transfers in Mexico were found to lead to increased farm income, which was attributed to their role in relieving farmer credit constraints and stimulating farm investment and output (Cord and Woton, 1998, unpublished; Sadoulet et al., 2001). 11 See Roe et al. (2002), for an analysis of the effects of US decoupled payments on farm consumption and savings decisions, using an intertemporal, economy-wide general equilibrium model. 12 We consider the sensitivity of this result due to the specification of farm programmes. The value reported here is the most distorting version of the programmes in which all US direct payments are treated as being coupled, this gives an upper bound on the effect US domestic support programmes have on world market prices. We also assume LDC closure in which developing countries have trade-productivity links and can expand employment of unskilled labour without increasing the wage to unskilled labour. 3

References Agricultural and Market Access Database (AMAD) (2000) Global Trade Analysis Project (GTAP) version 5.1, Harmonized Tariff Schedule of the United States. Balassa, B. (1989) New Directions in the World Economy. McMillan, London. Burfisher, M.E. and Jones, E. (eds) (1998) Regional Trade Agreements and US Agriculture. Agricultural Economics Report No. 771. US Department of Agriculture, Washington DC.

Burfisher, M.E., Robinson, S. and Thierfelder, K. (1992) Agricultural and food policies in a US–Mexico Free Trade Area. North American Journal of Economics and Finance 3(2), 117–140. Burfisher, M., Robinson, S. and Thierfelder, K. (2002) Developing countries and the gains from regionalism: links between trade and farm policy reforms in Mexico. American Journal of Agricultural Economics 84(3), 36–748.

120 Burfisher, M.E., Robinson, S. and Thierfelder, K. (2003) Trade and welfare effects of the FTAA. In: Burfisher, M. (ed.) US Agriculture and the Free Trade Area of the Americas; Agriculture Economic Report No. 827. US Department of Agriculture, Washington DC. Burfisher, M., Robinson, S. and Thierfelder, K. (2004) In: Anania, G. (ed.) Regionalism: Old and New Theory and Practice, in Agricultural Policy Reform and the WTO: Where are We Heading? Edward Elgar Press, New York. Collender, R. (1999) Agricultural boom and bust: will history repeat in the 1990s? Agriculture Outlook, 22–26. De Melo, J. and Robinson, S. (1992) Productivity and externalities: models of export-led growth. Journal of Economic Trade and Economic Development 1(1), 41–68. Devarajan, S., Lewis, J. and Robinson, S. (1993) External shocks, purchasing power parity, and the equilibrium real exchange rate. World Bank Economic Review 7(1), 45–63. Dewbre, J. and Mishra, A. (2002) Farm Household Incomes and Government Payments. Paper presented to the American Agricultural Economics Association, Long Beach, California. Diao, X., Somwaru, A. and Roe, T. (2001) A Global Analysis of Agricultural Reform in WTO Member Countries, in US Department of Agriculture. Agricultural Policy Reform in the WTO: The Road Ahead. Agricultural Economic Report No. 802, pp. 25–40. Diao, X., Diaz-Bonilla, E. and Robinson, S. (2002) Scenarios for Trade Integration in the Americas. Trade and Macroeconomics Division. Discussion paper No. 90, International Food Policy Research Institute, Washington DC. Diaz-Bonilla, E., Robinson, S., Thomas, M. and Cattaneo, A. (2000) Food Security and Trade Negotiations in the World Trade Organization: A Cluster Analysis of Country Groups. Trade and macroeconomics discussion paper No. 59, International Food Policy Research Institute, Washington DC.

S. Robinson and K. Thierfelder Diaz-Bonilla, E., Diao, X. and Robinson, S. (2003) How Much Does It Hurt? The Impact of Agricultural Trade Policies on Developing Countries. IFPRI Issue Brief, August 2003. http://www.ifpri.org/media/trade20030826. htm Dimaranan, B.V. and McDougall, R.A. (2002) Global Trade, Assistance, and Production. The GTAP 5 Database, Center for Global Trade Analysis, Purdue University, West Lafayette, Indiana. Findeis, J. (2002) Subjective Equilibrium Theory of the Household: Theory Revisited and New Directions. Paper presented to the Workshop on Farm Households-Firms Unit, Wye College, Imperial College, London. Gardner, B.L. (1992) Changing economic perspectives on the farm problem. Journal of Economic Literature 30, 62–101. Grossman, G. and Helpman, E. (1995) Technology and Trade. In: Grossman, C. and Rogoff, K. (eds) Handbook of International Economics, Volume 3. North Holland, Amsterdam. Harwood, J., Heifner, R., Coble, K., Perry, J. and Somwaru, A. (1999) Managing Risk in Farming: Concepts, Research and Analysis. Agriculture Economic Report No. 774, Economic Research Service, US Department of Agriculture, Washington DC. Hasha, G. (2001) EU Preferential arrangements: heightened competition for US. Agriculture Outlook, AGO-287, 16–20. Hoeckman, B., Ng, F. and Olarreaga, M. (2002) Reducing Agricultural Tariffs versus Domestic Support: What’s More Important for Developing Countries? World Bank Policy Research Working Paper No. 2918. Kilkenny, M. (1991) Computable General Equilibrium Modeling of Agricultural Policies: Documentation of the 30-Sector FPGE GAMS Model of the United States. Economic Research Service, Staff Report No. AGES 9125, US Department of Agriculture, Washington DC. Kilkenny, M. and Robinson, S. (1990) Computable general equilibrium analysis of agricultural liberalization: factor mobil-

Domestic Agricultural Reforms and Trade in Less Developed Countries ity and macro closure. Journal of Policy Modeling, 12, 527–556. Lewis, J.D., Robinson, S. and Thierfelder, K. (2003) Free Trade Agreements and the SADC economies. Journal of African Economies 12(2), 156–206. Lewis, W.A. (1954) Economic Development with Unlimited Supplies of Labour. Manchester School of Economic and Social Studies, 22. Organization for Economic Cooperation and Development (2001) Producer Subsidy Estimate Database, 2000. Robinson, S. (1991) Macroeconomics, financial variables, and computable general equilibrium models. World Development 19(11), 1509–25. Roe, T., Somwaru, A. and Diao, X. (2002) Do Direct Payments Have Intertemporal Effects on US Agriculture? International Food Policy Research Institute, Trade and Macroeconomics Division, Discussion Paper No. 104. Romer, P.M. (1994) The origins of endogenous growth. Journal of Economic Perspectives 8(1), 3–22. Rude, J. (2000) Green Box Criteria: A Theoretical Assessment. Publication No. 2007/E, Agriculture and Agri-food Canada, Economic and Policy Analysis Directorate. Sadoulet, S., de Janvry, A. and Davis, B. (2001) Cash Transfer Programs with Income Multipliers: PROCAMPO in Mexico. World Development 29(6), 1043–1056. Schiff, M. and Winters, L.A. (2003) Regional Integration and Development. The World Bank, Washington DC. Tokarick, S. (2003) Measuring the Impacts of

121

Distortions in Agricultural Trade in Partial and General Equilibrium. IMF Working Papers 00/110, The International Monetary Fund, Washington DC. United States Department of Agriculture (2001) Agricultural Policy Reform in the WTO. The Road Ahead. Agricultural Economic Report No. 802, Economic Research Service (USDA/ERS), Washington DC. United States Department of Agriculture (2003) Decoupled Payments: Household Income Transfers in Contemporary US Agriculture. Agricultural Economic Report No. 822, Economic Research Service (USDA/ERS), Washington DC. Vercammen, J. (2001) A Dynamic Analysis of Lump Sum Farm Subsidies. Paper presented at the joint Canadian Agricultural Economics Society/American Agricultural Economics Association annual meetings, Chicago, Illinois. Vercamman, J. (2003) A Stochastic Dynamic Programming Model of Direct Subsidy Payments and Agricultural Investment. Paper presented at the joint Canadian Agricultural Economics Society/American Agricultural Economics Association annual meetings, Montreal, Canada. Walter-Jørgensen, A. and Jensen, T.V. (2001) The CAP and the International Trade Negotiations. Ministeriet for Fodevarer, Landbrug og Fiskeri, Statens Jordbrugs-og Fiskeriokonomiske Institut, Copenhagen. Weyerbrock, S. (2001) East–West European integration: a general equilibrium analysis of alternative agricultural policies. Review of International Economics 9(3), 462–481.

6

Potential Coalitions and Convergence in the Doha Round KIM M. LIND* AND CHRISTIAN BJØRNSKOV† *Danish

Research Institute of Food Economics, Agricultural Policy Research Division, Rolighedsvej 25, 1958 Frederiksberg; †Department of Economics, Aarhus School of Business, Room P613, Prismet, Silkeborgvej 2, DK-8000, Århus C, Denmark

Introduction At the Ministerial Meeting in Doha in November 2001, members of the World Trade Organisation (WTO) decided to launch a new round of multilateral negotiations. All agreed that this round was supposed to be a development round that, in particular, addressed the needs and interests of developing countries. As approximately half of the population in a typical developing country is employed in agriculture, this sector has special significance for these countries, and although the outcome of the Doha meeting is a multifaceted amalgam of issues and interests, the negotiations on agriculture, therefore, form a vital component of the agenda. Above all, developing countries seem particularly interested in gaining improved access to developed countries’ food and agricultural markets and curbing the use of domestic support. Indeed, a number of studies show that such outcomes could prove considerably effective in furthering their development (e.g. World Bank, 2002 and Chapters 5 and 7, this volume). Yet, analyses of the initial negotiation positions of 120 members of the WTO revealed that there was substantial initial disagreement on how the round should proceed, particularly with regard to agricultural issues where key players strongly disagreed (Bjørnskov and Lind, 2002; Panagariya, 2002; Ruffer and Swinbank, 2003). Most notably, agricultural protection and support remains an unfinished agenda in the WTO. An early analysis of the initial positions on these issues (Bjørnskov and Lind, 2002) identified nine distinct groupings in the WTO. A conclusion was that the EU and Israel formed a rather isolated group. Conversely, the African group was relatively closer to a cluster consisting of the USA and most of the Cairns Group. Based on these findings, potential country groups were identified that could lead to the formation of alliances of developing countries in the Doha round, in order for them to gain bargaining power. 122

© CAB International 2006. WTO Negotiations and Agricultural Trade Liberalization: the Effects of Developed Countries’ Policies on Developing Countries (eds E. Diaz-Bonilla et al.)

Potential Coalitions and Convergence in the Doha Round

123

Since that analysis, other negotiating developments took place, including the Harbinson Proposal of early 2003, the WTO Cancún Ministerial in September of the same year and the so-called July Package in 2004. In this chapter, we update the results from Bjørnskov and Lind (2002), looking at which countries could collaborate with each other in the negotiations, and whether the Harbinson draft proposal reflected a genuine compromise between rival positions that ought to have furthered a convergence towards consensus. We provide broadly non-technical answers to these issues; interested readers are referred to the original work in Bjørnskov and Lind (2002) for a description of the more technical issues. The rest of the chapter is structured as follows. The first section describes the data related to members’ positions. The following section answers the question as to whether the Harbinson drafts have been genuine compromises by performing a cluster analysis; the penultimate section discusses problems specific to developing countries, and the final section concludes and draws some tentative policy implications for developing countries.

Positions of WTO Members and the Harbinson Draft This section introduces our measurement approach and tracks the movements in WTO members’ positions from November 2001 to December 2003. Furthermore, China has participated in a joint proposal allowing us to include this country in the analysis. The section also rates Stuart Harbinson’s draft proposals for a compromise in the negotiations.

Negotiation issues By reviewing all available official material during the last 4 years, 14 issues have been identified as being contended in the negotiations.1 The 13 disputed issues identified in Bjørnskov and Lind (2002) can be categorized under four headings: (i) market access issues, covering tariffs, tariff rate quotas, standards and the special safeguard clause; (ii) positions on export support and state trading practices; (iii) domestic support, i.e. the Green, Blue and Amber Boxes, the question of a future Development Box, de minimis levels and the aggregate support measure (AMS); and (iv) various non-trade concerns expressed by members, including positions on a broad vs. a narrow round, labour standards, environmental concerns, geographical indicators and the so-called ‘multifunctionality’ of agriculture.2 Table 6.1 lists the issues forming the foundation for the subsequent analyses. In addition, an extra set of topics has become important since the initiation of the Doha round. These are the so-called Singapore issues, which cover rules for investments, competition policy, transparency in government procurement and trade facilitation.

124

K.M. Lind and C. Bjørnskov

Table 6.1. Negotiation issues in the data set. Market access

Export support

Domestic support

Non-trade concerns

Included in the Tariffs Export subsidies analyses Tariff peaks and Export credits escalation Tariff rate quotas Special safeguard clause

Green Box Broad versus Blue Box narrow round Development Box Labour standards Aggregate measure and environment of support De minimis levels

Excluded from SPS standards the analyses

Amber Box

State trading

Geographical indicators Multifunctionality

The data reflecting the official positions of WTO members on these issues are based on information collected from a number of primary sources: proposals submitted to the agricultural negotiations since 2000, official statements and comments on other members’ proposals during these negotiations, and official statements and declarations during and following the Doha and Cancún meetings. It proved impossible to find sufficient information on 24 members of the WTO. The data therefore include only the positions of the remaining 122 members following the Cancún meeting, as of December 2003. Members’ positions on WTO negotiation issues are mapped into an ordered scale producing a dataset, which is subsequently used for statistical analyses in the next section. The scale consists of the ratings 0, 1, 2, 3 and 4, where the higher the number the more liberal/free trade oriented is the position.3 The rating 0 reflects support for expanding the current provisions for support and protection; a rating of 1 reflects support for keeping the current WTO provisions unchanged, but for including the so-called broad agenda issues, including a number of nontrade concerns; the rating 2 reflects limited support for reducing tariffs and domestic support, but with special and differential treatment or exemptions given to developing countries, in some cases also transition economies; the rating 3 reflects a desire to reduce tariffs or domestic support, i.e. to increase global market access and curb the use of support measures; and a rating of 4 reflects that a country wants to eliminate or substantially reduce tariffs and domestic support, and have a new round of negotiations as narrow as possible. This approach allows us to compare all countries on an issue-by-issue basis as well as on an aggregate basis, since the ratings can be used as a metric to calculate ‘distances’ between the positions of different countries. Here we also track the changes that may have happened in members’ negotiation positions, if those changes have been sufficiently substantial to require an adjustment by one or more ratings of a given country. It is necessary to stress that the present framework cannot capture all elements

Potential Coalitions and Convergence in the Doha Round

125

of the negotiation agenda (for instance, which specific formula is to be used for tariff reductions). However, we believe that it encapsulates the bulk of the information contained in members’ official statements and negotiation positions. In order to provide an example of the measurement approach, we rate the so-called Harbinson proposals4 on all 14 issues, as we have previously done with 122 member countries. Table 6.2 provides an overview of these ratings, showing that Harbinson has not presented a clear position on four issues: the desired broadness of the round, the special safeguard clause, the future of the Green Box and the so-called Singapore issues. In addition, Table 6.2 gives an overview of the ratings on China, which we extracted from the joint Brazilian–Indian–Chinese proposal for the September 2003 Ministerial Meeting in Cancún, Mexico. The table presents the ratings that Harbinson and China receive. It also presents the ratings on the issues averaged over all countries in the WTO that have a position on the particular issue. In the case of tariffs, the average position in the WTO is 2.37, whereas the most common position is 2. On the issue of tariff levels Harbinson calls for a significant reduction, and therefore receives a rating of 3. Likewise, China wants significant reductions in tariffs, and thus receives the same rating on the issue. In general, Harbinson is in line with the typical (median) position in the organization, although he seems more liberal than the average country; the same applies to China.

Table 6.2. Ratings of China and the Harbinson draft proposal. WTO Issue Broad vs narrow round Tariffs Escalation and peaks TRQ size Special safeguard Green Box Blue Box Development Box AMS levels Export subsidies Export credit De minimis levels Postmodern issues Singapore issues Average rating

Harbinson Reduce Reduce Discipline

Reduce In green box Reduce Eliminate Reduce Reduce Opposed

China – 3 3 2 – – 3 2 3 4 3 3 3 – 2.72

Reduce Increase Reduce Reduce Reduce Eliminate

average – 3 – 3 – 3 3 – 3 4 – – – –

2.32 2.37 2.88 2.72 2.18 2.37 3.20 1.49 2.65 3.20 2.30 1.83 2.29 2.27

2.61

2.43

median 2 2 3 3 2 3 4 2 3 4 2 2 2 3

Note. The WTO average is calculated for all countries that have a position on the issue. The median is the most common position on the issue for all countries that have a position. When there is no position, the average value for the WTO is utilized.

126

K.M. Lind and C. Bjørnskov

Movements In order to detect any convergence towards consensus, we here track the changes in negotiation positions that have occurred for the most influential countries during the last 2 years. First, the EU has a distinct position in the negotiations that places the Union at a distance from most other players. Indeed, the EU negotiation position was, and continues to be, isolated in the WTO (Bjørnskov and Lind, 2002). Furthermore, any potential movements in the ratings based on either the comprehensive proposal put forward in the agriculture negotiations or the recent reform of the Common Agricultural Policy are difficult to track, for three reasons. First, EU proposals link different issues by offering a ‘full package’. As such, the Union offers reductions in domestic support, tariffs and export subsidies, but with the caveat that these reductions come together with the inclusion of a range of non-trade concerns, thus broadening the negotiation agenda considerably.5 For example, the recent reform efforts aim to reduce overall domestic support. They do, however, also seek to enhance the scope for supporting various non-trade concerns. It is intended that these new ‘postmodern’ support measures be provided to farmers in the same way as the former traditional domestic support. Many countries therefore suspect that the EU is simply trying to re-label their agricultural support and thereby sidestep the WTO division of support in three Boxes (Swinbank, 2001).6 Nevertheless, as the EU has decided to propose a modest reduction in export subsidies, we adjust their rating on this issue upwards. Secondly, the USA continues to hold that total agricultural support should be reduced, and their official position has not changed in the WTO. However, by approving the Farm Bill in 2002, the USA moved in the opposite direction, mainly by making previous years’ ad hoc crisis aid permanent. We therefore adjust their rating on AMS levels downward compared to the situation in November 2001.7 Thirdly, very few countries have presented new precise proposals. A number of African countries have restated their concerns for ensuring special and differential treatment in the new round and for exempting staple crops from liberalization. Certain developing country members have also restated their support for the introduction of a Development Box. However, none of the official statements have given rise to changes in the ratings of these countries, as practically all have simply confirmed their commitment to specific earlier proposals. Overall, Table 6.3 summarizes the movements of the central players that have induced adjustments in ratings. It thus shows very few substantial changes. We use this set of adjusted ratings for the remainder of the chapter. The singular most important development tracked in Table 6.3 is, however, that of India. As a consequence of the joint proposal of different developing countries – which included India, as well as Brazil, China and others (the so-called Group-20) for the September Ministerial Meeting in Mexico, there have been changes in some specific aspects of the negotiating positions of those countries. Most notably, India now argues for eliminating

Potential Coalitions and Convergence in the Doha Round

127

Table 6.3. Movements in WTO members’ positions. Rating Member

Issue

African group Canada EU

Nov. 2001

Dec. 2003

Restated issues Average

2.57

2.57

No announcements Average

2.74

2.74

0

1

1.48

1.56

0

Export subsidies Post-modern issues Average

India

Green Box Blue Box Average

– 2.23

2 4 2.45

Restated issues Average

1.91

1.91

South Africa

Green Box AMS Average

4 2 3.08

3 3 3.08

USA

AMS levels Average

3 2.98

2 2.91

Japan

all Blue Box support and introducing appropriate disciplines for the use of the Green Box. This development is of potential importance to the convergence of positions in the organization, since India until now has taken a somewhat hesitant position.

Is there convergence? Table 6.3 displays the members and issues for whom we have identified significant changes from the positions in Bjørnskov and Lind (2002). This study analysed countries’ positions as of November 2001. The present study updates the positions to December 2003. The table shows that India received a rating of 0 on Green Box issues in November 2001, but in new proposals India has taken a more liberal standpoint on this issue and now receives a rating of 2. India’s ratings averaged over all issues have also moved in the more liberal direction from an average of 2.23 in 2001 to 2.45 in 2003. Based on the available information, we must therefore conclude that there has been little convergence towards consensus in the Doha round, as most major players have simply reiterated their initial negotiation positions. The central players – the EU and the USA – have held on to their initial positions, and there may even be a slight divergence detectable following the Cancún meeting in September 2003.

128

K.M. Lind and C. Bjørnskov

Identifying Potential Coalitions In this section, 122 WTO member countries that have submitted rateable proposals to the negotiations are subjected to a cluster analysis in order to identify groups of countries with common positions and objectives. Subsequently, the resulting clusters are discussed and their relative positions and distances from one another are studied in order to provide a picture of the potential negotiation coalitions in the WTO. By including the Harbinson proposal, which in this case counts as a ‘country’, we also provide an answer to the question of whether the proposal was a genuine compromise effort.

Cluster analysis Cluster analysis is a mathematical–statistical method for grouping objects with similar characteristics. In this case, the objects are members of the WTO, and the characteristics are their positions on negotiation issues. Cluster analysis thus provides a partitioning of the countries that shows groups that share similar positions – and thus similar objectives – in the negotiations. In comparison to other approaches, such as using ‘expert judging’ to form groups, cluster analysis has the discipline of defining variables and a metric to determine how close or not are the attributes used to form clusters. In addition, all results are, in principle, directly replicable for all interested parties, as our data are publicly available. The analysis in this section is built upon previous work in Bjørnskov and Lind (2002), where further details of the method and the data can be found. The analysis can be described with a simple analogy. Five cities, Copenhagen, Stockholm, Baltimore, Washington and Manila are to be classified into different groups. Cluster analysis is based on the distances between these cities, and would thus place the two Scandinavian cities in one group, the two American cities in another, and the Philippine capital in its own group, as it is a considerable distance from any of the other cities. Applying cluster analysis on the data described above does the same thing, although in 14-dimensional space. The analysis differs from previous work in that it includes the Harbinson proposals, China and the positional movements documented in the previous section, as well as the Singapore issues. In the analysis, all members apparently apply the same weight to all issues, which may a priori seem an unjustified assumption. It could, for example, be expected that two countries with distinct ratings on an issue could nevertheless easily find common ground when one or both countries place little weight on the issue. However, examining the documents underlying our ratings, it is obvious that countries tend to polarize their position on issues of particular interest to them. The ratings per se thus include an implicit weighting scheme, although they may not fully reflect the intensity of disagreement on the issues.

Potential Coalitions and Convergence in the Doha Round

129

This chapter employs two types of complementary cluster analysis methods. Cluster analysis is a multivariate technique for grouping elements according to their characteristics. The purpose of cluster analysis is to create a taxonomy where the elements are grouped together in clusters. These clusters are formed according to the objective that the elements within a cluster are similar to one another with respect to their characteristics, whereas elements in different clusters are dissimilar. Thus, the outcome of a cluster analysis should exhibit within-cluster homogeneity and betweencluster heterogeneity. A wide variety of techniques can be used to group the elements into clusters. The techniques differ according to the various properties of the techniques. One important property is the definition of distance or similarity between clusters, that can take a variety of forms. Furthermore, the techniques differ according to what type of clusters is formed, such as hierarchical, disjoint and/or fuzzy. In the analyses in this study all clusters are disjoint, but both hierarchical and non-hierarchical clusters are formed. A hierarchical cluster procedure is first carried out on the WTO member countries. The particular procedure employed is known as Ward’s method. Ward’s method is an agglomerative hierarchical clustering procedure, where each observation begins in a cluster by itself. Subsequently, the two closest clusters are merged to form a new cluster that replaces the two old clusters. Merging of the two closest clusters is repeated until only one cluster is left. In Ward’s method the distance or similarity between clusters is the sum of the squares between the clusters summed over all the variables. At each stage in the clustering procedure, the within-cluster sum of squares is calculated for all partitions obtainable by combining two clusters from the previous stage. The partition yielding the minimum sum of squares is selected for the next step in the hierarchical clustering procedure. Ward’s method is also denoted as the minimum-variance method. The Ward method has the property of tending to favour clusters that are not too dissimilar with respect to the number of members in the clusters. The hierarchical clustering procedure is irreversible, in the sense that when a country has been clustered together with other countries at some stage, those countries stay together at all subsequent stages. This property of the hierarchical clustering procedure is both an asset and a drawback. The drawback is that it is conceivable that later agglomerations of clusters could change the clusters so much that a country placed in a cluster at an early stage would, at a later stage, be a more natural candidate for another cluster. The pros are that the hierarchical procedure, by its very nature, tells a coherent story and can be a great help in facilitating both a theoretical interpretation of the unfolding order of clusters, as well as the particular partition selected. A set of statistical tests is used to derive number of clusters that provide the best taxonomy of WTO country positions.

130

K.M. Lind and C. Bjørnskov

In order to be able to refine the clustering resulting from employing Ward’s method, a non-hierarchical method is employed. In the nonhierarchical procedure clusters are based on least-squares estimation and Euclidean distances, which is also denoted a k-means method. Clusters are disjoint, meaning that observations are placed in one, and only one, cluster. The number of clusters has to be assigned before the procedure begins, which in this case comes from the hierarchical clustering. Subsequently, each cluster is assigned a seed, which can be derived in a number of ways. Then, as observations are assigned to clusters the seeds are updated as cluster means. An observation is assigned to the cluster with the nearest seed. An advantage of the non-hierarchical procedure compared to the hierarchical is that an observation can change cluster if the updating of clusters leads to another cluster being nearer to the observation. Table 6.4 displays the ten identified clusters containing countries that share similar positions in the negotiations. In the second last column the ratings averaged over all issues and countries in the cluster are presented. The last column provides the share of total WTO members’ population situated in the clusters. Table 6.5 presents additional information on the clusters’ average characteristics. Figure 6.1 illustrates the different average positions of the clusters by splitting them into market access, domestic support, export support and non-trade concerns. Compared with the results in Bjørnskov and Lind (2002) the updated results show both similarities and differences.8 First of all, the ‘average’ cluster identified in the previous study dissolves into several other clusters, which is a natural consequence of the added information acquired from members’ submissions in the period between the former study and the present one, as the former average cluster contained a number of countries with weakly identified overall positions. These countries, mostly transition and middle-income economies, are now grouped together in cluster 7 (‘Few positions’), which therefore becomes a residual cluster. The remaining average countries either join Norway in a new cluster grouped around wanting to enhance the scope for agricultural support, or become grouped in either the ‘Consensus’ cluster that includes Canada – the original average country, Mexico and China. The latter cluster thus covers 34 per cent of the total population in the WTO, or roughly one third of the entire global population. We label this cluster as ‘Consensus’ because it is the one closest to the WTO median (see below, Table 6.6) and because, according to the well-known median-voter theorem (see Mueller, 2003), it is the median rather than the average position that defines the chosen policy in a majority rule. We also note that three clusters remain virtually unchanged. The bulk of the African group, which has put forward joint proposals, are grouped in cluster 2. The Rest of Africa joins either the ‘Liberalization’ cluster or the Cairns group in cluster 6. Two developing countries focusing on the socalled multifunctionality of agriculture are still grouped together in cluster 1, while the European Union is joined by Iceland and Israel in cluster 10, stressing the wish for a broad round that includes various non-trade concerns. Finally, Cuba and India form their own cluster in wanting not

Potential Coalitions and Convergence in the Doha Round

131

Table 6.4. Clusters of WTO members and the Harbinson draft. Cluster

Members

1. DC Barbados, Democratic Republic of Congo multifunctionality

Rating

Population (%)

2.18

1.04

2. Africa

Angola, Benin, Botswana, Burkina Faso, Burundi, 2.56 Cameroon, Central African Republic, Chad, Republic of Congo, Djibouti, Gabon, Gambia, Guinea, Guinea-Bissau, Madagascar, Malawi, Mali, Mauritania, Mauritius, Mozambique, Niger, Rwanda, Senegal, Sierra Leone, Tanzania, Togo, Zambia

4.68

3. Elite development

Dominican Republic, El Salvador, Haiti, Kenya, Nicaragua, Nigeria, Peru, Sri Lanka, Uganda

2.75

5.03

4. Liberalization

Australia, Burma, Brunei, Côte d’Ivoire, Fiji, Ghana, Indonesia, Lesotho, Namibia, New Zealand, Singapore, Turkey, USA

2.75

8.93

5. Special treatment

Cuba, India

2.60

20.33

6. Cairns

Argentina, Bolivia, Brazil, Chile, Colombia, Egypt, Guatemala, Honduras, Malaysia, Pakistan, Paraguay, Philippines, South Africa, Thailand, Zimbabwe

2.97

13.91

7. Few positions

Albania, Antigua and Barbuda, Croatia, Czech Republic, 2.36 Dominica, Estonia, Grenada, Jamaica, Jordan, Kyrgyz Republic, Latvia, Lithuania, Mongolia, Poland, Slovakia, Slovenia, St Kitts and Nevis, St Lucia, St Vincent and the Grenadines, South Korea, Surinam, Switzerland, Trinidad and Tobago, Tunisia

2.94

8. Agricultural support

Bulgaria, Hungary, Japan, Norway

1.87

2.87

9. Consensus

Canada, China, Costa Rica, Ecuador, Harbinson, Mexico, Morocco, Swaziland, Uruguay, Venezuela

2.76

34.06

10. Broad round

Austria, Belgium, Denmark, Finland, France, Germany, 1.70 Greece, Iceland, Ireland, Israel, Italy, Luxembourg, Netherlands, Portugal, Spain, Sweden, UK

6.20

All WTO

2.47

100.00

Note: population gives the percentage of the full population covered by a given cluster. The 122 countries included in the analysis in total count 5.4 billion people. Rating is the average rating on all issues for the countries in the cluster.

only a broad negotiation round with ample room for special and differential treatment of developing countries, but also substantial reductions in agricultural support. With regard to the draft proposals laid forward by Stuart Harbinson, Table 6.4 shows that these drafts are placed in cluster 9 (‘Consensus’). In

132

Table 6.5. Characteristics of the clusters. 1 2 2.66 2.00 2.44 3.00 2.50 2.00 3.60 1.00 2.33 3.60 2.15 1.92 2.29 2.27 2.18 27.0
1.00 1.00 0.00 42.32 9,110 129.7 22.87 2.00

27 2.36 2.07 2.11 2.99 2.90 3.05 3.97 1.68 2.11 3.96 2.19 1.94 2.33 2.09 2.56 9.0
0.13 1.00 0.85 63.75 2,086 66.87 35.85 1.85

3

4

5

9 11 2 2.55 2.92 0.00 2.11 2.75 3.00 3.76 3.53 3.50 3.00 2.86 3.00 2.02 2.70 1.50 2.99 2.36 2.50 3.64 3.32 4.00 0.33 1.69 0.75 3.37 2.48 3.00 3.82 3.77 4.00 3.62 2.00 3.50 1.66 2.35 3.00 2.48 2.20 3.50 1.98 2.47 1.13 2.75 2.75 2.60 29.1 37.2 529.8 0.50
0.03
1.00 1.00 0.69 1.00 0.22 0.15 0.00 54.48 42.59 48.92 2,915 11,389 3,870 79.54 93.03 29.86 17.77 7.90 25.03 1.33 0.69 0.00

6 15 3.42 2.78 3.13 2.87 2.23 2.96 3.75 1.47 3.02 3.80 2.93 2.64 3.03 1.81 2.97 48.3 0.42 1.00 0.00 43.53 5,552 64.43 12.22 0.53

7

8

9

10

23 4 9 17 2.24 1.33 2.22 1.14 2.53 2.59 2.89 2.30 3.08 2.66 2.78 2.95 2.81 2.79 3.00 1.34 1.91 1.89 2.06 1.25 2.38 0.75 2.86 1.28 3.02 1.00 3.82 1.43 1.37 1.25 1.72 1.94 2.75 1.58 2.92 2.84 3.30 2.85 3.91 1.36 2.46 2.23 2.59 2.05 1.51 0.25 3.11 1.20 2.32 1.90 2.34 1.23 2.17 3.07 2.41 2.94 2.36 1.87 2.76 1.70 6.4 37.4 197.2 19.0
0.02 0.25
0.20
0.16 0.50 0.00 0.80 0.00 0.00 0.00 0.00 0.00 41.17 28.54 36.50 21.36 9,054 16,986 8,194 23,284 111.6 96.26 79.03 100.8 11.12 11.54 13.00 0.05 0.75 0.00 0.60 0.06

Note: data on GDP (adjusted for purchasing power), openness, import duty dependence (import duties as per cent of total government revenue), country size and rural population are from World Bank (2003); preferential agreements count the number of agreements with the EU, US, Canada or Japan. We measure political ideology using the categorization by Beck et al. (2001) who define the largest government party at any time between 1975 and 2000 according to whether they have a leftwing, centrist or rightwing political orientation. By coding leftwing parties –1, centrist parties 0, and rightwing parties 1, a crude measure of the ideology of government in any year is obtained. We thereafter average this measure over the period 1995–2001 in which the WTO has been in operation in order to measure average ideology of national administrations.

K.M. Lind and C. Bjørnskov

Countries in cluster Broad vs narrow round Tariffs Escalation and peaks TRQ size Special safeguard Green Box Blue Box Development Box AMS levels Export subsidies Export credit De minimis levels Postmodern issues Singapore issues Average rating Average country size (m) Political ideology DC percentage LDC percentage Rural population GDP per capita Openness Import duty dependence Preferential agreements

2

Potential Coalitions and Convergence in the Doha Round

133

4 3.5

Average rating

3 2.5 2 1.5 1 0.5 0 1

2

3

4

5

6

7

8

9

10

Cluster number Non-trade concerns

Export support

Domestic support

Market access

Fig. 6.1. Cluster positions.

other words, the draft proposal seems to be a genuine effort to reach a compromise that is in accordance with the position of the average WTO member, and thus a balance of special positions and wishes. Although Bjørnskov and Lind (2002) warned that the former average cluster was relatively weakly identified, a set of sensitivity analyses reveals that the placement of the draft in the consensus cluster is correct. However, it should be stressed that although cluster 9 seems slightly more liberal than the average WTO country, it is quite close to the median country. It is worth noting a few features with respect to the remaining clusters, reported in Table 6.5. First, in terms of sheer size, clusters 5 and 9 are prominent by virtue of each containing one of the world’s two largest countries, India and China, respectively. The third largest cluster measured by total population is cluster 6, which includes most of the Cairns group. Secondly, only clusters 8 and 10 include no developing countries while clusters 1, 2, 3, 4 and 6 include no developed countries. There is thus a clear division between most developed and developing countries, which is also reflected in statistics on the average rural population. The African group in cluster 2 consists of 85 per cent of the least developed countries, reflected in the fact that average GDP per capita is 2086 US$ – these are the poorest members of the WTO. Cluster 3 is also rather poor, on average. Some of these differences underlie the distances between the clusters, reported in Table 6.6. By reporting the distances as Euclidean distances in the 14-dimensional space stretched out by the data, the table displays which clusters are most likely to be able to cooperate in the WTO negotiations. Thus, the distance between, say, clusters 2 and 4 is not the same as the distance between clusters 2 and 3 plus the distance between clusters 3 and 4. An interesting finding here

134

Table 6.6. Distances between clusters. Cluster 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.

DC multifunc. Africa Elite development Liberalization Special treatment Cairns Few positions Agricultural support Consensus Broad round

WTO average WTO median Harbinson

1

2

3

4

5

6

7

8

9

10

– 1.51 2.63 1.69 4.13 2.31 1.55 3.96 2.16 4.52

1.51 – 3.04 2.07 4.17 2.39 2.18 4.74 2.05 5.21

2.63 3.04 – 2.76 3.45 2.21 2.17 5.04 2.69 5.28

1.69 2.07 2.76 – 4.23 1.90 1.69 4.33 1.78 4.62

4.13 4.17 3.45 4.23 – 3.80 3.65 5.83 3.27 5.93

2.31 2.39 2.21 1.90 3.80 – 2.21 5.41 1.70 5.53

1.55 2.18 2.17 1.69 3.65 2.21 – 3.46 2.08 3.71

3.96 4.74 5.04 4.33 5.83 5.41 3.46 – 5.05 2.98

2.16 2.05 2.69 1.78 3.27 1.70 2.08 5.05 – 5.02

4.52 5.21 5.28 4.62 5.93 5.53 3.71 2.98 5.02 –

1.21 2.10 2.65

1.82 1.90 2.77

2.39 2.83 2.95

1.43 1.98 2.08

3.77 4.04 3.39

2.10 2.62 1.89

0.60 1.98 2.21

3.64 4.70 4.75

1.84 1.76 1.65

3.76 4.69 4.65 K.M. Lind and C. Bjørnskov

Note: reading from left to right, the bold distance gives the cluster closest to the row cluster. The distances are Euclidean distances. The bottom row reports the distance to the Harbinson proposals.

Potential Coalitions and Convergence in the Doha Round

135

is that the consensus cluster containing Canada, China and Harbinson, is situated surprisingly close to the liberalization cluster, with a distance of only 1.70. These clusters are also close to the Cairns and African group clusters. Countries in these clusters clearly have relatively similar objectives and could probably benefit from combining their bargaining power. The other result to note is that cluster 10 (‘Broad round’) – the EU, Israel and Iceland – remains isolated in the WTO. The closest cluster is the ‘Agricultural support’ cluster, but with a distance of 2.98, this hardly reflects a broad agreement. Finally, we note that although cluster 7 is closest to the average in the organization, this simply reflects that countries in this cluster have very few positions; the cluster is thus average by construction. The cluster with robustly identified positions closest to the average is the African group, while the liberalization and consensus clusters are also relatively close to the average. Although the WTO is based on consensus, i.e. that all countries have to agree, the relevant benchmark with which to evaluate the divergence in the organization is the median position (see before the reference to the median voter theorem: Mueller, 2003). It would, hence, probably be the objectives of this cluster that would prevail in a hypothetical majority vote in the organization. In this respect, it is worth stressing that the cluster closest to the median country is the consensus cluster. Any convergence towards consensus in the future is therefore likely to come together around this specific set of positions.

Harbinson versus central players As the previous section documents, there are only very modest movements towards consensus in the Doha round. Harbinson’s draft proposal is an attempt to further this process and thus speed up the negotiations. However, a number of members have complained that the drafts do not take their specific concerns and positions into consideration, i.e. these members implicitly indicate that the drafts do not present a genuine compromise. We therefore devote this section to answering this question: is Harbinson’s draft proposal a genuine compromise? We do this by entering the set of ratings of the draft in a cluster analysis alongside those of the 122 members that have stated official positions on a sufficient number of issues. If the draft reflects a compromise position, we would expect it to end in the ‘average’ cluster and preferably close to Canada, which was identified as the most average, centrally placed member in Bjørnskov and Lind (2002). As noted above, it is not the case that Harbinson is placed at the average. Yet, the Harbinson proposals are situated close to Canada in the consensus cluster that seems to be defined around the politically relevant median position. The inevitable conclusion is thus that Stuart Harbinson made a genuine effort to present a proposal that would have prevailed if the WTO had had majority rule. In other words, the proposals seem to have hit the middle of the road in the organization. They were nevertheless ineffective in this respect, which warrants a closer look at the reasons for the lack of consensus.

136

K.M. Lind and C. Bjørnskov

To get a better overview of how Harbinson has positioned himself in relation to the central players, Table 6.7 gives correlations and distances to the proposals, as well as the average ratings; distances to the identified clusters are reported in Table 6.6. These distances should be interpreted as relative, which is the reason for including the percentage of the maximum distance. So although many members would find it difficult to agree with the Harbinson proposals, the table documents that some central members are substantially more in line with the draft than others. Table 6.7 in general confirms the main conclusions of Bjørnskov and Lind (2002). Norway, which we include as an illustrative example, as it has the most extreme position in the organization, is placed furthest away from the draft. The country is placed 6.52 from the draft, or 63 per cent of the maximum distance. The table also reveals the isolation of the EU position, since the Union is placed 52 per cent of the maximum distance away from Harbinson. Moreover, the correlation between the EU and Harbinson positions is virtually zero. At the other end, China and Canada are the central players that most agree with the draft, which is also reflected in finding that the distance from the consensus cluster is the smallest between any cluster of countries and the Harbinson proposals. Those countries are placed only 17 and 19 per cent, respectively, of the maximum distance away and have high correlations with the Harbinson position. It is worth noting that they are also substantially closer than the average or median country. India, on the other hand, has a high correlation but has some distance to the proposals, which reflects that the country has a few issues on which it takes a distinctive position. This is also reflected in the cluster analysis, which places India and Cuba in their own group, which is placed 3.39 from the Table 6.7. Correlations with central players.

Country African group Canada China EU India Japan Norway South Africa USA Harbinson Average Median

Average rating

Distance from Harbinson

Percentage of max. distance

Correlation

2.55 2.71 2.83 1.66 2.46 2.06 1.33 3.09 2.86 2.72 2.45 2.47

2.97 1.95 1.78 5.39 3.92 4.74 6.52 2.42 3.32 – 2.09 2.66

28.59 18.77 17.14 51.89 37.74 45.63 62.77 23.30 31.96 – 20.12 25.61

0.24 0.66 0.71 0.01 0.70 0.11 0.05 0.53 0.26 – 0.54 0.39

Note: as we rate member positions on a discrete scale, Harbinson cannot achieve a zero distance to the average position. Given that we rate the proposal on ten issues, the minimum obtainable distance to the average is 0.89; the maximum distance to Harbinson is 10.39. Correlation is with the Harbinson proposal.

Potential Coalitions and Convergence in the Doha Round

137

Harbinson proposals. Finally, the African group, the USA and South Africa (which has a somewhat different negotiation position than most African countries), which Bjørnskov and Lind (2002) identified as potential negotiation partners, are all placed about 23–32 per cent of the maximum distance away from Harbinson. We can therefore conclude that no country is in full agreement with the draft, although certain countries seem to be positioned close to it. These countries are grouped in the consensus cluster. The findings in Table 6.7 also do not appear to support the EU position that Harbinson’s proposals were not a true compromise. The European complaints about the proposals seem, instead, to result from the fact that the Union has positioned itself very far from the middle of the WTO. With respect to most developing countries, it may be somewhat worrying that they are positioned at some distance from the Harbinson proposal as well as from key players such as Canada. We therefore devote the next section to discussing a set of problems of particular interest to these countries.

Specific Problems for Developing Countries As noted above, some developing countries seem particularly worried about specific components of the Harbinson proposals and the negotiation agenda in general. This section discusses a set of potential problems based on Bjørnskov and Lind (2003), who provide a rigorous analysis of the potential problems of developing countries that may underlie their positions. The reasons for the reluctance to accept broad based liberalization are, however, discernible in Table 6.5. One of the problems is often referred to as preference erosion. Generally, liberalization leads to more trade, which is beneficial to all involved parties. However, in situations where developing countries are allowed lower import duties on exports to developed countries, multilateral liberalization can erode the preference margin enjoyed by developing countries and thus reduce their exports. For example, Yu and Jensen (2003) find that most LDCs will suffer welfare losses from further trade liberalization, which may warrant increased technical and financial cooperation between developed countries and LDCs. This dilemma is reflected in the negotiation positions of developing countries in the WTO. Clusters 2 and 4, which are entirely composed of developing countries – including most of the African group – are less inclined to lower tariffs than the average but more inclined to increase the size of TRQs, which are often the way that trade preferences are implemented. The same does not hold for cluster 6, where most countries are members of the Cairns group, and none are LDCs. The problem of preference erosion thus represents a problem to many developing countries that affects their negotiation positions. As Krueger (1997) stresses, bilateral liberalization such as preference agreements can be an impediment to multilateral trade liberalization by diminishing the benefits to countries that rely on such agreements.

138

K.M. Lind and C. Bjørnskov

The second problem to be stressed is that of erosion of tariff revenues. Many developing countries rely heavily on import duties as a source of government revenue. Four countries in our sample (Madagascar, Sierra Leone, Swaziland and Uganda) obtain approximately half of their total revenue from import duties. For those countries, multilateral liberalization would also imply that they would have to lower tariffs substantially and thereby forego revenue. Substituting towards other sources of revenue such as taxation in developing countries, since they are often characterized by poor infrastructure and lack of institutional development which make it difficult to collect taxes, does not seem a plausible solution to this problem. Nor does increasing the disbursement of foreign aid from developed countries, as countries with high dependence on import duties often tend to be dependent on foreign aid. This problem is easily discernible in Table 6.5, as countries in clusters 1, 2 and 5 on average obtain more than 20 per cent of their total government revenue from import duties. In conjunction with the problem of preference erosion, this tends to make countries’ negotiation positions even more tentative. Both of these problems seem to lead developing countries to be reluctant to embrace substantial trade liberalization under the auspices of the WTO. The problems are of a nature that precludes easy solutions, yet solutions have to be found if multilateral liberalization is to benefit all countries in the medium term. Otherwise, developing countries may find it in their interest to hinder further advancement in the negotiations.

Conclusions This chapter has sought to examine the ongoing negotiations in the World Trade Organisation by following movements, integrating Stuart Harbinson’s draft proposal in the analysis, and identifying groups of countries with similar negotiation objectives. We looked into the issue of which countries could rationally cooperate in the round of negotiations. Related to that we analysed whether the drafts that the chairman of the agricultural committee, Stuart Harbinson, brought forward reflected a genuine compromise between rival positions that ought to have furthered a convergence towards consensus. We identify ten clusters of countries that probably should be negotiating together, as they share most views on how the negotiations should proceed. Most of the African group proves to have similar objectives and are thus grouped together. Developing countries in general appear on the liberal side in the WTO, at least regarding liberalization in industrialized countries. On the other side, the European Union and a small group of highly protectionist countries are isolated in the organization by focusing on non-trade concerns and being reluctant to reducing agricultural domestic support. Cooperating with major players among developed countries could solve developing countries’ problem of bargaining power

Potential Coalitions and Convergence in the Doha Round

139

in the organization. Based on the cluster analysis, the optimal partners for most developing countries seem to be Canada and the Cairns group, while partnering with the traditionally development-friendly European countries emerges as the least favourable option. Secondly, up to the time of collation of the data we found only limited evidence of convergence of the negotiation positions. With the introduction of additional domestic support in the 2002 Farm Bill, the USA took a step away from the liberalizing position they initially supported in the negotiations. The EU took one step forward and another backward with the introduction of the modification of the Common Agricultural Policy. Although the Union appeared ready to cut export subsidies and overall domestic support, these concessions came in a package entailing increased support for non-trade concerns and a considerable broadening of the negotiation agenda. The EU therefore moved only marginally closer to the majority of WTO members. Nevertheless, the joint proposal advanced by Brazil, China, India and other developing countries of the G-20 slightly changed the situation by implying that two clusters containing most of Africa and the USA, and Canada, China and Indonesia, respectively, were then situated close to each other in the WTO. Finally, Stuart Harbinson’s draft proposals seemed to be expressions of genuine compromise, since our cluster analysis places them in the ‘Consensus’ country cluster with Canada and China. Our findings have some broader implications for the round of negotiations in the WTO. First, we can reconfirm the main finding in Bjørnskov and Lind (2002) that the EU is isolated in its demands in the round. Of the major players in the WTO, the EU and Japan are by far the least willing to liberalize, being in sharp contrast to the USA and most developing countries. However, neither the USA nor Harbinson seem to have taken specific developing country worries sufficiently into account when submitting their proposals to the organization. Most developing countries are eager to move forward, yet their positions seem to reflect how broad liberalization might erode preferential access to the North American and European markets in particular, and could cause problems of financing government budgets as tariff revenue decreases. Our findings thus underline Krueger’s (1997) argument that preferential agreements and other agreements that liberalize only selectively and in a discriminatory manner can hinder multilateral liberalization. With respect to the immediate future of the negotiations, during the period analysed certain clusters have been coming together as a result of subtle changes in negotiation positions. This suggests the implication that a majority of members – and a majority of the global population – should be able to reach an agreement. The success or failure of the Doha round therefore rests with a limited number of rich countries that have to find ways of handling political bias and powerful interest groups for the entire membership to come to a consensus on how to move forward in world trade.

140

K.M. Lind and C. Bjørnskov

Notes 1

We originally identified substantially more topics but eventually ended up with the 14 issues, as both an informal analysis and the ratings described below showed that the additional issues were not particularly disputed. 2 In the text we refer to the variety of labour and environmental concerns (that are in turn strongly associated with other ‘softer’ topics, such as the preservation of traditional landscapes) as ‘postmodern issues’. 3 Note that we use ‘liberal’ in the European sense of the word. It should also be noted that the five-step scale used is a conservative choice. In many cases, the language in the official documents would probably allow for a finer division of negotiation positions. 4 There were two Harbinson proposals: the original one and a slightly revised one. The differences between the two proposals lie mainly in details and in the formula approach for tariff reduction, so the ratings do not differ. 5 The EU has attempted to focus on postmodern issues in its preferential trade agreements. The Union, for example, has offered extra tariff margins for countries adopting a ‘social clause’, i.e. a contract that forces countries to adopt certain labour standards. In addition, another tariff margin is offered to countries that guarantee that they will adopt certain environmental standards (EU, 2001). Only Moldova has signed the social clause; no other country has shown interest in the offers. See Oxley (2003) for a background on linking environmental standards and trade sanctions. 6 Swinbank (2001) and Ruffer and Swinbank (2003) document that EU support for multifunctionality and other non-trade concerns will operate as the current Blue Box support. The Union works actively for changes that would place all support for postmodern issues in the Green Box and thus exempt it from future reductions. Jones (2002) argues that authorizing members to use sanctions based on non-trade concerns would undermine the global trading system. 7 The EU and the USA also put forward a joint proposal for the negotiations in the Cancún Ministerial Meeting in late August. However, the proposal was so vague that we have not been able to extract sufficient meaning in order to induce changes in the ratings. 8 An important methodological issue to be taken into account when comparing clusters is that although the average value of the rating gives an indication of similarity, what really counts is the ‘shape’ of the 14-dimensional graph for each group.

References Beck. T., Clarke, G., Groff, A., Keefer, P., Walsh, P. (2001) New tools in comparative political economy: the database of political institutions. World Bank Economic Review 15, 165–176. Bjørnskov, C. and. Lind, K.M. (2002) Where do developing countries go after Doha? An analysis of WTO positions and potential alliances. Journal of World Trade 36(3), 543–562. Bjørnskov, C. and Lind, K.M. (2003) Progress or Retreat in the Doha Round? Analysing

Underlying Policies in the WTO and the Harbinson Proposal. Research note, Danish Research Institute of Food Economics, Frederiksberg, Denmark. EU (2001) Proposal for Council Regulation Applying a Scheme of Generalised Tariff Preferences for the Period 1 January 2002 to 31 December 2004. Commission of the European Communities, Brussels. Jones, K. (2002) The WTO Core Agreement, Non-trade issues and institutional integrity. World Trade Review 1, 257–276.

Potential Coalitions and Convergence in the Doha Round Krueger, A. (1997) Free Trade Agreements versus Customs Unions. Journal of Development Economics vol. 54, 169–187. Mueller, D.C. (2003) Public Choice. Cambridge University Press, UK. Oxley, A. (2003) Environmental trade sanctions. What is at stake? Policy 18(4), 17–22. Panagariya, A. (2002) Developing countries at Doha: a political economy analysis. World Economy 25(9), 1205–1233. Ruffer, T. and Swinbank, A. (2003) Stock-Take of the WTO Agriculture Negotiations: Implications for Developing Countries. Oxford Policy Management, Oxford. Swinbank, A. (2001) Multifunctionality: a

141

European Euphemism for Protection? Paper presented at the FWAG Conference in Stoneleigh, November. World Bank (2002) Global Economic Prospects and the Developing Countries 2002. World Bank, Washington DC. World Bank (2003). World Development Indicators. CD-ROM and on-line database. World Bank, Washington DC. Yu, W. and Jensen, T.V. (2003) Tariff Preferences, WTO Negotiations and the LDCs. The Case of the ‘Everything But Arms’ Initiative. Working paper 04/03, Danish Research Institute of Food Economics, Frederiksberg, Denmark.

7

Assessing the Harbinson Draft on Modalities in the WTO Agriculture Negotiations1 SØREN E. FRANDSEN, HANS G. JENSEN, KIM M. LIND, POUL P. MELGAARD AND WUSHENG YU Danish Research Institute of Food Economics, Rolighedsvej 25, 1958 Frederiksberg C, Denmark

Introduction According to the timetable for the WTO negotiations decided at the Ministerial Meeting in Doha, an agreement on modalities concerning agriculture should have been reached by 31 March 2003. To provide impetus to the process towards this objective the then chairman of the special session on agriculture, Ambassador Stuart Harbinson, presented a compromise proposal on 12 February 2003 (WTO/TN/AG/1). This proposal (along with the second revised drafting of the same proposal in 18 March which introduced some changes in the form of less ambitious reductions and longer transitions periods) was, and still remains, the only tabled negotiating document that presents a comprehensive and detailed compromise proposal with concrete numbers on reductions, ceilings and transition periods. Since the Harbinson drafts there have been other attempts at formulating a framework for the negotiations to move forward, both prior to, and during, the September 2003 WTO Ministerial meeting in Cancún. Those texts, different from the Harbinson drafts, do not take a specific stand in the negotiations, although some wordings are pointing to a particular direction. After the failure at Cancún, negotiations began to move again only with the approval, in August 2004, of the so-called ‘July Package’. However, this text is also a general framework, without quantitative proposals that could be evaluated. For this reason, the Harbinson drafts still remain the most concrete and comprehensive compromise proposal on the table. Moreover, in the progressively more detailed work based on the July Package, some sections of the Harbinson proposal (as in the case of export disciplines and some aspects of the Green Box) have again become the substance of conversation (WTO, 2004). 142

© CAB International 2006. WTO Negotiations and Agricultural Trade Liberalization: the Effects of Developed Countries’ Policies on Developing Countries (eds E. Diaz-Bonilla et al.)

Assessing the Harbinson Draft on Modalities in the WTO Negotiations

143

The purpose of the chapter is to perform a preliminary evaluation of the economic consequences of the draft proposal (section III), as an input for the current negotiations. But first the Harbinson draft is described (section II).

The Harbinson Draft Description of the Harbinson Draft Harbinson Draft on market access TARIFFS For developed countries, the simple average reduction is 60 per cent for tariffs over 90 per cent, with a minimum per-tariff reduction of 45 per cent; for tariffs falling in the range of 15–90 per cent the simple average reduction shall be 50 per cent, with a minimum per-tariff reduction of 35 per cent; finally, for tariffs under 15 per cent, the simple average reduction shall be 40 per cent, with a minimum per-tariff reduction of 25 per cent. For developing countries, a certain flexibility by declaring a number of so-called Strategic Products (SP) is allowed, for which reduction commitments are smaller than for non-SP products. The design at least partially addresses the special and differential treatment issues that are demanded by the developing countries and that are agreed upon by the developed countries. The contentious issues of tariff escalation and tariff peaks are also taken into account, as higher tariff rates need to be reduced more than lower ones, and the difference between average reduction rates and minimum reduction rates is reasonably small. Compared to the proposals presented by other countries, the Harbinson Draft on tariff reduction is truly a compromise. In fact, this multi-step formula approach is a hybrid of the Swiss formula and the Uruguay Approach. Although the simple average cut requirement coupled with the minimum cut embodies the UR approach and hence the EU proposal, cutting high tariffs more than low tariffs (as reflected in the bigger percentage cut assigned to higher tariffs) is clearly the emphasis of the US and the Cairns Group proposals. The magnitude of the proposed reductions for developed countries is in general bounded by the reduction in the US and the EU proposals, with the US proposal being the upper bound and the EU proposal being the lower bound. For example, for a high tariff rate of 100 per cent, the Harbinson formula will probably result in a new tariff of 40 per cent (at most 55 per cent if applying the minimum cut), while the Swiss formula will bring it down to 20 per cent and the EU formula will probably set the new level at 64 per cent – or even as high as 75 per cent if the minimum reduction is applied. For developing countries, the Harbinson scheme is similar to the Cairns Group proposal, but the implied reduction is likely to be smaller under Harbinson. The quota-free and duty-free access for LDCs, as advocated by the EU and the African group, appears in the Harbinson text (under the heading of Domestic Support).

144

S.E. Frandsen et al.

In view of these points, it seems that Harbinson does offer a balanced compromise on tariff reductions. TARIFF RATE QUOTAS For developed countries, Harbinson proposes expanding TRQ volumes to 10 per cent of domestic consumption. Members are also allowed to have the flexibility of expanding up to 25 per cent of all TRQs by only 8 per cent of domestic consumption of these products, provided that the expansion of a corresponding number of TRQs reaches 12 per cent of their respective domestic consumption. For developing countries, there is no expansion requirement for their SP products. Again, Harbinson positions his first attempt on TRQ to be somewhere between the EU, which does not want any changes in relation to expanding the market access through TRQ, those who want to set the quota at a significantly higher level and those who seek to eventually abolish this system. In fact, the Harbinson Draft does not want to reduce the in-quota tariff per se, a position similar to the EU but quite different from the USA and the Cairns Group, who want to remove the in-quota tariff. On the other hand, Harbinson advocates the Cairns Group position on expanding the quota to a certain percentage of domestic consumption, although the increase is smaller in the case of Harbinson. The African Group’s request on special measures for small-scale farmers is not directly addressed, but Harbinson does provide some provisions for developing countries in general. SPECIAL SAFEGUARDS According to Harbinson, Article 5 shall cease to apply for developed countries at the end of, or 2 years after, the implementation period for further tariff reduction. Developing countries shall have the flexibility to apply special safeguard measures on SP products along the line of Article 5 provisions. This position is quite close to what the Cairns Group proposes in addressing the needs raised by many developing countries. It differs from the US position, as the latter does not address the need of developing countries, and it also differs from the EU position, as the EU also wants this article to cover developed countries. IMPORTING STE Harbinson drafts some possible new provisions for importing STE (State Trading Enterprise) regarding the definition of importing STE, the permissible conduct of the STE and the notification requirements. Tables 7.1, 7.2 and 7.3 provide some statistics on tariffs and the reductions proposed by Harbinson.

Harbinson Draft on export competition EXPORT SUBSIDIES The Harbinson text states that all reductions on export subsidies shall be based on final bound budgetary outlay and quantity commitments. For developed countries, export subsidies on products

Brazil

Coffee, roasted Sugar, raw Rice, unmilled grain Wheat, grain Bananas, fresh Grapes, fresh Fruit juice, grape Tomatoes, fresh Milk and cream, powder and condensed Butter Cheese Beef, frozen Pork, frozen

EU

B

A

W

B

A

35.0 35.0 55.0 55.0 35.0 28.8 35.0 35.0 31.5

13.0 19.0 13.0 13.0 13.0 13.0 17.0 13.0 27.0

22.0 16.0 42.0 42.0 22.0 15.8 18.0 22.0 4.5

0.4 68.9 66.8 90.8 16.4 15.3 138.9 57.4 22.5

0.4 68.9 66.8 90.8 16.4 15.3 138.9 14.4 22.5

55.0 55.0 55.0 55.0

19.0 27.0 13.0 13.0

36.0 28.0 42.0 42.0

93.4 73.2 43.3 31.9

93.4 73.2 43.3 31.9

Japan W

B

A

Malawi W

B

0.0 0.0 0.0 0.0 125.0 0.0 522.1 108.9 413.2 125.0 0.0 3,562.6 511.9 3,050.7 125.0 0.0 451.3 80.4 370.9 125.0 0.0 22.5 15.0 7.5 125.0 0.0 12.4 12.4 0.0 125.0 0.0 19.1 19.1 0.0 125.0 43.0 3.0 3.0 0.0 125.0 0.0 450.4 146.4 304.0 125.0 0.0 1,031.1 225.1 0.0 29.8 29.8 0.0 50.0 38.5 0.0 4.3 4.3

806.0 0.0 11.5 0.0

125.0 125.0 125.0 125.0

A

USA W

B

A

W

10.0 25.0 10.0 0.0 10.0 10.0 25.0 10.0 10.0

115.0 0.0 0.0 0.0 100.0 91.6 32.1 59.5 115.0 20.3 2.3 18.0 125.0 5.8 4.4 1.4 115.0 0.0 0.0 0.0 115.0 0.3 0.3 0.0 100.0 6.2 11.7
5.5 115.0 4.0 2.4 1.6 115.0 133.7 133.7 0.0

25.0 25.0 10.0 10.0

100.0 107.3 100.0 63.5 115.0 31.1 115.0 0.0

48.9 33.9 26.4 0.0

58.4 29.6 4.7 0.0

Assessing the Harbinson Draft on Modalities in the WTO Negotiations

Table 7.1. Bound and applied tariff rates and water for selected products and countries in 2000, percentage. (From WTO Integrated DataBase, Eurostat, FAOSTAT, USDA, OANDA, own calculations.)

B – bound tariff rates; A – applied tariff rates; W – water (water is a WTO-term for the difference between bound and applied rate).

145

146

Table 7.2. Bound and applied tariff rates and water for selected products and countries after implementation of Harbinson, percentage. (From own calculations.) Brazil

Coffee, roasted Sugar, raw Rice, unmilled grain Wheat, grain Bananas, fresh Grapes, fresh Fruit juice, grape Tomatoes, fresh Milk and cream, powder and condensed Butter Cheese Beef, frozen Pork, frozen

EU

Japan

B

A

W

B

A

W

B

A

23.5 23.5 36.9 36.9 23.5 19.3 23.5 23.5 21.1

13.0 19.0 13.0 13.0 13.0 13.0 17.0 13.0 21.1

10.5 4.5 23.9 23.9 10.5 6.3 6.5 10.5 0.0

0.2 34.5 33.4 36.3 8.2 7.7 55.6 28.7 11.3

0.2 34.5 33.4 36.3 8.2 7.7 55.6 14.4 11.3

0.0 0.0 0.0 0.0 0.0 0.0 0.0 14.3 0.0

0.0 208.8 1,425.0 180.5 11.3 7.4 9.6 1.8 180.2

0.0 108.9 511.9 80.4 11.3 7.4 9.6 1.8 146.4

36.9 36.9 36.9 36.9

19.0 27.0 13.0 13.0

17.9 9.9 23.9 23.9

37.4 36.6 21.7 16.0

37.4 36.6 21.7 16.0

0.0 0.0 0.0 0.0

412.4 14.9 25.0 2.6

225.1 14.9 25.0 2.6

Malawi W

USA

B

A

W

B

A

W

0.0 99.9 913.1 100.1 0.0 0.0 0.0 0.0 33.8

75.0 75.0 75.0 75.0 75.0 75.0 75.0 75.0 75.0

10.0 25.0 10.0 0.0 10.0 10.0 25.0 10.0 10.0

65.0 50.0 65.0 75.0 65.0 65.0 50.0 65.0 65.0

0.0 36.6 10.2 3.5 0.0 0.2 3.7 2.4 53.5

0.0 32.1 2.3 3.5 0.0 0.2 3.7 2.4 53.5

0.0 4.5 7.9 0.0 0.0 0.0 0.0 0.0 0.0

187.3 0.0 0.0 0.0

75.0 75.0 75.0 75.0

25.0 25.0 10.0 10.0

50.0 50.0 65.0 65.0

42.9 31.8 15.6 0.0

42.9 31.8 15.6 0.0

0.0 0.0 0.0 0.0

B – bound tariff rates; A – applied tariff rates; W – water (water is a WTO-term for the difference between bound and applied rate). S.E. Frandsen et al.

Brazil B

A

EU W

Coffee, roasted
33.0 0.0
52.5 Sugar, raw
33.0 0.0
72.2 Rice, unmilled grain
33.0 0.0
43.2 Wheat, grain
33.0 0.0
43.2 Bananas, fresh
33.0 0.0
52.5 Grapes, fresh
33.0 0.0
60.2 Fruit juice, grape
33.0 0.0
64.2 Tomatoes, fresh
33.0 0.0
52.5 Milk and cream, powder
33.0
21.8
100.0 and condensed Butter
33.0 0.0
50.4 Cheese
33.0 0.0
64.8 Beef, frozen
33.0 0.0
43.2 Pork, frozen
33.0 0.0
43.2

B

A

Japan W

B

A

Malawi W

B

USA

A

W

B

A

W


40.0
50.0
50.0
60.0
50.0
50.0
60.0
50.0
50.0


40.0 0.0
50.0 0.0
50.0 0.0
60.0 0.0
50.0 0.0
50.0 0.0
60.0 0.0 0.0
66.8
50.0 0.0

0.0
60.0
60.0
60.0
50.0
40.0
50.0
40.0
60.0

0.0 0.0 0.0
75.8 0.0
70.1 0.0
73.0
25.0
100.0
40.0 0.0
50.0 0.0
40.0 0.0 0.0
88.9


40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0

0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0


43.5
50.0
43.5
40.0
43.5
43.5
50.0
43.5
43.5

0.0
60.0
50.0
40.0 0.0
40.0
40.0
40.0
60.0


60.0
50.0
50.0
50.0


60.0
50.0
50.0
50.0


60.0 0.0
76.8
50.0
50.0 0.0
50.0
35.1
100.0
40.0
40.0 0.0


40.0
40.0
40.0
40.0

0.0 0.0 0.0 0.0


50.0
50.0
43.5
43.5


60.0
12.2
100.0
50.0
6.3
100.0
50.0
41.1
100.0 0.0 0.0 0.0

0.0 0.0 0.0 0.0

0.0 0.0 0.0
92.4 0.0
56.4
20.9
100.0 0.0 0.0
40.0 0.0
68.4
100.0 0.0
100.0
60.0 0.0

Assessing the Harbinson Draft on Modalities in the WTO Negotiations

Table 7.3. Change in bound and applied tariff rates and water for selected products and countries due to Harbinson, percentage. (From own calculations.)

B – bound tariff rates; A – applied tariff rates; W – water (water is a WTO-term for the difference between bound and applied rate).

147

148

S.E. Frandsen et al.

representing at least 50 per cent of the outlay shall be reduced each year by 30 per cent of the previous year’s value for 5 years, and shall be eliminated in year 6; for the remaining subsidies, per year reduction of 25 per cent of the previous year’s outlay shall be applied for 9 years and the subsidies shall be eliminated in year 10. For developing countries, export subsidies on products representing at least 50 per cent of the outlay shall be reduced by 25 per cent each year for 10 years, and shall be eliminated in year 11; for the remaining subsidies, a reduction of 20 per cent per year shall be applied for 12 years and the subsidies shall be eliminated in year 13. Exemptions for developing countries under Articles 9.4 and 9.1(d) and (e) shall be maintained. Most of the proposals aim at eliminating export subsidies in a relatively short time frame (the USA: 5 years; Cairns Group: 3 years for developed countries; India: 2 years), while the EU proposes only a 45 per cent cut in budgetary outlays. Against this background, Harbinson tries to strike a balance by seeking the elimination over a longer period. The special treatment for developing countries will also be able to continue and smaller cuts over a longer time span are also proposed for these countries. OFFICIALLY SUPPORTED EXPORT CREDITS, EXPORT CREDITS GUARANTEES AND INSURANCE PROGRAMMES

Harbinson sets out the terms and conditions on export credits and other forms of government export financing and proposes to disallow nonconforming support. The terms and conditions include maximum repayment term, cash payment, payment of interest, minimum interest, repayment of principles and risk premiums. This position is in line with the proposal of the Cairns Group and is certainly more specific and concrete than the US proposal. However, the EU and India ask for stronger disciplines as both want to subject these measures to the same disciplines/reduction commitments as those applied for export subsidies. As such, Harbinson is again positioned in the centre. FOOD AID Harbinson offers specific proposals on a possible replacement of paragraph 4 of Article 10, including, but not limited to: the conditions under which food aid can be given, the forms in which food aid can be provided, the procedures to carry out food aid transactions and the reporting requirements. This position incorporates many of the points raised in the Cairns Group proposal and accommodates the EU’s call for disciplines on food aid in kind. Table 7.4 provides a summary of the provisions on export competition in the Harbinson Draft.

Harbinson Draft on domestic support GREEN BOX Harbinson wants to maintain the provisions in Annex 2 with some amendments. These amendments either tighten the eligibility and criteria for detailed Green Box measures – for example, for direct payment,

Assessing the Harbinson Draft on Modalities in the WTO Negotiations

149

Table 7.4. Export subsidies in US$ million, 1995–1998. (From OECD, 2002.)

Australia Canada EU Hungary Norway USA

1995

1996

1997

1998

0 37 6386 41 83 26

0 4 7064 18 78 121

0 0 4943 10 102 112

1 0 5968 12 77 147

decoupled income support, structural adjustment assistance and regional assistance programmes, the bases of the payments are defined as ‘a fixed and unchanging historical base period’; or cap the payment to a certain level. The text proposes various SDT (Special and Differential Treatment) for developing countries on domestic support. Many of the points made by the Cairns Group proposal feature prominently in the Harbinson Draft, including the emphasis on fixed and unchanged historical periods for a number of Annex 2 measures, and the extent and length of several types of payments. The EU’s wish to have several types of non-trade concerns respected through the Green Box receives mixed consideration. The issues of food safety and labelling are not touched upon in Harbinson. Food security for developing countries, however, received considerable attention in the proposals of special and differential treatment, as did the enhancement to Article 6.2. Animal welfare is added to the Harbinson text, along with environmental programmes, although the payments on such programmes shall be less than the extra cost involved in complying with the government programmes, which differs from the EU’s total exemption request. BLUE BOX Blue Box measures shall be capped at the average notified level during the period 1999–2001 and bound at that level in members’ schedules. Thereafter, these payments shall be reduced in equal instalments over a 5-year period by 50 per cent. An alternative proposal is that these payments shall be included in the calculation of total AMS for developed countries, and therefore subject to the reduction commitment listed below. For developing countries, the reduction shall be 33 per cent over 10 years. This position is mid-way between the ‘maintaining’ proposal of the EU and the ‘eliminating’ proposals from the USA, the Cairns Group and Canada. AMBER BOX For developed countries, the final bound total AMS shall be reduced by 60 per cent in equal annual instalments over a 5-year period. For developing countries, the reduction of AMS shall be 40 per cent over a 10-year period. The de minimis level of 5 per cent for developed countries shall be reduced by 0.5 percentage points per year for 5 years, resulting in halving the de minimis levels. For developing countries, the 10 per cent de minimis levels shall be maintained.

150

S.E. Frandsen et al.

Virtually all the proposals agree with the reduction of AMS in developed countries to some degree, ranging from the EU’s 55 per cent and the Cairns Group and Canada’s 100 per cent reduction from the final bound levels, to the US reduction to 5 per cent of production values. Harbinson’s position of a 60 per cent reduction is slightly more than what is offered by the EU, but far less than the 100 per cent reduction. It is quite difficult to compare the US proposal with the Harbinson without knowing the magnitude of the 5-per cent production values. Therefore, a resort to more data is required in order to pin down this relative position. The other important part of the Amber Box is the de minimis exemption. The reduction proposal of 2.5 percentage points (or halving the current levels) is halfway between the US ‘maintaining’ proposal (in the implementation period) and the EU’s ‘eliminating’ proposal. In fact, this position is close to that of the Cairns Group, although the latter does not specify how much reduction shall take place. Maintaining the de minimis levels for developing countries receives support from the members’ proposals – for example, the Cairns Group proposal. The EU does not apply the reduction to developing countries, either. The African Group and India’s requests for increased flexibility for developing countries are partially fulfilled in Harbinson by allowing these countries to credit negative product-specific support to non-product-specific support. Other requests, such as allowing developing countries with zero AMS to raise it above zero (by the African Group), are not accommodated. Table 7.5 shows actual domestic support and Table 7.6 displays the implied changes in domestic support due to the Harbinson Draft. Evidently, the required changes in agricultural policies vary widely among the members. In particular, the EU has to reduce its support by a third. Also, India and the USA would have to reduce substantially, although only by half of the EU’s requirements.

A preliminary appraisal of the Harbinson Draft based on members’ positions The cluster analysis presented in Chapter 6 in this book suggests that the Harbinson Draft seems to fulfil its intended role in forming a possible ‘bridge’ between all the proposals. However, a major player – the EU – has a position removed from the Harbinson Draft. Thus, the Harbinson Draft’s ‘bridge’ does not seem to reach the EU. But if Harbinson had positioned his draft closer to the EU it would, in turn, have been far removed from most of the other WTO members. The Canadian position is quite close to the Harbinson Draft. Among some of the key players Canada is positioned in the average group. The African group proposal is reasonably, but not entirely, close to the Harbinson Draft. A large number of countries in this group are LDCs, and the rest are developing countries. Thus, the Harbinson Draft does take into account some of the considerations of the developing countries, and in particular of the least-developed countries. In Chapter 6 it was also shown that the US and Cairns Group positions are relatively closer to the African group, while the EU

Assessing the Harbinson Draft on Modalities in the WTO Negotiations

151

Table 7.5. Domestic support in US$ million. (From WTO.)

Argentina Australia Brazil Canada EU India Japan New Zealand Pakistan South Africa USA

Year

AMS bound

Applied AMS

Blue Box support

Green Box support

De minimis support

Total

1999/00 2000/01 1997/98 1999 1999/00 1997/98 1999 2000/01 1999/00 2000 1999

81 254 893 3,015 69,446 0 36,359 122 0 290 19,899

80 115 74 632 47,874 0 6,572 0 0 63 16,862

0 0 0 0 19,787 0 817 0 0 0 0

298 711 2,167 1,177 19,926 75 23,601 89 5 392 49,749

0 11 94 742 308 98 290 0 0 1 7,435

378 837 2,335 2,551 87,895 172 31,280 89 5 457 74,046

Note: ‘Year’ shows the year in which data on domestic support are available for the particular country. ‘AMS bound’ is the upper bound on AMS (Amber Box) agreed upon in the Uruguay Round. ‘Applied AMS’ is the amount of AMS support used by the country. ‘Blue Box support’ shows support that is distorting, but allowed under the AoA. ‘Green Box support’ contains non-distorting support. ‘De minimis support’ shows allowed very distorting support. ‘Total’ is the sum of AMS applied, Blue Box support, Green Box support and de minimis support.

appears in opposition to the stated positions of most developing countries. There are three main reasons for this result. First, the EU is against elimination of export subsidies where the developing countries regard these as dumping. Second, the EU wants to maintain provisions for distorting support in the Blue Box – and to some extent the Amber Box – whereas developing countries argue for their elimination. Third, developing countries see the introduction of various forms of non-trade concerns as hidden protectionism, whereas the EU is a strong advocate of standards on labour and the environment and of support for the multifunctionality of agriculture. In summary, a few observations are offered: ● ●

●

●

On issues where quantitative targets can be established (e.g. tariffs), the Harbinson Draft seeks to find the middle ground. On issues where there are multiple dimensions (e.g. TRQs), the Harbinson Draft seeks to make progress on the more important dimensions (e.g. expansion of quotas), rather than push all the dimensions forward simultaneously (e.g. maintaining the in-quota tariff). On more complex issues where members disagree on different dimensions (Amber Box and de minimis levels), the Harbinson Draft seeks to reach compromise on all these dimensions. The issue of special and differential treatment is explicitly respected through a proposal format consisting of two parts, one for developed countries and the other for developing countries. This treatment is often expressed in the form of lower reduction commitments, longer implementation period, more exemptions and less strict rules.

152

Table 7.6. Required changes in domestic support in US$ million.

Argentina Australia Brazil Canada EU India Japan New Zealand Pakistan South Africa USA

Year

 AMS bound

1999/00 2000/01 1997/98 1999 1999/00 1997/98 1999 2000/01 1999/00 2000 1999


32
152
357
1,809
41,667 0
21,815
73 0
116
11,939

 Applied AMS
31
14 0 0
20,096 0 0 0 0 0
8,902

 Blue Box support

 Green Box support

 De minimis support

 App. PSE

 App. PSE, %

0 0 0 0
9,894 0
409 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0

0
3 0
222
92 0
87 0 0 0
2,231


31
17 0
222
30,082 0
496 0 0 0
11,133


8.3
2.0 0.0
8.7
34.2 0.0
1.6 0.0 0.0 0.0
15.0

S.E. Frandsen et al.

Note: ‘Year’ shows the year in which data on domestic support is available for the particular country. ‘’ is symbol for ‘Change in’. ‘AMS bound’ is the upper bound on AMS agreed upon in the Uruguay Round. ‘Applied AMS’ is the of AMS support used by the country. ‘Blue Box support’ shows support under this category. ‘Green Box support’ shows support under this category. ‘De minimis support ’ shows the amounts of support falling under the de minimis levels. Approximate PSE is the sum of Applied AMS, Blue Box support, Green Box support and de minimis support. A caveat is warranted. These calculations have been applied on fairly crude and aggregated data. The picture may change somewhat when the details are applied, although it is the impression of the authors that the table provides a fairly good approximation.

Assessing the Harbinson Draft on Modalities in the WTO Negotiations ●

●

●

153

’New’ issues raised by developed countries – such as non-trade concerns – have the potential for further discussion but have not yet received sufficient gravitas in the current text. Many rule changes are proposed, such as TRQ administration, import and export STE, criteria and eligibility with respect to domestic support, food aid and special safeguards. In general, proposals with explicit quantitative targets and detailed texts on rule changes receive more attention than those with nonspecific targets and non-operational texts. One example is the Cairns Group proposal, whose various components have been used by Harbinson. This does not, however, suggest that the Harbinson Draft favours one member over the other. It is probably due more to the fact that Harbinson needs to have an operational starting point.

In conclusion, it seems that the Harbinson Proposal has achieved the two objectives that it was meant to achieve (defining a negotiation mandate and trying to close the differences among WTO members). The critical reaction received from the participants was understandable, since a truly balanced compromise implies concessions from all the parties involved. But what would have been the quantitative impact of this proposal? This question is discussed next.

Implications of Agricultural Liberalization: the Harbinson Draft This section tries to quantify the global economic impacts of the Harbinson Draft and, in particular, the macroeconomic effects in terms of economic welfare and gross national product in a number of developing and developed countries, as well as the impacts on global agricultural trade and production. Modelling of the Harbinson Draft The economic analysis is based on an economic model of the world economy, with particular emphasis on global trade and production covering 24 different product categories (of which 12 are primary agricultural products and eight are processed food products) in 40 countries and regions. The starting point of the analysis is the Global Trade Analysis Project (GTAP) database and model, (Hertel, 1997). The model is used to project the world economy (and the database) from 1997 to 2013 (constructing a baseline), including economic and population growth in general, the implementation of the Uruguay Round Commitments, the EU Everything But Arms Initiatives (free market access for the Least Developing Countries), the Agenda 2000 Reform of the Common Agricultural Policy, as well as the enlargement of the European Union, with the admission of ten new Central and Eastern European members (as laid down in the Copenhagen Agreement, December 2002), see Box 7.1.

154

S.E. Frandsen et al.

Box 7.1. Assumptions shaping the baseline scenario 1997–2013. Projections of the World Economy from 1997 to 2013 Regional income, capital and population growth and continued sector-specific total factor productivity increase Uruguay Round Agreement Full implementation of URA, e.g. if export subsidy commitment (in value terms) is binding, the export subsidy rate is reduced Agenda 2000 Reform Intervention prices are reduced (import tariff and export subsidy reductions) Hectare and livestock premiums and milk quotas are adjusted according to reform National Envelopes and new premiums are introduced All direct payments are deflated by 2 per cent per year (the (maximum) budgetary outlays are fixed in nominal terms). Setaside reflects the 10 per cent requirement and the sugar quota is unchanged Blair House Agreement concerning oilseeds is abolished European Union Enlargement Expansion of the CAP (including the common financing of the agricultural policy) to Cyprus, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Malta, Poland, the Slovak Republic and Slovenia following the outlines for domestic support (direct payments, production quotas and other supply management instruments) as laid down by the Copenhagen Agreement. All tariffs and export subsidies as well as non-tariff barriers between the EU and the CEEC are abolished. At the same time all sectors in the CC10 are given the same level of protection against third countries as found in the EU at the time of accession. Everything But Arms All EU (including ten new member countries) tariff rates reduced to zero on imports from Malawi, Mozambique, Tanzania, Zambia, Uganda and the two aggregate regions Other Southern Africa and the Rest of sub-Saharan Africa. Harbinson scenario 2013 Market access Tariff Reductions are implemented using the following template: In the EU-25, USA, Bulgaria, Romania and the ROW region the following cuts are implemented: Tariff rate equivalents > 90% 60% reduction Tariff rate equivalents 15–90% 50% reduction Tariff rate equivalents < 15% 40% reduction In Africa, China, Latin America and the Caribbean the following cuts are implemented: Tariff rate equivalents > 120% 40% reduction Tariff rate equivalents 20–120% 33% reduction Tariff rate equivalents < 20% 27% reduction All export subsidies are eliminated Domestic support Domestic support is reduced by 34.2 per cent in the EU-25 15.0 per cent in the USA 1.5 per cent in all other countries

Assessing the Harbinson Draft on Modalities in the WTO Negotiations

155

Like most other economic analyses of such trade talks and agreements, this study is conducted at an aggregate level, which for many products does not get to the detail of, say, six- or eight-digit commodity-level identification, which has implications for bound and applied rates, or for the details of individual direct support schemes. Therefore, this study is an attempt at quantifying the possible broad impacts of implementing the Harbinson Draft. The results found should therefore be interpreted carefully, by taking into consideration the particular assumptions and data that drive the results. Nevertheless, the study does give a first indication of the magnitude and directions of the impacts of implementing the Harbinson Draft. The assumptions underlying the analysis can be found in Box 7.1 (see also FØI (2003) for greater detail). The Harbinson Draft is represented in our simulations as follows. First, all agricultural tariffs are reduced as described in the box (60 per cent cuts for the highest tariffs in the developed countries to 27 per cent cuts for the lowest tariffs in the developing countries). Second, all export subsidies are eliminated completely (which mainly covers the EU, as described above). Third, domestic support measures are reduced by 34.2 per cent in the European Union, 15.0 per cent in the USA and by 1.5 per cent in the remaining countries in the rest of the world. These cuts reflect the level and composition (Green, Blue, Amber and de minimis support) of the domestic support measures in the individual countries, as explained above.

GDP and welfare Based on these assumptions, it is estimated that the global real income will increase by 100 billion US dollars per year (current 2003 prices) compared to the projected 2013 world economy database, corresponding to a global income gain of approximately 0.2–0.3 per cent (Tables 7.7 and 7.8). To put this number into perspective, this corresponds to twice the amount of the official development assistance being disbursed to the developing countries in 2001. The global welfare gains of 100 billion US dollars are similar to those projected by other studies of comparable liberalization scenarios. A study by Harrison et al. (2003) reported global welfare gains of 186 billion US dollars following a 50 per cent cut in tariffs and export subsidies. Likewise, François et al. (2003) arrived at much the same figures (196 billion US dollars) for a 50 per cent cut in tariffs; however, this study included services as well as goods. In the Brown et al. (2003) study a vastly larger figure (1040 billion US dollars) – again for a 50 per cent cut – was found. These gains, however, included the dynamic effects of foreign direct investments. As Anderson (2004) pointed out, the results of the static gains probably underestimated the true gains because dynamic effects such as growth effects were not taken into consideration.2

156

Table 7.7. Impacts of the Harbinson Draft: sector and macroeconomic indicators, EU-25. (From simulation results.)

EU-25

GDP (%)

Allocative efficiency (US$m)

Terms of trade (US$m)

CAP transfers (US$m)

Exports (%)

Imports (%)

Land prices (%)

0.2 0.2 0.1 0.4 0.2 0.1 0.3 0.1 0.3 0.2 0.3 0.1 0.3 0.2 0.1 0.1 0.3 0.1 0.2 0.5 0.1 0.1 0.2

1,131 759 6,726
202 1,051 3,307
140 3,689 2,381 919 276 397 1,729 6,167
14
238
1
473
51
203
1,316
73
30

677 401 2,823 670 1,652 2,359 510 1,624 1,385 487 497 261 855 3,191 18 69 21 110 17 67 174 39 47

281 504 1,784
97 277 1,612
63 1,101 857 274
50 162 803 1,580
23
53 8
85
17
79
438 1
42

150
106 2,147
692
920
543
578 1,066 227 103
162 8 208 1,218
7
252
33
505
45
167
974
119
34


6.6 6.8
3.1
13.2 0.3 15.8
4.0 22.9 1.4 47.1 14.7
8.0 108.0 22.4 0.9
4.6
6.2
2.1
25.5
18.5
9.9
6.3
19.6


0.1 7.2 3.6 3.5 8.8 8.1
0.2 8.9 3.0 6.3 5.8 3.3 17.6 7.7 0.8
1.2
1.0
0.9
3.7
6.3
3.5 0.2
1.9


31.7
16.9
31.0
30.4
29.2
27.2
38.8
25.8
25.8
24.9
30.0
31.8
25.5
26.8
11.0
25.7
27.4
24.8
32.8
32.2
26.5
27.7
28.8

–

25,792

17,954

8,297


11

–

–

–

S.E. Frandsen et al.

Belgium/Luxembourg Denmark Germany Greece Spain France Ireland Italy Netherlands Austria Portugal Finland Sweden UK Cyprus/Malta Czech Republic Estonia Hungary Latvia Lithuania Poland Slovakia Slovenia

Agricultural

Total welfare (US$m)

Bulgaria Romania North Africa Botswana Rest of South African Customs Union Malawi Mozambique Tanzania Zambia Zimbabwe Other Southern Africa Uganda Rest of sub-Saharan Africa China USA Latin America and the Caribbean Rest of the World

GDP (%)

Total welfare (US$m)

Allocative efficiency (US$m)

Terms of trade (US$m)

0.0 3.7 2.3 0.4 0.3

22 1,346 5,640 30 777

0 1,616 8,526 39 657

15
221
3,084 0 119

0.1 0.1 0.0 0.0 0.3 0.3 0.0 0.1


16 1
64
8 2
15
7
74

5 7 3 2 26 87 7 311

0.4 0.0 0.2

7,867 14,744 3,160

0.4

CAP transfers (US$m)

Agricultural Exports (%)

Imports (%)

Land prices (%)

0 0 0 0 0

15.8 59.3 444.2 78.8 29.2

6.1 248.3 47.1 18.2 46.9


1.1
10.0 13.4 12.6 0.5


21
6
61
9
24
127
14
369

0 0 0 0 0 0 0 0


5.9
6.7
9.1
21.0 1.0
10.2
5.4
6.6

53.5 9.7 11.0 9.3 29.6 5.2 15.5 16.6


5.6
4.4
9.7
9.4
5.9
23.1
2.7
8.9

9,696 4,937 4,925


2,557 8,815
2,117

0 0 0

19.8 17.6 43.0

20.1 25.6 26.7


3.9
6.4 5.8

39,782

50,123


8,557

0

63.3

50.0


16.5

Regions other than EU-25

–

73,188

80,966


8,219

0

–

–

–

EU-25 (from Table 7.7)

–

25,792

17,954

8,297


11

–

–

–

Total World

–

98,980

98,919

78


11

25.7

25.7

Assessing the Harbinson Draft on Modalities in the WTO Negotiations

Table 7.8. Impacts of the Harbinson Draft: Sector and macroeconomic indicators, regions other than EU-25. (From simulation results.)

157

158

S.E. Frandsen et al.

The global welfare gain in dollar amounts is not evenly distributed among the developed and developing countries, as 80 per cent of the global gains accrue to the enlarged European Union (25.8 billion US dollars), the USA (14.7 billion US dollars) and Canada, Japan, Australia and New Zealand (40 billion US dollars) – the latter group of countries constituting by far the largest share of the ‘Rest of the World’ aggregate. The remaining part of almost 20 billion US dollars accrues to a relatively large number of developing countries, including the least developing countries. However, in percentage real income gain, the benefits are more evenly distributed: the economic gains in, for example, North Africa reach 2.3 per cent of the GDP, in Botswana 0.4 per cent, Rest of South African Customs Union 0.3 per cent and China 0.4 per cent as compared to the (relative) income gain in the EU (0.2 per cent), USA (0.0 per cent) and Rest of the World (0.4 per cent) (see Figs 7.1 and 7.2). To understand the results of the individual countries it is important to note that in all regions, the allocative efficiency (i.e. reallocation of resources of production to their most efficient use) is in general positive (as is the impact on GDP in all the regions), whereas the terms-of-trade effects (the ratio of export to import prices) differ significantly between countries. Also, removing parts of the direct support in the EU (i.e. reducing domestic support by 34 per cent on average) leads to a redistribution of financial resources among the EU members, and therefore welfare gains and losses correspondingly. Decomposing the results further also illustrates that most of the economic gains arise from efficiency gains from countries’ own liberalization – the one critical factor in explaining the relatively large welfare gains in, for example, the European Union, although some foodexporting countries, such as Australia and New Zealand, realize important welfare gains from very significant and positive terms-of-trade effects. For a number of the developing countries the terms of trade effect is quite significant in explaining the (negative) welfare consequences. This price effect depends on whether a country is a net exporter or net importer 0.6 0.5 0.4 0.3 0.2 0.1

Fig. 7.1. The Harbinson Draft: impacts on GDP in EU-25.

Poland

Cyprus/Malta

Czech Republic

Italy

Germany

France

Finland

Slovakia

Hungary

Slovenia

Austria

Denmark

United Kingdom

Belgium/Luxembourg

Spain

Latvia

Sweden

Netherlands

Estonia

Ireland

Greece

Portugal

Lithuania

0.0

Assessing the Harbinson Draft on Modalities in the WTO Negotiations

159

0.6000 0.5000 %

0.4000 0.3000 0.2000 0.1000 Bulgaria

Tanzania

Zambia

United States of

Uganda

Mozambique

Malawi

Rest of subSaharan

Latin America

Zimbabwe

Other Southern

Rest of South

China

Rest of the World

Botswana

0.0000

Fig. 7.2. The Harbinson Draft: impacts on GDP in other regions excluding EU-25.

of the affected agricultural products (i.e. rising food import prices in heavily import-dependent developing countries leads to a term of trade loss), and on the extent and value of possible preferential access opportunities to the high-priced (and protected) developed market (i.e. the European Union and the Everything But Arms initiative granting free access for all commodities from the least-developed countries to the European market). Such preferences are by definition eroded in any trade liberalization scenario. Nevertheless, the simulation illustrates that the developed countries, such as the EU and the USA, could fully compensate the developing countries for lost welfare and still be better off if such a step – combining trade liberalization and development aid – was initiated. The welfare losses in the least-developed countries are relatively minor (some US$200–250 million), as compared to the combined EU–US gain of more than US$40 billion. Therefore, there is in principle a clear economic potential for compensating potential losers in such a trade liberalization scenario. Also, combining such initiatives with domestic reforms in the developing countries could modify – or even dominate – the possible negative impacts of changes in world market prices and the erosion of trade preferences. Moreover, considering other dynamic aspects linked to employment, technological or investment effects not factored in here, would imply even better welfare results for developing countries. The results from the simulations also demonstrate the significant negative impacts on European land prices, as they typically fall by 25–30 per cent due to the reductions of, in particular, the existing (and distorting) direct payments to European farmers.

Trade The impact of the Harbinson Draft on world agricultural trade volumes is also very significant, corresponding to an increase of some 25 per cent

160

S.E. Frandsen et al.

(Table 7.9). The somewhat different impacts across the regions and countries globally naturally reflect differences in the structure of agricultural production in the individual countries and regions considered. Impacts of the Harbinson Draft on agricultural trade volumes for the individual agricultural commodities are also very noteworthy, as the global export of most products expands significantly, with a few exceptions such as sugar cane and beet and bovine animals (products not being traded) (Table 7.9). For example, increases in global trade in processed rice, wool, beverages and tobacco, bovine meats and dairy products were 67, 35, 49, 28 and 24 per cent, respectively. While global trade in almost all the agricultural commodities increases, global production expands only marginally, mirroring a significant geographical redistribution of global agricultural production.

Concluding Remarks The Harbinson drafts still remains the most concrete and comprehensive compromise proposal on the table. Moreover, some of its components are Table 7.9. Impacts of the Harbinson Draft: global agricultural export and production. (From simulation results.) Commodities Paddy rice Wheat Other grains Vegetables, fruit, nuts Oilseeds Sugar cane and beet Plant-based fibres Other crops Bovine animals Other animal products Raw milk Wool Natural resources Bovine meat products Other meat products Vegetable oils and fats Dairy products Processed rice Sugar Other processed foods Beverages and tobacco Textiles/wearing apparel Manufactures Services Total world

Export (%)

Production (%)

50.3 26.0 17.6 17.2 20.2
1.5 25.0 16.2
1.9 1.7 11.0 34.9 0.5 28.2 18.2 14.8 24.1 66.6 26.9 26.1 49.2 5.8 1.3 2.1


2.2
0.4 0.3 0.2 0.2
0.5 2.2
0.3 0.0 0.3
1.4 0.6 0.0 0.1 0.2 0.3 0.2 0.2 0.1 0.4 1.0 0.6 0.1 0.1

3.3

–

Assessing the Harbinson Draft on Modalities in the WTO Negotiations

161

being considered as part of the current negotiations, after the July 2004 Package. The Harbinson draft was an attempt to bridge the material differences among the WTO countries. On a range of issues members’ views are still separated and, consequently, the suggested compromises in the draft met with reservations from some and stark opposition from others. Nevertheless, the draft offered a base with concrete formulations from which to continue the negotiations. The analysis undertaken in this study clearly demonstrates that there are significant global, as well as national, income gains to be reaped from implementing the proposals in the Harbinson Draft. The simulation shows that world trade in agricultural commodities could increase by 25 per cent and the estimated real income gains constitute US$100 billion per year, which corresponds to twice the amount of Official Development Assistance being disbursed to the developing countries. Or, to put it differently, the simulation shows that undertaking trade liberalization in agriculture has the potential to create economic benefits in developed and developing countries that vastly exceed the existing transfers from rich to poor countries. Therefore, resources could be released that could be used to assist growth and development in the poorer regions of the world, while at the same time a freer and more transparent trading regime is established.

Notes 1 2

A longer version of this analysis can be found in FØI (2003). In Anderson (2004) a review of studies concerning trade liberalization is found.

References Anderson, K. (2004) ‘The challenge of reducing subsidies and trade barriers’. Policy, Research working paper series, no. WPS 3415 , World Bank, Washington DC. Brown, D., Deardorff, A.V. and Stern, R.M. (2003) Multilateral, regional and bilateral trade-policy options for the United States and Japan. The World Economy 26(6), 803–828. FØI (2003) Note on the Harbinson Draft in the WTO Agriculture Negotiations. Danish Research Institute of Food Economics, Frederiksberg, Denmark.

François, J.F., Van Meijl, H. and Van Tongeren, F. (2003) Trade Liberalization and Developing Countries Under the Doha Round. Timbergen Institute Discussion Papers
8

Food Security and the World Trade Organization: a Typology of Countries1 EUGENIO DIAZ-BONILLA,* MARCELLE THOMAS,† SHERMAN ROBINSON‡ AND ANDREA CATTANEO§ *Executive

Director for Argentina and Haiti, Inter-American Development Bank, 1300 New York Ave., NW, Room NE1137, Washington DC 20577, USA; †Markets, Trade, and Institutions, International Food Policy Research Institute, 2033 K Street, NW, Washington DC 20006-1002, USA; ‡Department of Economics, School of Social Sciences, University of Sussex, Falmer, Brighton, BN1 9SN, UK; §Economic Research Service, USDA, 1800 M Street NW Room S-4213, Washington DC 20036-5831, USA.

Introduction The Agreement on Agriculture (AoA) negotiated during the Uruguay Round of international trade negotiations stipulated in Article 20 the need to continue agricultural negotiations within the World Trade Organization (WTO). It also indicated that those negotiations should take into consideration, among other things, ‘non-trade concerns’, and the preamble to the AoA mentions food security as an example of those concerns. In the current agricultural negotiations the issue of food security has been raised by both industrialized and developing countries. Richer countries that are net food importers are preoccupied with maintaining an ‘adequate’ ratio between total domestic food production and the level of trade needed to satisfy food requirements at the national level. Developing countries’ concerns are different and focus on whether the current Agreement on Agriculture may help or hinder important policy objectives such as elimination of poverty and hunger (as cause and consequence of food insecurity), and whether further negotiations may improve on the existing text or further compromise the attainment of those objectives in poor countries. These various claims and circumstances suggest the need for differentiating among the approaches and status of countries in relation to food security, in general, and in the context of WTO negotiations, in particular. 162

© CAB International 2006. WTO Negotiations and Agricultural Trade Liberalization: the Effects of Developed Countries’ Policies on Developing Countries (eds E. Diaz-Bonilla et al.)

Food Security and the World Trade Organization

163

Currently, the main categories of countries considered by the WTO include: developed countries, developing countries and, within the latter, the Least Developed Countries (LDCs) and Net Food Importing Developing Countries (NFIDCs). No formal definition of the first two categories exists, and the selection process works through self-identification and negotiation with other members. Additionally, a country can be considered ‘developing’ under some WTO legal texts but not under others, depending on the negotiations among member countries. As of February 2005, the WTO has 148 members and 33 observers. More than 80 per cent of the members and 90 per cent of the observers are developing countries or republics of the former Soviet Union. The LDCs are defined by the General Assembly of the United Nations by three criteria including low income, human resource weakness and economic vulnerability measures (UNCTAD, 2001). Among the 50 LDCs, 32 have become WTO Members and 10 are WTO observers, 8 of which are in the process of accession. The 22 NFIDCs have been selected through a procedure that takes place in the Committee on Agriculture of the WTO: countries wanting to be considered in that category must present data showing that they are net foodimporting countries and the other WTO members may accept (or not) the petition based on that evidence. Within the WTO legal framework, all those categories have different legal implications. For the coming negotiations to consider in detail food security concerns under WTO rules, two issues need to be addressed. The first is the relevance of the current classification of countries with respect to their food security status. Of these categories, only the NFIDCs are defined with respect to a particular food security indicator (i.e. net food imports), although, as will be argued below, it may not be the most appropriate. The second issue is whether the current legal texts, which define WTO commitments on the basis of these categories of countries, really address the issue of food security through that differential treatment. Both questions are related: if the categories are badly defined to capture food security concerns, then it is unlikely that the differential treatment under WTO rules will deal with those concerns in a meaningful way. But even if these categories capture the variety in the situations of food (in)security, the question of the adequacy of current and future WTO rules and commitments to treat these differences must still be answered. This chapter contributes to the first issue of the adequate classification of countries, as an input to the second and separate discussion on the specific rights and obligations under the WTO and their implications for food security.2 A classification of countries is presented using various dimensions of food security and clustering methods. A methodological innovation is the application of the theory of ‘fuzzy sets’ in conjunction with more traditional methods of cluster analysis. The rest of the chapter is organized as follows. The next section presents a rationale for the selection of food security indicators considered in this analysis. The third section briefly describes three approaches to cluster analysis (hierarchical, k-means and fuzzy), and presents the results, including

164

E. Diaz-Bonilla et al.

a classification of countries and their food security profiles. The fourth section discusses the suggested typology of countries in detail and considers the variety in national situations. Finally, the last section concludes with some implications from the food security profiles identified in this study for a better definition of the trade rules in the current WTO negotiations.

Food Security Indicators at the National Level Food security can be analysed at the global, national, regional, household and individual levels. Since the World Food Conference of 1974, definitions of food security have moved from the global and national levels to the household and individual levels, where the problem of food (in)security takes a concrete human dimension (Maxwell, 1996). At the same time it was recognized that poverty and lack of income opportunities, rather than food supply per se, have been the main obstacles to access to food (Sen, 1981). The 1996 World Food Summit included several of those different components when it asserted that ‘food security exists when all people, at all times, have physical and economic access to sufficient, safe, and nutritious food to meet their dietary needs and food preferences for an active and healthy life’ (FAO, 1996). But availability and access are only preconditions for adequate utilization of food. Food availability and even access do not determine unequivocally the more substantive issue of malnutrition or nutrition insecurity at the individual level, where other factors such as health, women’s education and women’s relative status in the society appear central (Smith and Haddad, 2000). This chapter, acknowledging that the deeper issue of nutrition insecurity requires analyses at the household and individual levels, none the less takes a national perspective (the level at which the negotiating categories are defined within the WTO) and focuses mainly on food availability issues using consumption, production and trade measures (Table 8.1). The selection of the indicators has been guided by the need to be relevant to the analysis of categories within trade negotiations, and not because of other possible purposes.3 Calories per capita (CALCAP) and proteins per capita (PROTCAP) measure average consumption levels at the national level. While national averages have limitations as indicators of household and individual food and nutrition security, Smith and Haddad (2000) showed that aggregate calories (which they label food availability) are important in explaining changes in malnutrition as defined by anthropometrical measures of children. This cluster analysis uses indicators for both calories and proteins, thus improving upon a calories-only measure. Food production per capita (PRODCAP) is an indicator of the ability of countries to feed themselves through domestic production. It considers both the notions of insurance and national autonomy advocated by some developed countries, and the more pressing concerns of poverty and hunger in developing countries.

Food Security and the World Trade Organization

165

Table 8.1. Food security indicators. Indicator

Description

Units

Source

CALCAP PROTCAP PRODCAPa

Calories per capita per day Proteins per capita per day Food production per capita: total food production in every year multiplied by the 1989–1991 world price in US dollars divided by total population of the corresponding year Ratio of total exports, including services, to food imports

Calories Proteins US$

FAOSTAT 1999 FAOSTAT 1999 FAOSTAT 1999

Ratio

Share of non-agricultural population

Ratio

FAOSTAT (1999) and World Development Indicators, World Bank (2000) FAOSTAT (1999) and World Development Indicators, World Bank (2000)

EXPTOIMPb

NAGRPOP

Note: Each indicator value is the average of the last five years of available data, which for most of the countries is 1993–1997. a FAO’s definition of food includes cereals, oils, livestock products, as well as fruits, pulses, roots and tubers, other vegetables, cocoa and sugar. It captures better the combined contribution of calories, proteins and micronutrients than narrower definitions of food, particularly those based only on cereals. b This variable is usually measured as food imports over total exports. The inverse is used so that higher (lower) values of this variable indicate more (less) food security similarly to the other variables.

The ratio of total exports to food imports (EXPTOIMP) is an indicator of food access. It measures the ability of countries to finance their food imports out of total export revenues, which include merchandise and services, such as tourism. This variable, which has been used in early studies of food security (Valdés and Konandreas, 1981), is a better indicator of food security than the net food trade measure (i.e. food exports minus food imports) used to determine the category of NFIDC in the WTO. Whether a country is a food importer or exporter does not reflect how much of its export revenues it must allocate to access food imports, and consequently, how vulnerable it may be to changes in food prices and international food availability. This indicator also highlights the broader role of trade and the possible impact of trade negotiations on food security, which concerns not only the availability of food in world markets, but also the generation of export income to finance those imports. In this context the important issue is whether total exports have increased because of the negotiations by more than the food import bill.4 A country whose food import bill goes up due to the specific trade agreement considered may still be less vulnerable if at the same time its total exports increase by a larger amount.

166

E. Diaz-Bonilla et al.

The fifth indicator, the share of non-agricultural population (NAGRPOP), highlights the possible distributive impact between rural and urban populations from changes in trade and agricultural policies. Several developing countries have indicated their concern that further liberalization of agricultural and trade policies may create problems for their large agricultural populations, where poverty is still concentrated (WTO, 2000a,b). But at the same time, it is also important to note the shift in the locus of poverty, food insecurity and malnutrition from rural to urban areas that different developing countries are experiencing, some over several decades, others as a more recent phenomenon (Haddad et al., 1999; Garrett and Ruel, 2000). Among other issues, urbanization in developing countries is raising new questions on the impact of trade policies on food security. Trade protection for food products is equivalent to a tax on food consumption, with the proceeds of that tax transferred to food producers, while agricultural liberalization (if domestic markets operate adequately) should result in a reduction in the implicit tax burden for food consumers. Similar profiles of trade protection (or trade liberalization) will have different implications for developing countries with important contingents of urban poor affected by food insecurity, than for poor countries where a majority of the population affected by poverty and food insecurity live in rural area and work in agricultural production. So, while a higher value for the first four indicators (consumption per capita of calories and proteins, food production per capita and total exports per unit of food import) is associated with greater food security, the ratio of urban population is somewhat more ambiguous in its implications.

Cluster Analysis: Data and Methodology This analysis uses a data set of 167 countries including 137 WTO members (more than 90 per cent of the total) and 19 WTO observers (close to 60 per cent of that category), as well as 44 LDCs (90 per cent of the LDCs), and all 22 NFIDCs so far defined as such under WTO rules. Clustering methods are used to derive food security profiles for the 167 countries based on the five measures of food security described in Table 8.1 (the data set available in Diaz-Bonilla et al., 2000, Appendix II). The selection of the variables is crucial because the derived clusters would reflect only the structure of the data as defined by those variables. An implication is that two countries that belong to the same cluster are considered similar only with respect to the variables selected, but they may well be very different in terms of other variables not considered. Cluster analysis is a form of data dimensionality reduction, which compacts information from an entire population or sample into information about specific, smaller groups (Cherkassky and Mulier, 1998; Hair et al., 1998)5. Although cluster analysis can be characterized as descriptive, atheoretical and non-inferential, some statistical aspects must be addressed: whether the sample data represent the population, whether

Food Security and the World Trade Organization

167

multicollinearity exists and whether outliers can be identified (Hair et al., 1998). In this study, the sample is close to the whole population, so there is not an issue about how representative it is. The degree of multicollinearity between calories per capita and proteins per capita is higher than that between the other variables, giving more weight to the consumption indicator, a desirable condition given its relevance to food security (see Diaz-Bonilla et al., 2000). Finally, cluster analysis is very sensitive to the presence of outliers, which may result from either extreme values of some of the variables or a unique combination of them. In this application, those outliers are identified early in the analysis and treated separately. Two main issues in cluster analysis are how to form clusters, and how many clusters to form. Clustering methods can be classified into two general categories, hierarchical and nonhierarchical, depending on their algorithms to form clusters. One main difference is that in hierarchical methods, once an object is allocated to a cluster, it remains there for the whole exercise, while nonhierarchical methods allow for reclassification of objects as clusters are formed. All methods try to maximize the differences between clusters relative to the variation within the clusters as they are formed.6 The hierarchical algorithm is used first to determine the number of clusters. The method also helps to identify outliers and provides the corresponding cluster centres (country average value of the five indicators). The k-means, a non-hierarchical method, uses the cluster centres and the number of clusters, computed previously, to define the food security profiles for each group of countries. It improves on the hierarchical method because it allows countries to be rearranged between clusters as clusters are formed. Finally, the ‘fuzzy’ analysis helps resolve any remaining ambiguity between individual country profile and its cluster general structure. While the k-means algorithm is deterministic (i.e. countries are either in a cluster or they are not), the fuzzy algorithm allows degrees of membership in different clusters. Fuzzy cluster analysis incorporates what is called event ambiguity, a form of uncertainty different from well-defined, unambiguous events than can be random (Yen and Langari, 1999). Fuzzy analysis measures the degree to which an event occurs, not whether it occurs. In this exercise, each cluster can be viewed as an event category, and every country, depending on the value of its indicators, will be a member to a different degree (a value between 0 and 1) of each cluster. A country is classified in the cluster in which it has a dominant degree of membership. But, in some cases, countries may have significant degrees of membership in more than one cluster. The k-means method provides the distance, but not the direction, of each country’s variable from its cluster’s centre. Conceivably, two countries might have similar k-means indicators of distance, but one country could lean towards a more food-secure profile, while the other might be more similar to a food-insecure one. The fuzzy approach, on the other hand, by showing the degree of membership in the different clusters, clarifies these ambiguities.

168

E. Diaz-Bonilla et al.

In all three methods, variables are converted to z-scores (subtracting the mean and dividing by the standard deviation), to avoid giving more weight to any one variable because of its unit of measure. Because no statistical inference exists in clustering, the ‘correct’ number of clusters cannot be determined by objective criteria, although a number of approaches have been suggested. One of the most common procedures is to evaluate the changes in the agglomeration coefficient, a measure of homogeneity between countries of the same cluster, using a dendogram computed by the hierarchical algorithm.7 The number of clusters selected is 12. The analysis shows that a lower number of clusters would have resulted in large decreases of homogeneity among members of the same cluster, and that a higher number would not have improved homogeneity significantly within clusters, except among clusters of developed countries, which are not the focus of this study. The results also identify two outliers, Thailand and New Zealand (Diaz-Bonilla et al., 2000). All three methods classified 129 countries (78 per cent) in the same cluster, the remaining 36 countries have the same cluster membership in two out of three methods, and no country is classified differently by each of the three clustering methods. Of the 36 countries for which only two methods agreed, 21 (58 per cent) are classified similarly by the hierarchical and the k-means approaches, while the remaining 15 are classified similarly by the fuzzy and the k-means methods. Countries are allocated to one of the 12 clusters where at least two of the three methods classified them. The next section analyses in detail these 12 clusters and uses the results of the fuzzy analysis to help clarify two questions: first, for the countries not classified unanimously by all three methods, what is their level of ambiguity in membership, and the direction in which they are ambiguous? Second, for the countries in which the three methods agree, are there cases of ambiguity that may have implications for food security analysis (for instance, a food-neutral country with non-trivial membership in food-insecure clusters)?

Typology of Countries For each cluster identified by the cluster analysis, average z-scores of the indicators are computed, and the 12 clusters are sorted in ascending order by the average value of their indicators in the following three general categories of food security: ‘food-insecure’ clusters have most of the centres for the different variables falling below
0.5 (minus half a standard deviation from zero); Clusters 1 to 4 fall into that category (Fig. 8.1). ‘foodneutral’ clusters have centres falling between –0.5 and +0.5 (plus or minus half a standard deviation around zero); Clusters 5 to 8 fall into that category (Fig. 8.2). Finally, ‘food-secure’ clusters have centres with values above +0.5 and include Clusters 9 to 12 (Fig. 8.3). Figure 8.4 illustrates the relative position of the 12 clusters in a different diagram, where the average value of the z-score variables for the

Food Security and the World Trade Organization

169

1.0

Cluster mean

0.5 0.0 –0.5 –1.0 –1.5 –2.0 CALCAP

PROTCAP

PRODCAP

EXPTOIMP

NAGRPOP

Variables Cluster 1

Cluster 2

Cluster 3

Cluster 4

Fig. 8.1. Food-insecure groups.

1.5

Cluster mean

1.0 0.5 0.0 –0.5 –1.0 CALCAP

PROTCAP

PRODCAP

EXPTOIMP

NAGRPOP

Variables Cluster 5

Cluster 6

Cluster 7

Cluster 8

Fig. 8.2. Food-neutral groups.

combined consumption of calories and proteins, is plotted against the trade indicator showing the burden of the food bill (also in z-score values). The solid lines at the value of –0.5 across both axis of the chart divide the space into four main quadrants separating the food-insecure clusters from the rest (the dotted lines at the +0.5 value and other quadrants differentiating among clusters that are food-neutral or food-secure): Clusters 1 and 2 appear in the quadrant that is consumption-vulnerable and trade-stressed (south-west quadrant), with the values below –0.5 on both dimensions; Cluster 3 is in the quadrant which identifies consumption vulnerability but not trade stress (south-east quadrant);

170

E. Diaz-Bonilla et al.

4.0

Cluster mean

3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 CALCAP

PROTCAP

PRODCAP

EXPTOIMP

NAGRPOP

Variables Cluster 9

Cluster 10

Cluster 11

Cluster 12

Fig. 8.3. Food-secure groups.

11

1.5

12 10

9

avcalpro

1.0 7

0.5

8 0.0

6

4

5

–0.5 2

3

–1.0 1 –1.0

–0.5

0.0

0.5

1.0

1.5

2.0

xtlmfood

Note: In the above diagram the average value of the z-score variables for the consumption of calories and proteins (avcalpro), is plotted against the trade indicator showing the burden of the food bill (xtlmfood, also in z-score values). The solid lines at the values of
0.5 across both axes of the chart divide the space into four main quadrants separating the food-insecure clusters from the rest (the dotted lines at the +0.5 values add other quadrants differentiating among clusters that are food-neutral or food-secure). Fig. 8.4. Scatter plot of consumption per capita versus trade indicator.

Food Security and the World Trade Organization

171

Cluster 4 is in the trade-stressed quadrant but is above the level of –0.5 for consumption (north-west quadrant). The rest of the clusters appear in the intermediate or high levels of consumption and trade security (north-east quadrant), with both dimensions above the –0.5 value.

Food-insecure group Very food-insecure countries Countries in Cluster 1 have the lowest levels of consumption measured in calories (1983) and protein (49 g) per capita, and of food production per capita (US$82). Their food imports represent over 20 per cent of their total export earnings compared to the world-weighted average of 6 per cent (therefore they are considered ‘trade stressed’), and more than 75 per cent of their population is rural (Table 8.2). This group includes 30 countries, all of them LDCs except Kenya, a country classified as NFIDC within the WTO.8 Most of the countries in this group are from Africa (23 out of a total of 30). Twenty-three countries are WTO members and two are WTO observers (Table 8.3). Most countries have been included in the same cluster by all three methods. A few – Angola, Cambodia, Madagascar, Mali, Nepal and Uganda – have a dominant degree of membership in Cluster 3, which is also a foodinsecure cluster but has a lower burden of food imports (less ‘trade stressed’) than Cluster 1 (see Diaz-Bonilla et al., 2000, Appendix IV, showing the values for the fuzzy membership of all countries).9 Food-insecure countries with an urban profile Although Cluster 2 shows higher levels of consumption and production than Cluster 1, it is also ‘consumption vulnerable’ and trade stressed. But these countries are far less rural than those in other food-vulnerable clusters, with more than 70 per cent of the population classified as urban (Table 8.2). This raises the issue of urban food insecurity, which has its own special characteristics (see Garret and Ruel, 2000). While countries in the previous cluster may be more concerned about food insecurity in the countryside and the impact of agricultural imports on poor agricultural producers, in countries with larger urban populations, and where conceivably an important percentage of poor and food-insecure groups may be urban dwellers, policies aimed at agricultural trade protection have a clear trade-off: they may maintain higher incomes for poor producers, but they may also act as a tax on poor urban consumers (both effects depending on other policies and on the interaction of markets and institutions).10 Among the 14 members of this cluster, two are LDCs and five are NFIDCs. Most members are Latin American countries or republics of the ex-Soviet Union. Except for Tajikistan, all of the countries are either WTO members (11) or observers (two) (Table 8.3).

172

E. Diaz-Bonilla et al.

Table 8.2. Final cluster means. (From author’s calculations.)

Cluster 1 Cluster 2 Cluster 3 Cluster 4 Cluster 5 Cluster 6 Cluster 7 Cluster 8 Cluster 9 Cluster10 Cluster11 Cluster12

CALCAP (calories)

PROTCAP (g)

PRODCAP (US$)

EXPTOIMP ratio

Share of food imports to total exports (per cent)

1982.9 2229.2 2244.6 2581.5 2602.3 2672.9 2976.1 2827.7 3231.3 3271.8 3303.7 3374.1

48.6 58.8 52.6 70.8 66.5 72.8 82.7 78.4 100.1 97.7 103.3 107.5

81.8 117.6 120.3 157.2 210.4 124.1 135.1 233.3 254.2 304.2 520.6 923.9

4.9 5.3 14.1 4.8 11.3 19.8 9.1 25.6 18.6 35.9 17.7 32.7

20.4 19.0 7.1 20.8 8.8 5.0 11.0 3.9 5.4 2.8 5.7 3.1

NAGRPOP ratio 0.23 0.71 0.41 0.39 0.75 0.41 0.82 0.83 0.88 0.93 0.93 0.93

Note: Share of food import to total exports is the inverse of the indicator EXPTOIMP.

This cluster shows substantial convergence between the different clustering methods: but for three countries – Botswana, the Dominican Republic and Mongolia, fuzzy clustering shows dominant membership in other clusters. For Botswana and Mongolia the accumulated membership in food-insecure clusters is dominant, so no ambiguity follows. The Dominican Republic, on the other hand, has more than 40 per cent membership in food-neutral Clusters 5 and 6 as against 43 per cent in food insecure Cluster 2 (Diaz-Bonilla et al., 2000, Appendix IV). One reason for this ambiguity is that the Dominican Republic is the least trade-stressed country in Cluster 2, with a food bill of about 7 per cent of total exports (close to the average for the world and for developing countries). The Dominican Republic is considered a NFIDC within WTO, but some of its food imports are linked to an expanding tourism industry, and may not necessarily reflect food security concerns. Still, this country has a total degree of membership in food-insecure clusters of 54 per cent, and, therefore, its classification is maintained. Food-insecure countries with consumption vulnerability Cluster 3 shares low consumption and production levels with Clusters 1 and 2 (2245 calories and 53 g of protein per capita per day), and a large rural population with Cluster 4 (around 60 per cent). But the burden of the food bill, which is around 7 per cent of total exports, is within intermediate levels (Table 8.2). This cluster can be characterized as ‘consumption vulnerable’ but ‘trade neutral’ (Fig. 8.1). Four of the 17 countries in Cluster 3 are LDCs and two are NFIDCs. All are members or observers of the WTO and include eight developing countries from Africa, seven from Asia and the Pacific and

Food Security and the World Trade Organization

Table 8.3. Country membership in Clusters 1 to 12. Cluster, no. of countries and food group

WTO status

LDC

NFIDC

1 (30) FI

WTO members

Angola, Bangladesh, Burkina Faso, Burundi, Cambodia, Central African Republic, Chad, Dem. Republic of Congo, Gambia*, Guinea, Guinea-Bissau, Haiti, Madagascar, Malawi, Mali, Mozambique, Nepal, Niger, Rwanda, Sierra Leone, United Rep. of Tanzania, Uganda Ethiopia, Yemen Afghanistan, Comoros, Eritrea, Liberia, Somalia

Kenya (Gambia*)

Djibouti, Lesotho

Botswana, Cuba, Dominican Republic, Honduras, Peru

WTO observers Others 2 (14) FI

WTO members

WTO observers Others 3 (17) FI

4 (13) FI

Others

El Salvador, Georgia, Mongolia, Nicaragua Armenia, Azerbaijan Tajikistan

Solomon Islands, Togo, Zambia

WTO observers Others

Laos

WTO members

Benin, Mauritania, Senegal

WTO observers Others

Sudan, Vanuatu Kiribati

Côte d’Ivoire, Sri Lanka

Bolivia, Cameroon, Republic of Congo, Ghana, Guatemala, India, Namibia, Papua New Guinea, Philippines, Zimbabwe Vietnam

Pakistan, Saint Lucia

Albania, Grenada, Saint Kitts and Nevis, Saint Vincent/Grenadines Seychelles Continued

173

WTO members

174

Table 8.3. Continued. Country membership in Clusters 1 to 12. Cluster, no. of countries and food group

WTO status

5 (18) FN

WTO members

LDC

NFIDC

Others

Jamaica, Trinidad and Tobago, Venezuela

Belize, Brazil, Colombia, Costa Rica, Ecuador, Fiji Islands, Guyana, Kyrgyzstan, Nigeria, Paraguay, Surinam, Swaziland Croatia, Macedonia (former Yug. Rep.), Uzbekistan

WTO observers Others 6 (5) FN

WTO members

Myanmar

Antigua and Barbuda, Gabon, Indonesia China

WTO observers Others 7 (14) FN

WTO members

Maldives

WTO observers

Cape Verde

Others 8 (9) FN

WTO members

Barbados, Dominica, Egypt, Jordan, Mauritius, Morocco, Tunisia

Brunei Darussalam, Estonia, Kuwait, Macau, Mexico Algeria, Lebanon, Russian Federation, Saudi Arabia Bahamas, Islamic Rep. of Iran, Libyan Arab Jamahiriya, Syrian Arab Republic

WTO observers Others 9 (16) FS

WTO members

Czech Republic, Germany, Iceland, Israel, Japan, Lithuania, Malta, Poland, Portugal, Romania, Slovenia, Turkey, United Arab Emirates, UK

E. Diaz-Bonilla et al.

Bulgaria, Chile, Republic of Korea, Latvia, Malaysia, Republic of Moldova, Panama, Slovakia, South Africa

Belarus, Kazakhstan

WTO members

Austria, China–Hong Kong SAR, Finland, Hungary, Norway, Sweden, Switzerland, USA Ukraine

WTO observers Others 11 (9) FS

WTO members

12 (3) FS

WTO observers Others WTO members WTO observers Others

Outliers (2) WTO members FS

Argentina, Belgium–Luxembourg, Canada, France, Greece, Italy, Netherlands, Spain, Uruguay

Australia, Denmark, Ireland

New Zealand, Thailand

Food Security and the World Trade Organization

10 (9) FS

WTO observers Others

* Gambia is both a LDC and a NFIDC; FI – food insecure; FN – food neutral; FS – food secure.

175

176

E. Diaz-Bonilla et al.

two from Latin America (Table 8.3). Three countries from the Cairns Group (Bolivia, Guatemala and the Philippines) appear in this group. The majority of countries in this cluster are classified differently by at least two methods. For Bolivia, Côte d’Ivoire, Ghana, Guatemala, Papua New Guinea, Solomon Islands and Sri Lanka, the ambiguity in classification is within the food-insecure clusters and therefore does not lead to an ambiguity in their overall food-insecure profile (Diaz-Bonilla et al., 2000, Appendix IV). In the case of India, Namibia, the Philippines and Vietnam the fuzzy method gives them dominant membership in Cluster 6, a food-neutral cluster. Except for the Philippines, these countries are all net food exporters and the incidence of the food bill on total exports (‘trade stress’) is low for all of them: 4.5 per cent for India, 5.3 per cent for Vietnam and about 6 per cent for Namibia and the Philippines. These countries exemplify a possible policy dilemma: because they are not trade stressed, they could expand food imports to improve their low levels of consumption (if food availability were the problem); but, at the same time, they would be concerned about the impact of additional food imports on their large poor agricultural populations. All in all, two out of three methods classify them as food insecure (in this case the hierarchical and the k-means), and all four countries are retained in the food insecure Cluster 3. Food-insecure countries with trade stress Cluster 4 has an opposite profile to Cluster 3: it shows higher consumption levels (2581 calories and 71 grams of protein) but also a heavier trade burden with a food bill of almost 21 per cent of total exports (Table 8.2). Cluster 4 has 13 members, including six LDCs and two NFIDCs.11 All countries except one, Kiribati, are WTO members or observers (Table 8.3). Although the inclusion of some bigger countries in this group, such as Pakistan, Sudan and Senegal, conforms to the notion of having intermediate consumption yet being trade stressed, the classification of some small islands from the Caribbean and the Pacific in this group is less clear and has to be approached with caution: some of them show important membership (from the fuzzy clustering results) in the food neutral Clusters 5 to 7 (DiazBonilla et al., 2000, Appendix IV); agricultural production data may not be reliable and do not include fisheries, which for some of those islands may represent an important component of production; the tourism industry has an impact on the external food balance and the receipts from tourism services may not be properly reflected in balance of payment accounts; finally, the distinction between rural and urban population, which the fifth indicator measures, may not be relevant in a small island. Of the countries classified by only two methods, Albania shows the greatest levels of ambiguity. The fuzzy method places this country outside its general group of food-insecure clusters, in Cluster 7, a food-neutral group (Diaz-Bonilla et al., 2000, Appendix IV). Albania’s profile combines relatively higher levels of consumption of calories and protein than the

Food Security and the World Trade Organization

177

average for Cluster 4, with a substantial level of trade stress (see country data in Diaz-Bonilla et al., 2000, Appendix II). That profile is similar to Cluster 7, where higher consumption is combined with borderline tradestress value. Still, Albania’s high trade vulnerability justifies its final classification in the food insecure Cluster 4.

Food-neutral group Indicators of food neutral Clusters 5 to 8 have z-scores between the
0.5 to +0.5 range, with some values above +0.5: consumption and urban population in Cluster 7, and trade ratios in Clusters 6 and 8 (Fig. 8.2). All clusters in this group are urban, except Cluster 6, which includes China. All of them show levels of consumption of calories and protein, and of production per capita, above those of Clusters 1, 2 and 3. The range is from 2600 calories and 66 grams of protein in Cluster 5 to 2976 calories and 83 grams in Cluster 7. They are also less trade stressed than Clusters 1, 2 and 4, particularly Clusters 6 and 8, which have a food bill at or below 5 per cent of total exports (Table 8.2). Average food neutral countries Cluster 5 has average values for the five variables close to the 0 level, the mean of the z-scores (Fig. 8.2). It includes three NFIDCs, and five countries in this cluster are members of the Cairns Group (Brazil, Colombia, Costa Rica, Fiji and Paraguay). All countries are members or observers of the WTO (Table 8.3). All 18 countries in this cluster are classified identically by all three methods, with the exceptions of Ecuador and Trinidad and Tobago. Although classified in different clusters by one of the methods, these last two countries remain within the general group of food-neutral countries. Although Kyrgyzstan, Swaziland and Uzbekistan have some degree of membership in food-insecure groups, they are classified unanimously by all three methods, and show more than 60 per cent membership in foodneutral clusters. The composition of Cluster 5 appears to reflect correctly intermediate levels of food security. Rural and trade-secure food-neutral countries Cluster 6 has levels of consumption of calories and protein slightly above average, is not trade stressed and is rural. All the countries in this group are WTO members. China is in Cluster 6. Myanmar is the only LDC (Table 8.3) and Indonesia is a member of the Cairns Group. Fuzzy clustering classifies India as a food-neutral country in Cluster 6, while the other two methods include India in Cluster 3, a food-insecure cluster (Diaz-Bonilla et al., 2000, Appendix IV). India and China both have a large proportion of rural population and low food bills relative to total

178

E. Diaz-Bonilla et al.

exports, but India’s consumption and production indicators are similar to food-insecure groups, while China shows food-neutral profile. So India remains in Cluster 3. For the small number of remaining countries (Antigua and Barbuda, Gabon, Indonesia and Myanmar), the fuzzy analysis confirms their inclusion in a food-neutral cluster. Food-neutral countries with high consumption and trade stress Cluster 7 includes 22 countries, of which two are LDCs and five are NFIDCs. Most of the countries are either developing or transition economies (Table 8.3). The World Bank considers four countries to be in the high-income category: the Bahamas, Brunei, Kuwait and Macau. Relative to other foodneutral clusters, Cluster 7 combines high consumption levels similar to foodsecure clusters, but also a heavy incidence of the food trade bill. The level of trade stress is an issue for some of the countries in this cluster. Although the food import bill for the group, on average, is around 11 per cent of total exports, some countries have high to very high values of this indicator. This is particularly so in the cases of Cape Verde and Maldives (both LDCs), of Lebanon (neither LDC nor NFIDC), but also of Egypt (an NFIDC), Dominica (Commonwealth of), Jordan and Algeria. With these levels of trade stress, should these countries be considered food insecure rather than food neutral? Given their high levels of consumption of calories and protein in some cases comparable to foodsecure countries, and that they are less rural – those trade-stressed countries are classified by the clustering algorithms in Cluster 7 because the grouping depends on the combined variables: clearly, if two groups of countries have similar levels of trade stress, the group with middle to lower consumption will be more vulnerable than the group with higher levels of consumption. Among those countries with high trade stress, Cape Verde and Maldives, the two LDCs countries, also have the largest membership in food-insecure clusters (about 15 and 35 per cent, respectively). Urban and trade-secure food-neutral countries Finally, among the food-neutral clusters, Cluster 8 is the most food secure, with higher levels for all indicators than Clusters 5, 6 and 7. In Cluster 8, all nine countries are members of the WTO. Chile, Malaysia and South Africa belong to the Cairns Group. Thailand, which is treated as an outlier because of the lowest food import bill (less than 2 per cent) – but otherwise would be in Cluster 8, is also a member of the Cairns Group. Food-secure group12 Finally, Clusters 9 to 12 are food secure, with most of the variables’ z-scores above the +0.5 value, that is, consumption of calories and protein well in excess of 3200 and 97 grams, production per capita above US$254, food

Food Security and the World Trade Organization

179

import bills between 3 and 6 per cent of total exports (i.e. these countries are trade secure), and above 88 per cent of total population is urban (Fig. 8.3). The main differences among them are the levels of production per capita, which range from US$254 for Cluster 9 to US$924 for Cluster 12 (Table 8.2). These groups have levels of consumption and production – as well as a trade ratio for food imports – that seem to provide more than enough margin to achieve food security under any likely event, domestic or international. Those clusters are labelled food-secure countries with intermediate production and trade indicators (Cluster 9); food-secure countries with intermediate production (Cluster 10); food-secure countries with intermediate trade indicators (Cluster 11); and very food-secure countries (Cluster 12). All industrialized countries (considered in the category of high-income OECD countries by the World Bank) fall into food-secure clusters, as well as some developing countries and former socialist countries, labelled middle income by the World Bank. All European Union (EU) members are in foodsecure clusters, as well as all the applicants for future membership, except for Bulgaria, Latvia, Slovakia (all in Cluster 8, the more food-secure of the food-neutral clusters) and Estonia, which is in Cluster 7. Two industrialized members, Australia and Canada, and two developing countries, Argentina and Uruguay, belong to the Cairns Group. New Zealand, which would be classified in Cluster 12 but was identified as an outlier because of its very high level of production per capita (US$1589), is also a member of the Cairns Group. Considering Cluster 12 and New Zealand, the four very food-secure countries are divided equally between the Cairns Group and the European Union (Table 8.3). These four food-secure clusters appear to be very robust in membership and profiles across clustering methods.

Conclusions The results have implications for the two issues identified in the introduction: first, the usefulness of the categories currently utilized in the WTO for discussion of food security concerns; and second, the relationship between the definition of appropriate grouping of food-(in)secure countries and WTO commitments, both current and future. As mentioned, this chapter concentrates on the first issue, while some implications of the cluster analysis for the legal obligations in the WTO are discussed elsewhere (Diaz-Bonilla et al., 2002b). The main conclusion is that some of the categories used by the WTO appear inadequate for capturing food security concerns. Consider for instance the category of ‘developing countries’. Concerns about the wide variety of countries that are self-identified as developing countries with special treatment have existed for some time in the General Agreement on Tariffs and Trade (GATT,) and now in the WTO. In the specific of food security developing countries appear scattered across all levels of food (in)security, except in the highest food-secure Cluster 12.

180

E. Diaz-Bonilla et al.

Another category with a weak correlation with food security indicators is that of NFIDCs: excluding Gambia (that as an LDC fits the profile of a food-insecure country; see below) only 11 of the remaining 21 countries in this group are classified in food-insecure clusters (or 12 if Egypt is considered here because of the high food import to total export ratio). Being a net food importer appears to be only a weak indicator of food vulnerability. Venezuela, for instance, a net food importer that is also a large oil exporter, has a food bill of around 5 per cent of total exports, below the average for developing countries. Another NFIDC with important levels of oil exports is Trinidad and Tobago. In both cases high levels of food imports reflect only the comparative advantage of their production structure. Additionally, some countries may be net food importers just because of a dominant tourist industry (like Barbados, which has an income of US$7000 per capita, the highest of all NFIDCs). In any case, the seven NFIDCs considered here in the food-neutral group (excluding Egypt) have food imports that represent about 9 per cent of total exports, while for the foodinsecure NFIDCs (including Egypt) the average is above 16 per cent. However, although being a NFIDC may not be a good indicator of serious food security problems, it does not mean that this category of countries should be dismissed. This classification, negotiated during the Uruguay Round, has implications under the Ministerial Decision on LDCs and NFIDCs, and constitutes an acquired right.13 The current membership in NFIDCs does not have to be changed and it certainly remains valid for those goals of the Ministerial Decision separate from food security considerations. But addressing the latter concerns requires a more precise approach based on specific indicators, such as the ones suggested here. The category of LDCs, on the other hand, corresponds most to countries suffering from food insecurity, even though this issue is not explicit in its definition. Only three out of the 44 LDCs covered in this study are not among the food insecure Clusters 1 to 4: Cape Verde, Maldives and Myanmar. These countries have incomes per capita four to six times higher than the LDCs’ average of US$235 (UNCTAD, 2000). Still, Cape Verde and Maldives are two of the most trade-stressed countries in Cluster 7, and with a non-trivial (although not dominant) membership in food-insecure clusters. This analysis would leave only Myanmar out of the countries with clear food security problems. On the other hand, countries such as El Salvador, Georgia, Mongolia and Nicaragua (all WTO members), which are neither LDC nor NFIDC, have a food security profile similar to the more vulnerable ones. Limiting the special and differential treatment related to food security problems to only LDCs or NFIDCs would leave them out. For the WTO negotiations, the analysis presented here suggests that to define specific rights and obligations in the WTO using the category of LDCs appears an appropriate starting point, even though food security issues are not part of the criteria for the definition of LDC. But it may not be enough. A possible approach would be to consider for special treatment under food security provisions both LDCs – as defined by the United

Food Security and the World Trade Organization

181

Nations – plus all other countries classified as food insecure according to some objective indicators, such as those used here. Without having to resort to formal cluster analysis, a more limited approach would be to use the consumption of calories and protein per capita as indicators of consumption vulnerability, and the food import bill as a percentage of all exports (merchandise and services) as an indicator of trade stress, in order to identify the countries most at risk. Values of the indicators, computed from an average of the last three or five years, would be those yielding zscores below
0.5 (around 2380 calories and 62 g of protein per day per capita, and about 13 per cent for the food import bill over total exports). Countries would move in and out of the food-insecure category so defined, depending on their performance. Those food-insecure countries would receive a treatment similar to LDCs for rights and obligations related to domestic support and their own market access. Also, they would be considered for the food aid, financial support, and technical assistance envisaged in the Ministerial Decision on possible negative effects of the agricultural reform programme on LDCs and NFIDCs. The issue of special access to other countries’ markets for LDCs, and the additional benefits conferred upon LDCs because of reasons other than food security, would still be limited to only those countries specified by the United Nations. The quantitative limits suggested would help differentiate developing countries that might need special treatment in terms of food security from those that do not. It is also relevant to ask about the food security situation of the developed countries. Several developed countries have advanced the notion of food security as part of the ‘multifunctionality’ of agriculture, or, more generally, among non-trade concerns (Norway RMA, 1998). Our typology, however, confirms the common-sense perception that all developed countries are food secure. The term ‘food security’ appears to have a very different meaning in developed and developing countries. For policy implications and the agricultural negotiations, maintaining the same label for two altogether different situations only obscures the issues being analysed. The discussion of food security should be limited to the vulnerability of developing countries.

1

This chapter is an updated and shortened version of the paper by Diaz-Bonilla et al. (2000). 2 Several of the issues on the legal obligations in the WTO are discussed in DiazBonilla et al. (2002a,b). 3 Diaz-Bonilla et al. (2000), present a more detailed discussion of the indicators selected, and of others also considered but discarded. For instance, the fact that both developing and developed countries have claimed food security as a non-trade concern would have made inadequate the use of GDP per capita to discuss the trade implications of different food security classifications because it would have prejudged some of the issues being negotiated (i.e. whether developed countries can claim food security as a non-trade concern). By the same token, variables reflecting domestic

182

E. Diaz-Bonilla et al. distributional patterns (as opposed to the national aggregates utilized here), although they are crucial for the design of domestic programmes, are less relevant to analyse trade negotiation issues: trading partners would find awkward a country’s claim to special status under WTO simply because it has a very unjust income distribution, or because it discriminates against some internal groups. 4 While in the early 1960s food imports for developing countries represented around 20 per cent of total exports, that percentage had declined to around 6 per cent by the end of the 1990s. This decline was due to the fact that, although food imports of developing countries increased significantly, total exports grew even faster (Diaz-Bonilla et al., 2000). 5 Clustering is a more rigorous way of finding groupings, to the extent that it must follow the disciplines of defining the quantifiable variables and of establishing a metric to determine closeness or proximity across groups in a verifiable way. ‘Table classification’, although based on defined and quantifiable variables, is usually based on arbitrary cut-off values to form groups. ‘Expert judging’ is even weaker because this tends to be based on an unknown (or unobservable) set of variables that may be applied differently to different cases. 6 A more detailed description of the clustering methods, and how they are applied to this exercise, can be found in Diaz-Bonilla et al. (2000), Appendix I. The hierarchical and k-means methods were run using SPSS; the fuzzy method was programmed in the General Algebraic Modelling System, GAMS (see Brooke et al., 1998), by Cattaneo following Ross (1995). 7 A dendrogram is a chart that provides a graphical view of the agglomeration process and shows the increase of the agglomeration coefficient, at each level of combination of different clusters. At the start, when each country belongs to a separate cluster, the value of the coefficient is zero, and it increases as countries are combined together in a smaller number of clusters (See Appendix III in Diaz-Bonilla et al., 2000). 8 Gambia is an LDC but also applied and was recognized as NFIDC. 9 This paper and the appendices can be obtained from the authors. 10 The case of vulnerable rural groups that are net consumers of food must also be considered. 11 In 2001, Senegal was classified as an LDC, and is counted as such in this classification instead of a NFIDC. 12 Diaz-Bonilla et al. (2000) discuss in greater detail the food-secure groups. 13 This issue was the subject of a special Ministerial Decision agreed during the Uruguay Round negotiations. It is called the ‘Decision on Measures Concerning the Possible Negative Effects of the Reform Program on Least-developed and Net FoodImporting Developing Countries’ (GATT, 1994, 448–449).

References Brooke, A., Kendrick, D., Meeraus, A. and Raman, R. (1998) GAMS – a user’s guide. The Scientific Press, San Francisco, California. Cherkassky, V. and Mulier, F. (1998) Learning from data: concepts, theory, and methods. John Wiley and Sons, New York. Diaz-Bonilla, E., Thomas, M., Robinson, S. and Cattaneo, A. (2000) Food security and

trade negotiations in the World Trade Organization: a cluster analysis of country groups. Trade and Macroeconomics discussion paper 59. International Food Research Institute, Washington DC. Diaz-Bonilla, E., Thomas, M. and Robinson, S. (2002a) On boxes, contents, and users: Food security and the WTO negotiations. Trade and Macroeconomics division paper

Food Security and the World Trade Organization 82. International Food Policy Research Institute, Washington DC. Diaz-Bonilla, E., Robinson, S., Thomas, M. and Yanoma, Y. (2002b) WTO, Agriculture, and developing countries: a survey of issues. Trade and Macroeconomics division paper 81. International Food Policy Research Institute, Washington DC. FAO (Food and Agriculture Organization of the United Nations) (1996) Rome declaration on world food security and World food summit plan of action. World Food Summit, Rome, Italy: Food and Agriculture Organization of the United Nations. FAOSTAT (1999) Database of the Food and Agricultural Organization. . Garrett, J.L. and Ruel, M. (2000) Achieving urban food and nutrition security in the developing world. Focus 3: A 2020 vision for food, agriculture, and the environment. International Food Policy Research Institute, Washington DC. GATT (General Agreement on Tariffs and Trade) (1994) The results of the Uruguay Round of multilateral trade negotiations: the legal texts. Geneva, Switzerland. Haddad, L., Ruel, M. and Garrett, J.L. (1999) Are urban poverty and undernutrition growing? Some newly assembled evidence. Food Consumption and Nutrition discussion paper 63. International Food Policy Research Institute, Washington DC. Hair, J.F., Anderson, R.E., Tatham, R.K. and Balck, W.C. (1998) Multivariate data analysis, 5th edn. Prentice Hall, Upper Saddle River, New Jersey. Maxwell, S. (1996) Food security: a postmodern perspective. Food Policy 21(6), 155–170. Norway Royal Ministry of Agriculture (1998) Non-trade Concerns in a Multifunctional Agriculture – Implications for Agricultural Policy and the Multilateral Trading System. Paper presented by Norway to the WTO. http://www.landbruk.dep.no/landbruks departementet/multifunctionality/assets/im ages/NTC_paper.doc Ross, T.J. (1995) Fuzzy logic with engineering applications. International edition, McGraw-Hill.

183 Sen, A. (1981) Poverty and famines: an essay on entitlement and deprivation. Clarendon Press, Oxford, UK. Smith, L.C. and Haddad, L. (2000) Explaining child malnutrition in developing countries: a cross-country analysis. IFPRI Research Report 111. International Food Policy Research Institute, Washington DC. UNCTAD (United Nations Conference on Trade and Development) (2000) Impact of the reform process in agriculture on LDCs and Net Food-Importing Developing Countries and ways to address their concerns in multilateral trade negotiations. Background note by the UNCTAD secretariat. TD/B/COM.1/EM.11/2. Geneva, Switzerland. UNCTAD (United Nations Conference on Trade and Development) (2001) Statistical profiles of the Least Developed Countries. Prepared by the UNCTAD secretariat. Geneva, Switzerland. Valdés, A. and Konandreas, P. (1981) Assessing food insecurity based on national aggregates in Developing Countries. In Valdés, A. (ed.) Food Security for Developing Countries. Westview Press, Boulder, Colorado. World Bank (2000) World Development Indicators. On CD-ROM. Washington DC. WTO (World Trade Organization) (2000a) Agreement on Agriculture: Special and Differential Treatment and a Development Box. Proposal to the June 2000 special session of the committee on agriculture by Cuba, The Dominican Republic, Honduras, Pakistan, Haiti, Nicaragua, Kenya, Uganda, Zimbabwe, Sri Lanka and El Salvador. G/AG/NG/W/13. WTO (World Trade Organization) (2000b) Agreement on Agriculture: Green Box/Annex 2 Subsidies. Proposal to the June 2000 special session of the committee on agriculture by Cuba, the Dominican Republic, Honduras, Pakistan, Haiti, Nicaragua, Kenya, Uganda, Zimbabwe, Sri Lanka and El Salvador. G/AG/NG/W/14. Yen, J. and Langari, R. (1999) Fuzzy logic: intelligence, control, and information. Prentice Hall, Upper Saddle River, New Jersey.

9

A Proposal for Combating Acute Food Shortages Based on Sub-Saharan African Needs KIM M. LIND Food and Research Economic Institute, Agricultural Policy Research Division, Rolighedsvej 25, 1958 Frederiksberg C, Denmark

Introduction Storage of food for use in times of need has taken a more prominent position in the international food policy debate. The motivation for this debate is the increasing liberalization of the world economy and, in particular, liberalization of agricultural policies. It is argued that this trend, linked in part to past and ongoing WTO negotiations, is leading to declining levels of food stocks, thereby jeopardizing food security. Previously, due to agricultural policies, mostly in industrialized countries, vast amounts of food were stored as a result of political decisions to keep farmers’ income at desired levels. Consequently, after the reform of some of these policies, the use of stocks of various agricultural products generated by public interventions in times of low prices has decreased. Consequently, FAO has raised some concerns that the levels of food stock under a regime of liberal agricultural policies may not be sufficient to cover the needs in the case of a major disaster (FAO, 1995). This chapter considers food stocks for security reasons as an insurance against shortages due to bad harvest, wars or other disasters. Thus, the objective in this chapter is more narrow in focus than were previous studies of policies to counter food shortages. A vast literature exists on this topic. However, the focus has primarily been on general undernourishment, and thus, on ensuring an adequate supply of food for the population at large also in ‘normal’ times (see Konandreas et al., 1978; Huddleston et al., 1984; Ravaillon, 1996; Zeller et al., 1997; and Sarris, 2000). In this chapter, only disastrous events are considered for which immediate action must be taken in order to prevent widespread famine – with accompanying disease and death. The important problem of securing adequate food access for the millions of people suffering from undernourishment is not considered. One 184

© CAB International 2006. WTO Negotiations and Agricultural Trade Liberalization: the Effects of Developed Countries’ Policies on Developing Countries (eds E. Diaz-Bonilla et al.)

Proposal for Combating Acute Food Shortages

185

purpose of the chapter is to derive an ‘optimal’ level of food stocks based on the uncertainty of food supply and demand, and the objective of keeping that level of stocks. More specifically, the chapter aims at deriving distributions describing the probability of critical food shortages and the scale of the shortage. Although the focus of stocks for food security purposes is of global relevance, the focus of the paper is on sub-Saharan Africa. The developed countries are food secure, so they do not need to keep stocks for security purposes. The least-developed countries, on the other hand, are food insecure and Sub-Saharan Africa constitutes a sizeable part of the latter group (see Diaz-Bonilla et al., Chapter 8, this volume). In this chapter, instead of using the observed levels of stocks which, since World War II, have been greatly affected by the prevailing protectionist agricultural policies, and rather than looking at indirect measures of nutritional status, the basis for the analysis is the observed levels of calorie intake. In contrast to the level of stocks that may go up or down for a variety of reasons, calorie intake gives a more direct measure of food shortage and, consequently, the potential need for reserve stocks. In relation to food intake other measures could be relevant, such as the amounts of protein and various nutrients. However, these other measures are more related to the longer-term well-being of the individual, and are concerned with ensuring access to a steady, sufficient flow of proper diets, rather than coping with unexpected emergency situations. In the latter case, which is the focus of this chapter, the important objective is to ensure an adequate supply of calories. Based on that statistical analysis for sub-Saharan Africa, the chapter explores in the final section the possibility of a less expensive alternative to keeping stocks for security purposes, based on a financial fund set-up with the sole purpose of purchasing grain on the open market when acute food shortages occur.

Background Storage of food against hunger is an age-old policy. The first instance recorded in history is that of Joseph’s recommendation to the Egyptian pharaoh based on the pharaoh’s dream around 2000 BC (Genesis, 41). According to the story, cereals were stored during 7 years with good harvests, to be drawn upon during the following 7 years with bad harvests. In principle this was not an insurance scheme, but a case of consumption smoothing over time. To be an insurance scheme, some degree of uncertainty has to play a role, but in this case the outcomes (bad harvests, good harvests) were known in advance through revelation. Storage of food as consumption smoothing must have been even older as people need to store food for consumption between harvests or hunting/fishing seasons. Since the time of Joseph, ensuring an adequate supply of food has been a prime concern of political leaders. As the sentence panem et circenses (bread and circuses) by the Roman poet Juvenal around 100 AD shows, the

186

K.M. Lind

leaders of Rome were perfectly aware of the correlation between ensuring an adequate supply of food and political power. Now, again skipping something like 2000 years, the Common Agricultural Policy (CAP) of the European Union has been established, among other concerns, to secure an adequate supply of food within the union. At the time of the enactment of the CAP the degree of self-sufficiency of food products, at around 85 per cent, was considered too low. Thus, the CAP was designed with the primary aim of increasing the EU’s supply of food, which has subsequently led to big surpluses. Around 1970, the previously prevailing optimism about food conditions – and availability of natural resources in general – were being questioned. The old hypothesis of Malthus (1798) – that the growth of population would outstrip the growth of food production – was given new life, in particular, by the report of the Club of Rome (1972). This led to the World Food Conference in 1974. During this time there were unprecedented increases in world food prices as a result of large and unexpected purchases of grain from the Soviet Union. Simultaneously, climatic changes resulting from ‘El Niño’ negatively affected fish stocks in the Pacific region of Latin America, leading to substantial declines in the world’s protein supply. Added to that, bad harvests plagued soybean producers in the USA. The result was a substantial increase in the real price of food during a substantial part of the 1970s. One of the major results of the 1974 World Food Conference was to put the issue of food security on the international agenda. The International Fund for Agricultural Development (IFAD) was established as a result of the conference. Another result was the creation of the Committee on World Food Security in 1975. This committee has the ‘function of evaluating the adequacy of food stocks, especially cereals’ (FAO, 1983). To achieve the objectives of the committee, an estimate of a minimum safe level of world cereal stocks was carried out by the FAO secretariat in 1974. Three methods were employed: ●

●

●

The longer-term trends in area, yields and domestic production and consumption of cereals in exporting and net importing countries during 1955/56 to 1972/73 were analysed to measure deviations from the trends in these variables; this analysis provided the range of deviations observed in the past to serve as a possible indication of reserves needed to maintain the long-term trend in consumption in a year of crop failure. The maximum single yearly shortfall in actual production below trends during the period 1955/56 to 1972/73 for each type of grain at the world level was measured as an indicator of the maximum level of stock that would have been required in the past to maintain consumption levels. An historical analysis of past ratios of actual carry-over1 stocks to total disappearance (domestic consumption plus exports) in exporting countries, and to total domestic consumption in importing countries,

Proposal for Combating Acute Food Shortages

187

was carried out for all countries for which stock data were available. The aim was to identify those ratios that in the past (1961–1974) were related to ‘normal’ or ‘abnormal’ situations in the grain economy. This analysis provided a measure of the level of working and reserve stocks. All three of the methods used by FAO are based on maximum previously observed discrepancies between a supply and a demand measure. In the third method the level of working2 stocks is estimated. It is assumed simply to be the lowest level of stocks relative to demand. At this point, corresponding to 1973/74, all reserve3 stocks were considered to have been exhausted. The two first methods are used to estimate the combined level of stocks, which is equal to working stocks plus reserve stock. According to FAO (1983) the three methods yielded similar results. A minimum safe level of world carry-over stocks for all cereals should be within a range of 17–18 per cent of world cereals consumption. The major part of this level consists of working stocks, whereas the reserve element amounts to 5–6 per cent of world consumption. The figure of 17–18 per cent has been maintained by FAO since as a target level for assessment of the state of food security. Thus, after the trade liberalizations resulting from the Uruguay Round some concerns have been raised by FAO that world food stocks may fall below the minimum safe level so defined. Up until the Uruguay Round, agricultural policy often dictated intervention in the market through public purchases in order to diminish supply, thus maintaining agricultural prices at a politically acceptable level. Consequently, stocks of agricultural products increased and, as a side effect, FAO’s safe level of stocks was more likely to be maintained. In the post-Uruguay Round, however, because of the liberalization of the agricultural markets, this mechanism is expected to decline, implying lower levels of stocks overall. The methods employed by the committee have a very ad hoc nature. For instance, why is it that reserve stocks and only reserve stocks have been used at the low point? No compelling argument can be found. It could very well be that some of the working stock had been used too. Furthermore, the historical period under consideration (1955/56 to 1972/73) was a time of widespread agricultural protectionist measures. Thus, the actual stock levels may have had very little to do with the levels needed for keeping the supply lines operating smoothly. Another point is that the level of working stocks needed may not necessarily be stationary over time. Stocks have, in general, high upkeep costs providing incentives for reducing them. Indeed, popular concepts such as ‘Just in Time’ inventories provide some evidence that working stocks may be reduced. The advance of new effective means of communication, including the internet and cellular phones, as well as the expansion of roads and railways, all contribute to make possible the reduction of working stocks. Furthermore, the market situation could influence the level of stocks. Thus, in times of high prices, working stocks may be reduced, and vice versa in times of low prices. Overall, the rule of thumb that 17–18 per cent of world

188

K.M. Lind

food consumption provides a minimum safe level of food stocks is not supported by irrefutable arguments. Moreover, working processes, organizational advances and technological improvements may have changed the need for keeping food stocks (see also the early critical analysis of the FAO target figures by Sarris, 1985).

Estimating the Level of Food Stocks for Sub-Saharan Africa The FAO requirement for a minimum safe level of food stocks applies to the entire world. In practice, however, some countries are immune, from a disaster point of view, to fluctuations in food supply, while others are highly vulnerable. As shown by Diaz-Bonilla et al. (Chapter 8, this volume), all developed countries in the world are food secure. That is, they have enough production capacity and/or sufficient resources to sustain periods of food shortages. On the other hand, most of the least developed countries are food insecure and, therefore, exposed to the fluctuations of food supply and world food prices. Other developing countries show much more heterogeneity and may or may not be food insecure. This chapter narrows the area of interest to sub-Saharan Africa. African countries south of the Sahara are of particular interest in the development literature. Other developing country areas such as South America and Asia have experienced important improvements in their incomes and standards of living. While in the 1950s several African countries had income and growth exceeding that of Asian countries, the situation has been reversed since then. The sub-Saharan countries in particular seem stuck at a lowincome level, which is reflected in inadequate levels of consumption of calories per capita, and are thus exposed to risks of undernourishment and hunger. This vulnerability is exacerbated by the fact that many of these countries are net food importers. Domestic production is not sufficient to satisfy domestic demand and in order to import food the country is, therefore, dependent upon its foreign exchange reserves, which are not abundant. In fact, in the scenario of disasters and emergencies sub-Saharan countries are often not in a position to buy enough food in a timely manner. Therefore, the international system of disaster and emergency relief is required, which is operated through various organizations. The food aid needed in such circumstances is supplied by the major food exporting countries of the world, where the European Union is an important contributor. Actually, one of the reasons stated for the CAP is to provide a safe level of food supply is, although this presumably is directed only towards the EU member countries. However, the EU has, for a long time, been a net food exporter and, even though yields fluctuate according to the specific conditions within the harvest year, this has not changed the net exporting position at any time during the last couple of decades. In fact, the EU has built up large stocks of food that, in turn, are drawn upon in times of need by the food-deficit countries, primarily located in subSaharan Africa.

Proposal for Combating Acute Food Shortages

189

In Fig. 9.1 the sub-Saharan Africa countries are sorted by ascending GDP per capita based on purchasing power parity (PPP). Thus, Tanzania has the lowest GDP per capita and Mauritius the highest among the countries displayed here. The dashed line shows human consumption in calories per capita per day according to FAO, whereas the solid line shows the estimated human consumption, in calories per capita per day based on GDP per capita. Evidently, increasing income implies increasing calorie consumption, but the prediction power of GDP is very low. For instance, Nigeria, with US$833 per capita, consumes 2771 calories per capita while Angola, with US$1997 per capita, consumes only 1916 calories per capita. Consequently, as a predictor for nutritional status, the direct measure of calories per capita per day is superior to indirect measures.

Statistical analysis of critical food deficits According to FAO (1999) approximately 800 million people suffer from undernourishment. This is a consequence not so much from unexpected events in the form of disasters, conflicts and bad harvests around the globe – although such events occur on a regular basis – but from lack of access in stabilizing food sources. In this study, chronic undernourishment is separated from disastrous events because stocks are modelled as an insurance policy. Providing adequate food supplies for 800 million people is not a ‘stock’ issue: that would require a totally different policy

3100

Calories per capita per day

Regression line Calories per capita per day

2600

2100

Ta n

za M nia a Bu law Si E run i er th d M ra iop i oz L ia am eo bi ne qu Me N ali M Z ig ad am er ag b as ia c C ar Er had i N tre Bu R ige a rk wa ria in n a da C Gu Fa on in go ea Be so , Bi nin C D en em ss tra . R au lA K ep fri ca U en . n g ya C Re and on p a go ub , lic Se Re ne p. G ga am b To l C ia, go am T e he C Co roo ôt m n e o M d'Iv ros au o ri ire Le tan so ia G tho h G an ui a n Zi An ea C mb gol ap a a e b Sw Ve we az rde N ila Bo am nd ts ibia w G an M ab a au o rit n iu s

1600

Note. The sub-Saharan countries are sorted according to GDP per capita, PPP, in ascending order. Thus, Tanzania is the poorest country and Mauritius the richest. Fig. 9.1. Calories per capita per day regressed on GDP per capita, PPP.

190

K.M. Lind

comprising production, transportation and distribution, which is beyond the scope of this study. Thus, only the temporary negative shocks to food supply are considered. Certainly not all temporary negative shocks to food production pose a problem. On the contrary, such events are an integral part of food production that is exposed to several uncertain external factors. The amount and timing of rainfall vary from year to year, pests and diseases may attack the crops, and other detrimental events may occur. Thus, production uncertainty is inherent in food production. In affluent societies this uncertainty does not pose any difficulty to maintaining overall food supplies at an adequate level. Exports and/or imports function as buffers for overall food supply for the domestic population. In poor countries, however, large negative shocks may render the policy of adjusting exports and/or imports unfeasible.

Data Data for consumption of food are obtained from FAO’s database (http://apps.fao.org/). The database contains, among other statistics, food balance sheets for practically every country in the world, with annual data for the period 1961–1996. A food balance sheet shows supply and demand of food in metric tons in a particular year for a particular country/region. Food is divided into a large number of food items (around 100). Supply is comprised of domestic production, imports, stock changes and exports. In FAO’s terminology positive stock changes add to supply and positive exports subtract from it. Thus, total domestic supply equals production plus imports plus stock changes minus exports. Demand is comprised of feed, seed, processing, waste, other uses and food. Furthermore, in the food balance sheet per capita supply is shown, which is derived as demand for food in kilograms divided by population. Three nutrientspecific variables are also shown in the food balance sheet – calories per capita per day, grams of protein per capita per day and grams of fat per capita per day. The focus here is on calorie consumption, as the chapter is concerned with catastrophic events where food is needed for survival. In the longer term other nutrient features of food intake are, of course, of great concern. However, this important issue is not related to food storage as such, but has to do with longer-term developments of food availability and accessibility. Throughout the chapter – as well as in FAO’s food balance sheets – the term calorie is shorthand for kilocalories (1 calorie = 4.19 kJ). The supply of calories per capita per day is derived as the per capita supply of food in kg per year divided by 365 to obtain the per capita supply of food in kg per day. This is then divided by a conversion factor, which is the number of calories per kg of the particular food item. The result is the per capita supply of food in calories per day. For protein per capita per day and fat per capita per day similar calculations are done.

Proposal for Combating Acute Food Shortages

191

Sub-Saharan Africa, which is the focus of the study consists, according to FAO, of 52 countries. FAO have data for 46 of these countries, and the period considered for the estimations is 1961–1996. The remaining countries are British Indian Ocean Territory, Equatorial Guinea, Mayotte, Reunion, Saint Helena and Western Sahara. However, these countries have a small population of about 1.6 million people in total (see http://www. odci.gov/cia/publications/factbook), while the included countries have more than 500 million people, so the average results should not be biased by their omission in this analysis. For Peoples Democratic Republic of Ethiopia, data for the period 1993–1996 are calculated by combining the data from Eritrea and Ethiopia. Thus, the two latter countries are not represented individually in the study and data consist of annual observations in the period 1961–1996 for 45 countries. In particular, observations on calories per capita per day and population are used. Figures on population are likewise obtained from FAOSTAT. The definition of survival rates of energy intake is somewhat ambiguous. In WHO (1985) the survival requirement is defined as 1.27 times the basic metabolic rate (BMR).4 Thus, from the equation 1.27  BMR, in principle, it is possible to estimate survival energy consumption. However, the BMR changes according to weight, height, age and sex. Likewise, other factors contribute to BMR, such as pregnancy, where the BMR is increased. In FAO/WHO (1985) BMRs for various combinations of those characteristics are presented. This study uses the figures for men, 18–30 years of age, 1.7 m in height and weighing an average of 63.5 kg; and for women, 18–30 years of age, 1.6 m in height weighing an average of 54 kg. These figures are added and divided by 2, which assumes, as a reasonable approximation, that the number of males equals females. FAO/WHO (1985) figures for BMR in calorie requirements per day for the two groups are 1650 for men and 1290 for women. Thus, the average survival requirement for energy intake is calculated as 1.27  (1650 + 1290)/2 = 1867 calories per day. This method should not understate the need for energy (and rather this subsistence level may be high), because the populations of sub-Saharan Africa are, on average, very young, with a large number of children that have a lower BMR than adults.

Probability of food deficit The problem of assessing the probability of a food deficit is twofold. First, what is the probability of per capita consumption falling below the critical level? (an observation of calorie intake below the critical level will be called an event in what follows). Second, in the scenario of an event, what is the probability of the deficit being of a particular size? In both cases, the recorded events and sizes are approximated by an adequate probability distribution. The approximation is based on the statistical properties of data from sub-Saharan Africa.

192

K.M. Lind

Probability of an event occurring As outlined above the purpose of the study is to assess the need for the keeping of food stocks. Because supply of food follows an annual cycle the natural time period for the study is by year. Thus, the first step is to approximate the number of annual food deficits where food stocks may be of use in meeting these demands. Again, because of data availability the number of food deficits is equal to the number of countries experiencing food deficits. Data are recorded in the period 1961–1996 for 45 countries, yielding a total number of 1620 observations. With the critical level equal to the subsistence level (1867 cal.) 273 events are observed, or an average of 7.58 events per year. The mode of these frequencies is seven, meaning that seven events per year is the most frequent outcome. As they stand, the frequencies are not particularly useful for assessing the probability of a given number of outcomes per year. Certain outcomes, such as 11 or five events per year, have not been observed in the period under consideration. Thus, they would be assigned a probability of zero if based solely on the frequencies, which does not appear too satisfactory. Instead, a generalization is needed in the form of a distribution, which assigns probabilities over the entire range of possible outcomes. Hence, the problem is to seek among distributions in order to find a satisfactory generalization of the frequencies. A first requirement of potential distributions is given by the nature of the problem. Number of events per year can only take on discrete values, thus the search process is restricted to discrete distributions.5 One obvious candidate is a distribution that yields the probability of the number of events per time period known as the Poisson distribution. However, this result on the Poisson is based on certain assumptions: in particular, events are assumed to be independent. This may be questionable if the probability of observing a value below the critical level for a given country is higher if the value was below the critical level for the previous year. Nevertheless, whether the Poisson is an adequate approximation must be left to empirical analyses. The Poisson distribution has a number of appealing features. It is a well-known and relatively simple distribution that is fully described by only one parameter, and a maximum likelihood estimate of this parameter is easily derived. Here, it is postulated that the frequencies of the events can be described by T independently and identically Poisson-distributed random variables Xt, f ( x|λ ) =

λxt exp(− λ ), t = 1, 2 ,..., T . xt !

(1)

Thus, the joint density is of the form, T

∑ xt

λxt λi =1 f ( x1 ,..., xT |λ ) = ∏ exp(− λ ) = T exp(−Tλ ) i = 1 xt ! ∏ xt ! T

i =1

(2)

Proposal for Combating Acute Food Shortages

193

The ML-estimator6, λˆ , is, 1 T λˆ = ∑ xt T t =1

(3)

The ML-estimator is given by the average number of events per year. Because the Poisson distribution depends only on the λ -parameter, it is fully described by the ML-estimate. The Poisson distribution can be approximated by a normal distribution with mean and variance equal to the mean and variance of the Poisson distribution. The frequencies of events, the estimated Poisson and the approximating Normal distribution are shown in Fig. 9.2. Table 9.1 shows some descriptive statistics for the frequencies.

0.45 0.40

Frequency

0.35

Normal Poisson Frequencies

0.30 0.25 0.20 0.15 0.10 0.05 0.00 0

1

2

3

4

5

6

7

8

9 10 11 12 13 14 15 16 17 18

Number of events

Fig. 9.2. Frequencies, Normal and Poisson distributions.

Table 9.1. Normality test. Test Skewness Excess kurtosis Normality test,  2(2) 95% critical value for  2(2)

Test value 1.260 1.447 0.415 5.991

Note. Skewness is a measure of the symmetry of the distribution. If the distribution is symmetric skewness equals zero. The normal distribution is symmetric. Kurtosis is a measure of the thickness of the tails in the distribution. The normal distribution has kurtosis equal to three. Thus, excess kurtosis is defined as kurtosis minus three. A test for normality of the distribution can be derived from the two measures, skewness and excess kurtosis. The test is  2 distributed with two degrees of freedom.

194

K.M. Lind

Some skewness and excess kurtosis are observed, but according to the normality test these third and fourth moment measures are well inside the acceptable region. The value for the test is shown in Table 9.1 along with the test value for a significance level of 0.05. As seen, the null hypothesis of normality is convincingly accepted. Therefore, the Poisson distribution appears to provide an adequate description of the frequencies. Probability of the size of an event When an event occurs, the scale of the food deficit is, of course, of vital importance for any strategies aimed at alleviating food shortages. The analysis now uses the data not indexed by time but treated as one big sample, except that only observations where an event occurs (i.e. when the per capita consumption falls below the critical value) are included. As indicated, this happens in 273 cases, which constitute the sample of observations. The time and place (country) where the event happens is of no concern in this analysis. In particular, the place (country) where disaster strikes is of no importance because it is assumed that donors and administrators of the food aid system are willing to help all countries. Therefore, in this analysis the purpose is to obtain a probability distribution for the size of events upon which the system for the entire region should be based. When an event occurs, where the food consumption in calories per capita per day is below the critical value, the shortage is multiplied by the number of people in the country and the number of days in a year. The resulting number shows the annual food shortage in calories. These observations are ordered according to increasing size. Subsequently, the events are collected in intervals where the upper limit is 150 billion calories higher than the lower limit, and the number of events in each interval is derived. Almost half the events fall into the first interval (0–150 billion calories) and the frequency of an event declines rapidly with increasing scale of the event. In contrast to the probability of an event occurring, in this case there are no prior hints as to what sort of probability distribution to look for. However, the configuration of the frequencies immediately excludes many possible candidates. Thus, the normal distribution is clearly inadequate in this case, with the frequencies forming a highly asymmetrical shape. On the other hand, the rapid decline of the frequencies points to some sort of exponential feature. One candidate that is capable of producing such behaviour is the exponential distribution, where frequencies of food deficits are described by a random variable y with density function, f ( y|θ ) = θ exp( −θy ), y > 0.

(4)

The expected value of the exponential distribution is derived as, E[ f ( y|θ )] =

1 , θ

(5)

and the variance as, Var[ f ( y|θ )] =

1 . θ2

(6)

Proposal for Combating Acute Food Shortages

195

The exponential distribution is a continuous distribution as opposed to the Poisson distribution, which is a discrete distribution. Like the Poisson, it is a well-known, simple distribution that depends only on one parameter, . A property of the exponential distribution is that it is invariant to scale changes. In case the exponential description is an adequate approximation to the observed frequencies, Fj, Fj ≈ Pj = 1 − exp(−θy j ),

(7)

a linear in parameters relationship is derived, log(1 − Fj ) ≈ −θ y j .

(8)

The parameter, , in the exponential distribution is derived from this relationship. An OLS regression of log(1
Fj) on yj yields, log(1 − Fj ) ≈ −0.00127 y j

(9)

(0.00005)

where the number in brackets is the standard deviation and R2 is 0.77. Thus, the estimate of  is 0.00127. If the estimated exponential distribution is an adequate approximation to the observed frequencies, the latter should be dispersed more or less randomly around the straight line in Fig. 9.3, where the observed frequencies are shown together with the estimated exponential distribution. Admittedly, the observed frequencies display signs of systematic deviations from the estimated exponential distribution. Thus, for lower values of Y the corresponding estimated probabilities are lower than the observed frequencies (log(1
Hj) < log(1
Pj) => Pj < Hj). Nevertheless, overall the deviations are small as some 77 per cent of the variation in the observed frequencies is explained. Accordingly, the approximation in the form of the estimated exponential distribution is considered adequate for the present purpose. 0.000 Observed frequencies Exponential distribution

Frequency

–1.000 –2.000 –3.000 –4.000 –5.000 –6.000 0

1000

2000

3000

4000

Billion calories

Fig. 9.3. The observed frequencies and the estimated exponential distribution.

5000

196

K.M. Lind

The mean of the exponential distribution is derived as, E[Y ] =

1 , θ

(10)

where the estimate of  is 0.00127. Thus, the expected value or mean of the size of calorie deficits is 789.65 bn calories.

Summary of statistical analysis: how large the stocks should be? From the Poisson distribution an average of 7.58 events of food deficits occur every year. Multiplying this number by the average size of a food deficit obtained from the exponential distribution yields 7.58789.65 = 5,989.55 bn calories. Consequently, according to the statistical analysis, approximately six trillion calories are needed annually on average to cover the most basic nutritional requirements. By itself this number may not be particularly informative, since calories per se are not stored or traded. Instead, the calorie equivalent of food stocks can put some perspective on the numbers. The most stored food item, particularly in the EU, is wheat. Therefore, wheat is used as a reference point in the subsequent calculations. Furthermore, wheat is traded and consumed throughout the world, although the amount of wheat in daily diets varies across regions and cultures. Also, in comparison to other food items, it is relatively easy to transport and distribute. The average calorie content of wheat is derived from FAO’s food balance sheets. The total supply in the world of wheat in calories per capita per day – and total food supply of wheat – is used to derive the number of calories per kilogram wheat. This is done for all the years in FAO’s food balance sheets and the average is calculated. It should be mentioned that the figure does not change much over the time period. The resulting figure is 2827 calories per kg wheat. Using this to convert the average food deficit per year in calories yields 1,000,000  5,989.55/2827 = 2,117,278 metric tons of wheat. This value must be put in perspective to understand how small or large it is. In Fig. 9.4 is shown the European Union’s ending stocks of wheat for the period 1970/71 to 1999/00. Ending stocks are the stocks at the end of the harvest year, which is around 1 August. At this point the stocks are assumed to be at their lowest during the harvest year, and will soon be replenished with the new harvest. In Fig. 9.4 is also shown a line corresponding to the average food deficit in sub-Saharan Africa, calculated above. Evidently, the wheat stocks of the European Union are more than sufficient to cover the average needs. Furthermore, Fig. 9.4 shows only the wheat stocks, but besides that the European Union also keeps stocks of other grains, not to mention various other food items such as beef, butter and other dairy products.

Proposal for Combating Acute Food Shortages

197

25,000,000 EU stocks

Wheat stocks

20,000,000

Stock needed for security according to calculations

15,000,000

10,000,000

5,000,000

0 1970/71

1975/76

1980/81

1985/86

1990/91

1995/96

2000/01

Fig. 9.4. Ending stocks of wheat in European Union (m metric tons).

According to USDA, in 1999 global grain trade comprised 102 m. MT of wheat, 25 m. MT of rice and 91 m. MT of maize. These figures do not include trade among the EU countries themselves. Thus, in wheat equivalents a total of 230 m. MT was exported (and, of course, the same amount was imported). Of this amount an average of 2.1 m. MT would be needed for sub-Saharan Africa, and even considering the largest deficit experienced (3 m. MT), only slightly more than 1 per cent of the global grain trade is needed.

FAO’s recommendations Table 9.2 summarizes different policy options dealt with in this chapter. FAO’s recommended food stock policies as described in FAO (1983) are displayed in columns one and three. Column one shows the recommendation for the entire world. It is surprising how close the recommendation for total grain stocks is to the actual grain stocks in the world, as displayed in column two. Then again, FAO’s recommendation is derived from the actual behaviour of food stocks, and apparently this behaviour has not changed much since the recommendations were calculated based on numbers from the 1960s and 1970s. The cost of keeping the stocks recommended by FAO amounts to more than US$5 bn dollars per year. Most of these stocks, however, are not for insurance against hunger. According to FAO approximately two-thirds are working stocks that are kept to ensure the smooth and uninterrupted flow of food and foodstuffs, both in the production and the distribution processes. In reality, however, large parts of these stocks have been acquired as a result of the agricultural policies in the exporting countries. Thus, in periods of low demand and/or ‘too’ high supply, EU and US policies call for interventions in the markets, often by purchasing grain and subsequently placing it in storage buildings. Therefore, a large part of the actual stock level may have little to do with supply security.

198

Table 9.2. FAO stock policies. Based on FAO’s recommendation: World Food consumption in weq.a, m. t Total grain stocks in weq.b, m. t Reserve stocks in weq.c, m. t World price of wheatd, $/t Maintenance costs of wheat stockse, $/t/year Costs of maintenance of total stocksf, million $/year Cost of maintenance of reserve stocksg, million $/year Cost of purchasing the total grain stocksh, million $ Cost of purchasing the reserve grain stocksi, million $

2,159.4 377.9 118.8 93.7 14.1 5,328.4 1,674.6 34,397.3 10,119.6

Actual stocks: World 376.9 93.7 14.1 5,314.3 35,315.5

Based on FAO’s recommendation: sub-Saharan Africa 167.2 29.3 9.2 93.7 14.1 412.5 129.7 2,661.1 787.1

Required stocks: sub-Saharan Africa

Proposal: sub-Saharan Africa

2.1 93.7 14.1

0.0 93.7 14.1

29.6

0.0

196.8

196.8

Note: weq. is wheat equivalents. All quantities are in wheat equivalents. In 1999, 5,956,286,000 people in the world consumed an average of 2808 cal/cap./day according to FAOSTAT. Thus, the number of calories consumed per year can be calculated, and dividing this number by the number of calories per kg wheat (2827) the quantity of food consumption in the world in wheat equivalents is obtained. For sub-Saharan Africa in 1999, 589,140,000 people consumed an average of 2198 cal/cap./day according to FAOSTAT. b FAO’s recommendation that 17–18 per cent of total food consumption should be stored is used to calculate total recommended food stocks. Hence, food consumption in weq. is multiplied by 0.175 to obtain FAO’s recommended total grain stocks. The actual stocks are ending stocks for 1998/99, that is the stock level in mid-1999. According to USDA the quantity stored for wheat was 135.6 m. t, for rice 60.0 m. t and for coarse grains 155.3 m. t. The calorie content in these grains is derived by using FAOSTAT data, yielding 2,827, 3,649 and 2,982 calories per kg for wheat, rice and maize, respectively. Thus, the actual grain stocks in wheat equivalents yields 135.6, 77.4 and 163.8 m. t respectively, amounting to 376.9 m. t. c FAO recommends that 5–6 per cent of total food consumption should be stored, which is used to calculate total recommended reserve stocks. Hence, food consumption in weq. is multiplied by 0.055. d Average price for wheat in 1999. Source: USDA. e According to Dhar (1993) maintenance costs of storing amount to 15 per cent of the price per year. Thus, maintenance cost per t is obtained by multiplying world price of wheat with 0.15. f b  e. g c  e. h b  d. i c  d. a

K.M. Lind

Proposal for Combating Acute Food Shortages

199

FAO’s recommendation for a safe level of food stocks applies to the entire world. Nevertheless, in column three the recommendations are used for the sub-Saharan Africa region. This region is much more prone to famines and disasters than most, if not all, the other regions in the world. Average calorie consumption is at a low level; therefore, it does not take very major catastrophes to put dietary needs in jeopardy. Furthermore, natural catastrophes such as droughts and floods are common in the region – as are man-made disasters such as civil war. For these reasons, FAO’s recommendations for stock levels devised for the entire world should be increased significantly if applied to the sub-Saharan Africa region. In column four, however, the figures are calculated without increasing the recommendation. Thus, these numbers provide a minimum for the FAO stock levels in sub-Saharan Africa. According to the FAO recommendation at least 9.2 m. t of wheat equivalents need to be stored as reserves. Apart from purchasing this grain at a cost of US$787.1 m. in 1999, an annual cost of US$129.7 m. is induced by the reserve stock. It must be assumed that the grain used to alleviate famines and hunger is not sold on the market but donated to people in want. Thus, the reserve stock does not provide funds for itself and depends upon adequate and timely donations for replenishment and upon financial assistance in order to maintain the reserves. However, no information exists regarding the operations of these stocks. Information on size and timeliness of withdrawals and replenishments from the reserve stocks is needed to get a precise estimate of the costs of the operation. Likewise, the criteria for donating food supplies would be required in order to assess the likely scale of operations.

Statistical analysis Since none of this information is available, the statistical analysis based on the minimum calorie criteria is used to evaluate the FAO policy. Column four displays the figures based on the statistical analysis. On average, to achieve at least the minimum requirements for calorie intake, 2.1 m. t per year are required. This number is only about one fourth of the amount given by the FAO recommendation. A reserve stock of this size would, naturally, also cost only about one fourth of the FAO recommendation. However, the figure from the statistical analysis is an average, thus, in some years only small or no food donations are needed to alleviate sudden food crises, whereas in other years the wants will be much higher. In the examined period 1961 to 1996 the highest deficit in wheat equivalents was in 1993, where just below 3 m. t were needed to cover the basic calorie requirements. In 1993 prices, US$119.5/t, this would have required around US$360 m. Still, the quantity of the highest deficit is only about one third of the FAO recommendation.

200

K.M. Lind

Costs of keeping stocks The upkeep and maintenance costs of perishable food products are quite high, and particularly, if the product is meat. Meat needs refrigeration, which is energy consuming and thereby costly to maintain. But even with grain, costs may reach considerable amounts. Despite the numerous studies that have been conducted concerning stock needs, few seem to consider the magnitude of the costs. Implicitly, however, some measurements can be deduced. The CAP of the European Union has been heavily dependent on stocks to regulate prices received by farmers. To this effect, EU compensates farmers to stock up agricultural products by paying them a sum to cover the expenses of private stocks. A study by Dhar (1993) makes an attempt to quantify the costs. Based on the operations of the Food Corporation in India, it is estimated that the maintenance costs of the (buffer) stock of grain should be around 15 per cent of the price of the grain. However, the current cost of maintaining the stock is more than 30 per cent of the price of the grain, according to Dhar (1993). If wheat only is considered, in 2000 approximately 600 m. t of wheat were consumed in the world. This led to a required stock – according to FAO’s 17 per cent rule – of around 100 m. t of wheat. The maintenance costs of this stock are more than one billion dollars. Considering only the part of FAO’s requirement that refers directly to emergency needs, which is one third of the 17 per cent, the upkeep on these stocks still runs into several hundred million dollars. According to the calculation from the statistical analysis, an amount of 2,117,278 t of wheat are needed on average every year to supply the needs of sub-Saharan Africa. Admittedly, the rest of the world is neglected in the statistical analysis; however, the FAO requirements do seem quite an expensive way to ensure food security. The EU itself, through the workings of the CAP, indirectly provides another estimate. Thus, the CAP stipulates an intervention price for the supported products, which includes wheat. That intervention price is valid in the first month of the crop year. If the farmer postpones his sale to the EU through intervention then he receives an addition to the stipulated intervention price for every month after the first harvest month he has postponed the sale. This monthly addition is granted because the farmer incurs the costs of keeping his grain in stocks. Thus, this monthly addition is in essence an estimate of the costs of keeping stocks. In 2000 this addition was set at €1/t according to Landbrugsraadet (2000), which is equivalent to $0.92/t in 2000 prices. A crude calculation yields the annual costs of keeping one metric tonne of wheat as 12  €1 = €12 = $11.1 (at the average exchange rate for the year 2000), which is close to the estimate based on Dhar (1993). From the perspective of available resources the amount of wheat traded in the world was over 100 m. t in 2000. In wheat equivalents, the amount traded in the crop year 1999/00 was 104.3 m. t of wheat, 29.3 m. t of rice and 101.4 m. t of coarse grain, totalling 235.0 m. t of wheat equivalents. Furthermore, the figures based on USDA data exclude intra-

Proposal for Combating Acute Food Shortages

201

EU trade. Thus, allocating to food security in SSA a couple of million tonnes, or even twice that amount, through purchases from the world market, does not seem to have a large impact. The problem is that available resources must be ready before a crisis emerges in order to buy the required amount of grain. It is evident that if sufficient resources can only be made available through a complicated political process involving disputes over who is to pay what, then such a strategy will fail. The urgently needed food aid would be delayed, with potential devastating implications. An alternative strategy The problem of hunger and undernourishment is not, as such, a problem of too little food available. World food production provides sufficient amounts in terms of calories per capita (FAO, 1999). However, lack of access to food for poor people is the source of the recurrent food crises. Hence, researchers have focused on financial arrangements by which poor people can achieve credits to enable consumption smoothing over time, thereby alleviating short-term hunger periods. A vast body of work on both rural and urban credit arrangements for the poor exists (see Zeller et al., 1997). Often these credit arrangements need public support, one major reason being that poor people seldom have collateral (Zeller et al., 1997). Thus, purely market-based credit institutions are unlikely to address these problems. Furthermore, if creditors are only committed for a limited time period the credit institutions experience widespread loan defaults in the final year of external subsidies. Thus, credit institutions aimed at alleviating financial problems for the poor need to be based on a more permanent source of subsidy arrangement. In the case of food security at national level, and recalling that the actual needs constitute only a small part of both available stocks and current trade, the question is: why not just finance the buying of the required food from existing traders instead of keeping own stocks that entail huge expenses for maintenance? An international financial fund could be established with the sole purpose of functioning as financier for buying food from world markets in case certain emergency criteria are met. Konandreas et al. (1978) and Huddleston et al. (1984) advocate a similar arrangement. In Huddleston et al. (1984) the focus is stabilizing food supply through imports. Consequently, the financial facility targets fluctuations in the trade balance and exchange rates (see also Ravaillon, 1996). Furthermore, Sarris (2000) has suggested a fund with similar objectives based on option contracts. From Table 9.2 it can be seen that the fund would have to spend on average around US$200 m. per year to acquire grain. These costs are exactly the same as under a storing policy, since presumably the stocks have to be replenished each year on average by this amount. As mentioned, in the observed period the maximum amount required in 1999 prices was 360 US$ m. dollars. So, a fund with, say, US$400–450 m. in capital should suffice.

202

K.M. Lind

A financial fund must be financed and replenished continuously through a set of agreements among donor countries. The advantage of this strategy is that it is much cheaper than providing emergency aid through stocks. Stocks are, as shown above, very expensive to maintain. The advantage of relying on stocks is that the available resources are already at disposal. This in no way implies that they are in the right locations. On the contrary, stocks are primarily kept by big exporting countries that do not need them for security purposes. Thus, all the problems involved in the logistics and distribution of the emergency aid are present. The alternative strategy of having an international financing fund would also start the emergency aid from much the same place as under the stock strategy. When the fund buys up grain at the world market it would obtain the grain at much the same places as they are kept under the stock strategy: the ports of the big exporting countries. This concern over the future of the world’s disaster aid capabilities arose because of the liberalizations agreed upon during the Uruguay Round of GATT negotiations. In order to assuage anxieties, and also for other reasons voiced in particular by developing countries over even further liberalizations, the matter of emergency aid could be a negotiating topic during the ongoing WTO trade negotiations. The principal interested parties in a formal emergency aid fund would be the developing countries and, in particular, the least-developed countries, since they stand to benefit the most from such an arrangement. But the developed world also stands to benefit considerably by further liberalization of agriculture and, as the WTO ministerial meetings in Seattle and Cancún demonstrated, they need the consensus of the developing countries to move forward. Therefore, part of the price for the consent of the least-developed countries could well be the establishment of such a fund, along with binding agreements and commitments to the allocation of the financial resources. The fund needs to be able to conduct its business in a completely independent manner. One argument against such a fund could be that in times of worldwide short supply – and therefore high prices – the developed countries would be reluctant to allow major buy-ups from the market. Thus, if the operations of the fund are under political control, emergency aid could be impeded or delayed intolerably. It may be true that in particular years the costs of providing emergency aid will be high – as in the years of worldwide short supply. But, in the longer perspective, it is much cheaper to operate through buy-ups as opposed to spending millions of dollars every year on maintenance costs. Furthermore, in case of the stock-based approach, besides the maintenance costs the stocks have to be replenished when drawn upon for emergency aid. It is conceivable that the costs of replenishing the stocks could be lower than the costs of making the buy-ups in times of need. However, these conceivably lower costs are hardly millions of dollars cheaper. Moreover, a financial fund

Proposal for Combating Acute Food Shortages

203

would be earning interest on its capital. Admittedly, the funds would have to be in a highly liquid form, but still the fund would achieve a positive rate of return. In the stock-based strategy the capital is tied up in buildings and the grain stored, which earns no interest. On the contrary, the rate of return is negative because both buildings and grain depreciate over time and need to be maintained. Most countries in the world are food secure, so they do not need to store food in large inventories in order to insure themselves against bad harvests and other adverse events. This function can readily be left to the market, where the demand and supply situation in individual countries will adjust, with prices going up if supplies decrease. However, paying an ‘over-normal’ price for a temporary phenomenon will, over time, be cheaper than incurring the very sizeable costs a stock-based strategy entails. Nevertheless, this form of ‘insurance’ provided by the market excludes countries that do not have ready access to foreign exchange. In particular, poor countries, such as those of most of sub-Saharan Africa, face risks of grave disasters now and in the future. Therefore, an insurance system has to be set up for such countries. Again, the contention is that stockholding is too expensive a strategy. Instead, a financial fund with highly liquid assets should be established. This fund should have no other purpose than to provide food in kind when needed. Thus, when a country faces serious food shortages the fund should buy up food on the open market and put it at the disposal of the proper authorities. Whether the fund should also be involved in transportation and distribution of food aid is a separate issue, which will require additional donors’ contributions. Another prerequisite for this fund is that it should be continuously replenished. Thus, an automatic refinancing structure needs to be implemented. This is not an easy task, as the controversies surrounding financing of the UN and other international organizations indicate. However, the sizeable stockholding costs that can be saved if this fund is set up should provide an economic incentive for choosing such a strategy. Thus, approximately US$4 bn is being spent on stockholding each year, if FAO’s 17 per cent requirement is followed according to the calculations above. One third of the 17 per cent stockholding is for reserve purposes. If these reserve stocks could be saved more than US$1 bn would be saved each year. Of course, some administrative costs would be incurred by the financing fund, but hardly at the scale of stock maintenance costs. The fund would, on average, have to act 7.5 times per year according to the statistical analysis. Thus, the fund would need to be replenished at least every 2 months. Naturally, the fund would be inoperable if political compromises and dealings had to be undertaken each time. Instead, an automatic system for replenishing the fund needs to be implemented – perhaps a system based on countries’ relative income or following the UN contributions.

204

K.M. Lind

An argument in favour of stocks is that, as mentioned above, the food is already at hand and, thus, time is saved in the distribution process. However, the bulk of the world’s grain stocks are kept in and by the big exporting countries, and are therefore not located in the countries where they would be most likely to be needed. Furthermore, if grain is purchased on the world market the location of the purchased food is where it is stored by the vendors, primarily in the silos of the ports of the big exporting countries. That is, the purchased food grains would be located at the same places as reserve stocks. The argument in favour of strict criteria to be met before the fund takes action is to counter the problems of moral hazard that can arise. Thus, the fund would induce a positive incentive for countries or regions to declare themselves in a sudden emergency. Undoubtedly, in many countries and regions in sub-Saharan Africa, perennial problems of malnutrition exist, and as such they qualify for assistance. It is important, however, to emphasize that the fund is to be regarded as an insurance against suddenly arisen emergencies, not as a general organization to combat malnutrition. On the other hand, the decision making in the fund has to be flexible in order to enable quick and timely assistance. Obviously, the exact charter of the fund therefore must reflect both of these opposing considerations. The fund could be set up by the UN, which has institutions and expertise of working on problems of hunger. Alternatively, given the limited funds needed for the operation of such fund, an arrangement within the EU could be possible. This would also ensure shorter, and perhaps more direct, lines of communication between donors and fund operators, with potentially more adequate and timely financing.

Concluding Remarks With further liberalization of world agriculture there have been concerns that the decline of food stocks associated with reductions in production subsidies, mainly in industrialized countries, could endanger world food security, particularly in more vulnerable countries. This chapter reexamines the issue of what should be the adequate level of food stocks for insurance purposes, based on a stochastic approach that focuses on calorie availability. The calculations focus on sub-Saharan Africa, one of the developing regions most affected by food deficits and variability. The conclusions suggest lower levels of stocks than have been historically recommended, based on more ad hoc calculations that take as given historical stocks created by agricultural subsidies and protection. The chapter also explores the notion of a financial fund for food security purposes, as an alternative to maintaining physical stocks. It finds that the savings in maintenance and management of the latter would favour the implementation of such a financial fund.

Proposal for Combating Acute Food Shortages

205

Notes 1

Carry-over stocks are stocks held from the end of one crop year to the beginning of the next. It is the same as ending stocks for a particular crop year. 2 Working stocks are those required to assure the smooth and uninterrupted flow of supplies through the production and distribution lines. 3 Reserve stocks are those that can be drawn on to meet unexpected deficits and other emergency needs. 4 The 1.27 factor is obtained by 1  1/3 + 1.4  2/3. The 1/3 is the average proportion of time spent on sleep and, hence, the 2/3 is the period awake. In the period awake, even if no activity as such is carried out, the acts of washing, dressing and short periods of standing raise the energy consumption to 1.4 BMR. 5 A more formal way to present this is that the dominating measure is a counting measure on {0,1,2,…}. 6 It is easy to derive, T

T

t =1

t =1

lnL(λ | x ) = −Tλ + ln(λ )∑ x t + ∑ ln(x t !), ∂ lnL(λ | x ) 1T = −T + ∑ x t = 0 => ∂λ λ t =1 T ˆλ = 1 x , ∑ T t =1 t

where L(.) is the likelihood function.

References Club of Rome (1972) The Limits to Growth, by Donella H. Meadows, Dennis L. Meadows, Jorgen Randers and William W. Behrens III. Universe Books, New York. Dhar, U.R. (1993) Managerial issues in publicinventory holding: the experience of the Food Corporation in India. In Berck, P. and Bigman, D. (eds) Food Security and Food Inventories in Developing Countries. CAB International, Wallingford, UK. FAO (1983) Approaches to World Food Security. FAO economic and social development, paper 32, Food and Agriculture Organization of the United Nations, Rome. FAO (1995) World Agriculture: Towards 2010, Alexandratos, N. (ed.). John Wiley and Sons, Chichester, UK. FAO (1999) The State of Food Insecurity in the World (SOFI). Rome.

Huddleston, B., Johnson, D.G., Reutlinger, S. and Valdés, A. (1984) International Finance for Food Security. John Hopkins University Press, Baltimore, Maryland. Konandreas, P., Huddleston, B. and Ramangkura, V. (1978) Food Security: An Insurance Approach. Research report 4, International Food Policy Research Institute, Washington DC. Landbrugsraadet (2000) Raadsnyt om EU, News Brief no. 4. Malthus, T.R. (1798) An Essay on the Principle of Population. (Originally printed in London, for J. Johnson, St Paul’s Churchyard, 1798.) In HTML format by Ed Stephan at http://www.ac.wwu.edu/~ stephan/malthus/malthus.0.html Ravaillon, M. (1996) Famines and Economics. Policy Research Working Paper 1693, The World Bank, Washington DC.

206 Sarris, A. (1985) Safe Levels of Global Grain Carryover Stocks for World Food Security. Report prepared for FAO Commodities Division, FAO, Rome. Sarris, A. (2000) World cereal price instability and a market based instrument for LDC food import risk management. Food Policy 25(2), 189–209. WHO (1985) Energy and Protein

K.M. Lind Requirements. Report of a Joint FAO/ WHO/UNU Expert Consultation. WHO Technical Report Series No. 724, Geneva, Switzerland. Zeller, M., Schrieder, G., Von Braun, J. and Heidhues, F. (1997) Rural Finance for the Poor. Food Policy Review 4, International Food Policy Research Institute, Washington DC.

10

Thinking Inside the Boxes: Protection in the Development and Food Security Boxes versus Investments in the Green Box

EUGENIO DIAZ-BONILLA,* XINSHEN DIAO† AND SHERMAN ROBINSON‡ *Executive

Director for Argentina and Haiti, Inter-American Development Bank, 1300 New York Ave., NW, Room NE1137, Washington DC 20577, USA; †Development Strategy and Governance, International Food Policy Research Institute, 2033 K Street, NW, Washington DC 20006-1002, USA; ‡Department of Economics, School of Social Sciences, University of Sussex, Falmer, Brighton, BN1 9SN, UK

Background The discontent with the AoA from many developing countries has been a recurring theme since the completion of the Uruguay Round. It has continued during the current agricultural negotiations, initiated in March 2000 and reaffirmed at Doha in November 2001. Developing countries criticized the proliferation of boxes for domestic support, with the argument that many of the provisions included were exemptions that industrialized countries secured for their policies. They were also dissatisfied with the limited market access opportunities in developed countries’ markets, and the fact that many industrialized countries retained the right to utilize export subsidies. At the same time developing countries indicated their concern that the Special and Differential Treatment provisions1 of the AoA were not comprehensive enough, limiting the range of policies necessary for rural development, poverty alleviation and food security (UNCTAD, 2000). Developing countries have followed two main approaches to address the imbalances they perceive in the AoA. One has been to ‘play offence’, trying to limit the legal ability of rich countries to use high protection and large subsides. The other has been a mostly ‘defensive’ strategy, such as asking for higher permissible levels of agricultural protection and subsidization in developing countries themselves. Of course, both © CAB International 2006. WTO Negotiations and Agricultural Trade Liberalization: the Effects of Developed Countries’ Policies on Developing Countries (eds E. Diaz-Bonilla et al.)

207

208

E. Diaz-Bonilla et al.

approaches are not mutually exclusive, and they have been mixed in different proportions in many proposals presented by developing countries. Mostly within a ‘defensive strategy’, different developing countries have suggested the notion of a Development Box (DB) or a Food Security Box (FSB), encompassing a set of SDT provisions applicable only to developing countries or a sub-group of them, to allow them to further protect and subsidize their agricultural production. The idea of a Development Box was initially advanced by a group of developing countries (WTO document G/AG/NG/W/13 by Cuba, Dominican Republic, Honduras, Pakistan, Haiti, Nicaragua, Kenya, Uganda, Zimbabwe, Sri Lanka and El Salvador). But other countries also supported the idea (for instance ASEAN countries (G/AG/NG/W/55) and the African Group (G/AG/NG/W/142)). India (G/AG/NG/W/102) proposed a variation of similar ideas as a Food Security Box, which has been supported by Indonesia (G/AG/NG/W/115), Turkey (G/AG/NG/W/172) and other countries. The proposals about these boxes, particularly the Development Box, have been also backed by a variety of NGOs (see, for instance, Green and Priyadarshi, 2001 and CAFOD, Action Aid, Oxfam and IATP, 2002). However, some developing countries during the negotiations raised concerns about several of the suggested SDT measures packaged into the Development/Food Security Boxes, especially those that led to increases in protection, mostly because of the possible negative impact on trade among developing countries themselves (WTO , 2004).2 The range of policies suggested under the notions of a Development Box or a Food Security Box is rather broad (see Annex). Although some of the proposals include suggestions to limit the distorting policies of developed countries (‘playing offence’), and others cover some more general issues (such as the secular decline in commodity prices or the role of multinational firms), the core of the provisions suggested focuses on different SDT clauses to protect and subsidize agricultural production in developing countries (‘playing defence’). Within this defensive approach there have been several proposals related to domestic support, but the emphasis from many developing countries seems to be on the request for more flexibility to increase their levels of protection for reasons such as food security, rural development and protection of rural livelihoods and small farmers. Some proposals have argued for a ‘positive list approach’ declaring which agricultural products or sectors they would like included in the AoA provisions, and only those would be subject to AoA commitments (see for instance G/AG/NG/W/13, presented by 11 developing countries).3 Except for a broad flexibility granted through positive or negative lists, the other criteria for general exemptions have usually been included in specific modifications to legal clauses, mainly related to three issues: first, the possibility of maintaining or increasing border protection;4 second, revamping current safeguards (against import surges and collapsing world markets, as suggested by India and others);5 and, third, streamlining of countervailing measures (against subsidized exports, but also domestic subsidies, as suggested by different developing country members of the

Food Security Boxes versus Investments in the Green Box

209

Cairns Group).6 Of course, exclusion of products from WTO disciplines, or protection increased on a permanent basis, are different from the use of contingent instruments for specific trade events. Based on the numerous proposals, Stuart Harbinson, the Chairman of the Committee on Agriculture, presented a draft and a revision of the modalities for the negotiations, (TN/AG/W/1, 17 February 2003 and TN/AG/W/1/Rev.1, 18 March 2003). The Harbinson Proposal (HP)7 did not consider the idea of having positive (or negative) lists of products (which would have basically changed the notion of integrality of the list of agricultural products). It also proposed reductions of market access restrictions for all products and countries, although at different paces for developed and developing and excluding, as it has been the practice in the past, the 49 UN-designated Least Developed Countries from commitments. However, following in part the request of different developing countries, the HP considered the possibility for those countries to declare a number of agricultural products (to be defined during the negotiations) at a six-digit or four-digit level of the Harmonized Tariff System as ‘SP products’ for food security, rural development and/or livelihood security concerns. The HP suggested tariff reductions of 10 per cent with a minimum cut of 5 per cent for those products, but there was no obligation to expand tariff-rate quotas for them. Also for SP products, developing countries would be able to apply a new special safeguard mechanism based on Article 5 of the Agreement on Agriculture, but the operational aspects of this instrument, such as the number of products that can be designated as SP, and the possible operation of the SP safeguard, were to be worked out (there was a blank Annex on this topic). The notion of a new agricultural special safeguard mechanism (SSM) for developing countries in import-sensitive items was maintained in the joint EC–US paper on agriculture (JOB(03)/157, 13 August 2003). A group of developing countries, known as G-20, which considered that the joint EC–US proposal was still maintaining an excessive level of subsidization and protection in developed countries, included the notion of a SSM but also reintroduced the concept of Special Products, as previously discussed in the HP (WT/MIN(03)/W/6, 4 September 2003).8 The revision of the drafts prepared by Perez del Castillo, the Chairman of the General Council prior to the WTO Ministerial at Cancún (Mexico) in September 2003 (JOB (03)/150/Rev.1; Draft Cancún Ministerial Text and Annex A), considered the SSM and some special treatment for SP, as was also the case of the revised Annex A modified during the Cancún Ministerial by the Chair of the Conference, the Minister of Foreign Affairs from Mexico, Ernest Derbez. After the failure of the Ministerial, the so-called July Package of 2004 reaffirmed the notions of both a SSM and Special Products for developing countries. Moreover, it also included a new category of Sensitive Products, open to all WTO members, which may receive more flexible treatment in terms of market access (WT/L/579, 2 August 2004; Annex A). At the end of 2004 a group of developing countries (G-33) presented a proposal for an SSM that could be automatically triggered, apply to all products, and have both price- and volume-triggered safeguards (WTO, 2004; ICTSD, 2005).

210

E. Diaz-Bonilla et al.

Therefore, the ‘defensive’ aspects of the original proposals of the Development Box have been subsumed into the notions of Special Products and a new special safeguard, mostly as mechanisms to provide protection to agriculture in developing countries because of poverty and food security concerns. The negotiations will have to consider several issues regarding the operation of these measures in developing countries. First, whether the availability of the safeguard should or should not be limited to some number of pre-designated products.9 The broader the coverage in terms of products, the larger the possible negative effects on domestic and world welfare. Second is the issue of which countries can have access to that safeguard. The WTO framework does not make distinctions among developing countries. The 50 Least Developed countries, on the other hand, do not have any obligation, so conceivably they could operate tariffs under the WTO system without restrictions. A third question is the operation of the safeguard, which would, in principle, be based on the current Special Safeguard.10 Some proposals suggest the application of quotas (as in the normal Safeguard of Article XIX of GATT 1994), or prefer that the import taxes be applied from bound tariffs (which are higher than applied tariffs in many developing countries). The analysis of the SSM and SP is, in turn, part of broader debate on the merits of protection versus other policy approaches as the best way for achieving the objectives of rural development, poverty alleviation and food security in developing countries. Some of the developing country proposals and several NGOs have recommended the increase in protection of the agricultural sector, suggesting, implicitly or explicitly, that consumers in those countries be taxed to support producers. In a similar vein, it has been argued that the legal exemptions allowed for developing countries under the current AoA (Green Box, de minimis, Article 6.2), or the new domestic support alternatives considered in the HP,11 are of no use to developing countries, mainly because the policies permitted ‘cost money’ and are very difficult to implement due to their technical and human resource requirements. Usually, the conclusion of this line of analysis is that developing countries need additional flexibility, mainly in terms of the levels of protection allowed (see Solagral, 1999; UNCTAD, 2000, for the current AoA; IATP, 2003, for the HP). The implication of this line of argument is that protection ‘does not cost money’ and is easier to implement in poor countries than alternative approaches such as investments in agricultural technology, extension or infrastructure. This debate surrounding the use of consumers’ money through protection versus utilizing taxpayers’ funds through domestic support or agricultural investments as the main defensive strategies for achieving the objectives of rural development, poverty alleviation and food security, will most likely continue during the agricultural negotiations. The following section tries to contribute to that discussion by comparing increases in protection for some special products (the preferred instrument for those supporting the notion of a DB/FSB),12 with investments in agricultural technology (a policy already allowed under the Green Box). The increase in

Food Security Boxes versus Investments in the Green Box

211

protection utilized in the simulations is arbitrary, in order to be able to compare the implicit tax of increased tariffs with an equivalent explicit tax, now collected by the government, which is then invested in agricultural R&D, a typical Green Box measure. These are the two scenarios discussed in the next section. The objective is to make explicit that protection ‘costs money’ and that those funds can have different uses. The simulation utilizes a general equilibrium world model to trace the welfare, production, consumption and trade effects of those different policies. Of particular interest are the impacts on employment and food consumption (as indicators of poverty and food security concerns, the rationale provided by the developing countries defending protection as the main instrument of the DB/FSB), and the consequences for trade among developing countries (the main concern of other developing countries that presented reservations to some components of the DF/FSB). The model considers technological change in the production functions for agricultural sectors that is linked to government expenditures in agricultural research. Therefore, the simulations can evaluate other alternative policies, such as investments in agricultural technology that are allowed under the current Green Box.

Comparing Protection in a DB/FSB with Green Box Measures The model The framework of analysis is a general equilibrium model with a multiregion and multi-sector specification (a more detailed description of the model is in Diao et al., 2002). It includes 46 countries or regions and 37 sectors of production.13 The base year is 1997 and the data come from the database of the Global Trade Analysis Project (GTAP), version 5 (Dimaranan and McDougall, 2002). The model includes a representative private household in each country, which saves a proportion of disposable income and buys consumption goods. The household owns the firms in that country but also works for them, receiving wages and distributed profits, and there are also transfers (which may be negative) from the government. The country’s government collects trade and other taxes and spends all its tax revenues on government current expenditures or transfers to households. A capital account by country collects savings and buys investment goods. Producers within a country/region are aggregated into one representative firm for each sector, which produces the respective good or service, buying intermediate goods and hiring factors of productions. There are five factors of production: skilled labour, unskilled labour, capital, land and natural resources. In making production decisions, the firms choose the levels of labour and intermediate inputs to produce a single sectoral output, taking into account the price of sectoral outputs, the wage rate, the prices of intermediate inputs and the existent stock of capital. Sectoral outputs are either sold in the domestic market or exported to foreign markets.

212

E. Diaz-Bonilla et al.

The domestically produced and consumed good from each sector is different both from the export good generated in that same sector (with that differentiation captured through a CET function) and from the imported good corresponding to that sector (utilizing a CES function). The composite export and import goods from each sector are also differentiated by country of origin/destination based on constant elasticity functions acting as aggregators. Domestic and world markets for goods and services equilibrate through changes in endogenously determined prices. Domestic production and consumption prices interact with world prices, the exchange rate per country, different levels of border protection, and other consumption, production and export taxes or subsidies. Factor markets also adjust through the interaction of demand, supply and prices. In the simulations the supply of all factors of production other than labour is kept at the base levels, and there are no changes in intercountry savings and investments flows. Following a long tradition in development analysis (see for instance Lewis, 1954) labour markets in developing countries are run with endogenous (un)employment and nonflexible consumption real wages.14 This approach differs from many exercises with this class of world trade models that consider full employment with flexible wages as the equilibrating variable. Returns to factors of production (including wages) may vary across sectors in the same country due to other imperfections in their markets that are modelled as invariant to the policy experiments discussed here. The US nominal exchange rate is fixed at 1; therefore, the US dollar is the world numeraire, and world prices are expressed in US dollars. Every country has its own nominal exchange rate, which is allowed to float, and also a country-specific numeraire price index of domestic goods. Changes in the nominal exchange rate in a country affect the real exchange rate (defined as the ratio of the prices of traded goods to non-traded goods in each country/region), and there is a functional relationship between the real exchange rate and the trade balance in each country. A specific feature of the model utilized here is that it includes a parameter of technological change in the production functions of different primary agricultural products, which, in turn, depends on government expenditures in agricultural research and the returns to that research. The (value-added) production functions in the primary agricultural sector relate sectoral value-added Qt to the use of factors of production (e.g. labour, capital, natural resources),15 and the level of total factor productivity (TFP): Qt = TFPt  F1(Lt, Kt, NRt, INPt).

(1)

In turn, TFPt and its evolution over time is considered a function of government expenditures in agricultural technology and the return from those expenditures as TFPt = TFPo  (1 + g)t , and

(2)

g = (IRR/100)  (GovExpR&D/AgGDP),

(3)

where g is the rate of growth (divided by 100) of TFP; IRR is the internal rate of return of public expenditures in agricultural R&D; GovExpR&D/AgGDP

Food Security Boxes versus Investments in the Green Box

213

is the ratio of government expenditures in agricultural technology over agricultural value added; and TFPo is the constant calculated at the base year (which embodies changes in TFP up to that time). The relationship between g, IRR and GovExpR&D/AgGDP assumes that agricultural products are tradables in a small open economy. Then the change in the economic surplus for the economy due to expenditures in R&D can be approximated by the producers’ surplus (Alston et al., 1995). To calculate the IRR, Alston et al. (2000) consider that there is a perpetual flow of benefits B per year,16 and that the costs are approximated by a one-time expenditure C in year t0, so the formula B/IRR = C applies. Based on those criteria, a relationship between g, IRR and the ratio of government expenditures over agricultural value added can be derived.17 The values of the IRR are those that correspond to the averages reported in Alston et al. (2000), for each of the geographic regions considered in the model. The ratio of R&D expenditures over agricultural GDP comes from ASTI/IFPRI/ISNAR (http://www.asti.cgiar.org/index.cfm). The production functions, as all equations in the model, are calibrated to the base year data. In the simulation, total public expenditures in agricultural research are increased exogenously. That additional cost is financed with an income-tax equivalent, so that the balance of the public budget is not changed. For every year of expenditures, there is a corresponding increase in the level of production according to the values of the IRR reported in Alston et al. (2000), which accumulate over time.18

Simulations The simulations consider two scenarios. As indicated, several supporters of the DB/FSB consider that adjustments in protection for special products should be the main instrument in the new Boxes. Therefore, in Scenario 1, protection is increased by an arbitrary amount (50 per cent) on grains (cereals) that are usually included among the ‘food security crops’. This increase in protection is simulated only for those countries that supported the notion of a DB/FSB. The list of countries/regions considered as they appear in the GTAP database includes: Indonesia, Philippines, Thailand, Vietnam, Bangladesh, India, Sri Lanka, Rest of Southern Asia, Central America and Caribbean, Peru, Venezuela, Rest of Andean Pact, Turkey, Rest of Middle East, Morocco, Rest of North Africa, Botswana, Mozambique, Tanzania, Zambia, Zimbabwe, Rest of Southern Africa, Uganda and Rest of sub-Saharan Africa. In order to be able to compare that alternative with a different application of the resources transferred through protection, the total implicit tax on consumption resulting from higher protection is calculated as a price wedge between world and domestic prices of the products considered (which include both the explicit tax of higher tariffs on the imported component and the implicit tax of a higher price of the national component).

214

E. Diaz-Bonilla et al.

In Scenario 2, the value of the mostly implicit tax on households associated with protection is transformed into an explicit tax from the same amount now collected by the government. Those public funds are subsequently utilized to increase investments in agricultural technology for all primary products. Because the new tax revenue is equal to the increase in R&D expenditures the policy change is budget neutral. That tax and the new level of expenditures are then maintained over time.19 The simulation results are based on the accumulated incremental impact of those new levels of expenditures in agricultural research (i.e. the difference between the cumulated old and new levels of expenditures on agricultural R&D), over the same medium term considered for the first simulation on protection.20

Results The simulations are presented in Tables 10.1 to 10.5, where the scenarios are labelled ‘Protection’ and ‘R&D’. Those tables show the comparative static results between the base year and the configuration of the economic variables with the policy changes simulated. The increase in protection reduces world welfare (measured as equivalent variation in consumption) by about US$1 bn. As in similar cases, the countries increasing protection bear most of the welfare losses, here amounting to about US$790 m. The main negative impact occurs on countries in Northern Africa and the Middle East, which are large importers of cereals and have more problems in expanding domestic production to substitute for imports. But that increase in protection in the DB/FSB countries also reduces welfare for the rest of the world, both for other developing countries not participating in the increase in protection (a decline of about US$160 m) and for developed countries (a loss of US$230 m) (Table 10.1). While the consumption measure is the more adequate one in welfare terms, countries tend to look at trade negotiations from the perspective of production and employment. Therefore Tables 10.2 and 10.3 present changes in Gross Domestic Product at factor cost and labour employment after the increase in protection for the DB/FSB countries (first column). Although the changes are very small (just fractions of 1 per cent) they are almost universally negative, i.e. overall GDP and employment for the whole economy decline with increased protection, even though the protected sectors themselves expand (not shown here). This is in part related to the fact that other productive sectors using the protected crops as inputs (from livestock and poultry to flour mills and bakeries) are affected by higher costs of inputs.21 In addition, the increase in the cost of food – with sluggish real consumption wages – leads to increases in the real production wages (i.e. nominal wages divided by the sectoral production prices) for different activities. This may reduce the competitiveness of export-oriented and import-substituting manufactures (such as textiles and others), affecting also production and employment there. It is well known that one of the historical

Food Security Boxes versus Investments in the Green Box

215

Table 10.1. Welfare effect (equivalent variation, billion US dollars change from the base).

Eastern Asiaa India Rest of Southern Asia Central America and South Americab Turkey Northern Africa and Middle East Rest of Africac Development Box Countries All developing countries All developed countries World

Protection

R&D


0.07
0.04
0.03 0.00
0.01
0.63
0.01
0.78
0.84
0.23
1.07

0.12 0.05 0.10 0.07 0.03 1.29 0.05 1.70 1.76 0.40 2.16

Notes Indonesia, Philippines, Thailand, Vietnam. b Includes Central America and Caribbean, Venezuela, Peru and Ecuador. c Does not include South Africa. a

Table 10.2. Effect on GDP (factor cost, % change from the base).

Indonesia Philippines Thailand Vietnam Bangladesh India Sri Lanka Rest of Southern Asia Central America and Caribbean Peru Venezuela Rest of Andean Pact Turkey Rest of Middle East Morocco Rest of North Africa Botswana Mozambique Tanzania, United Republic of Zambia Zimbabwe Rest of Southern Africa Uganda Rest of sub-Saharan Africa

Protection

R&D

0.00
0.12
0.02
0.13
0.01
0.02
0.11
0.05
0.02
0.04
0.03
0.02
0.02
0.22
0.02
0.03
0.06
0.01
0.03
0.01
0.09
0.01
0.03
0.02

0.01 0.16 0.00 0.04 0.07 0.02 0.41 0.03 0.09 0.06 0.02 0.01 0.02 0.28 0.21 0.35 0.09 0.00 0.07 0.01 0.03 0.02 0.14 0.04

216

E. Diaz-Bonilla et al. Table 10.3. Total labour (% change from the base).

Indonesia Philippines Thailand Vietnam Bangladesh India Sri Lanka Rest of Southern Asia Central America and Caribbean Peru Venezuela Rest of Andean Pact Turkey Rest of Middle East Morocco Rest of North Africa Botswana Mozambique Tanzania, United Republic of Zambia Zimbabwe Rest of Southern Africa Uganda Rest of sub-Saharan Africa

Protection

R&D

0.00
0.09
0.03
0.13
0.01
0.02
0.10
0.05
0.01
0.01
0.01 0.00
0.01
0.17
0.02
0.02
0.05
0.01
0.01 0.00
0.08 0.00
0.02
0.01

0.01 0.12 0.00 0.04 0.08 0.02 0.38 0.03 0.07 0.03 0.01 0.00 0.01 0.22 0.21 0.34 0.08 0.00 0.05 0.01 0.01 0.01 0.12 0.03

reasons why many developing countries shifted internal terms of trade against agriculture was to maintain the cost of food and, therefore, of salaries, low enough to foster industrialization. While history shows that such strategy had heavy costs for the economy, the agricultural sector and for poverty alleviation (Little et al., 1970; Balassa, 1986), doing the opposite now forcing the terms of trade against industry does not seem the best way of inducing a sustainable, pro-poor development strategy. In addition to the production and employment effects, another reason usually invoked for increased protection is food security. Tables 10.4 and 10.5 present the results of the simulations for other variables that have been utilized in different studies of food security at the country level (see for instance Diaz-Bonilla et al., 2000): food consumption in real terms, and the ratio of food imports over total exports. Certainly, from the point of view of food security, it is more relevant to look at consumption, and not production, of food. For instance, Smith and Haddad, 2000, have shown that, along with other factors (such as women, education and health care), increases in total national availability of food (which include imports along with domestic production) have a positive impact on the decline of child malnutrition. With increased protection, however, and although there is more production of the goods protected (not shown in the tables here),

Food Security Boxes versus Investments in the Green Box

217

Table 10.4. Effect on food consumption in real terms (% change from the base).

Indonesia Philippines Thailand Vietnam Bangladesh India Sri Lanka Rest of Southern Asia Central America and Caribbean Peru Venezuela Rest of Andean Pact Turkey Rest of Middle East Morocco Rest of North Africa Botswana Mozambique Tanzania, United Republic of Zambia Zimbabwe Rest of Southern Africa Uganda Rest of sub-Saharan Africa

Protection

R&D

0.00
0.40 0.00
0.20
0.03
0.01
0.25
0.04
0.18
0.20
0.10
0.08
0.07
1.21
0.20
0.15
0.19
0.01
0.09
0.01
0.38
0.03
0.04
0.06

0.02 0.59 0.01 0.42 0.15 0.04 1.17 0.08 0.37 0.23 0.08 0.06 0.08 1.54 0.65 1.15 0.47 0.02 0.13 0.03 0.18 0.07 0.29 0.12

there is less consumption and utilization overall, because prices increase and imports are curtailed (Table 10.4). Even though, as before, the reductions are not large in percentage terms, except for the cases of the Middle East and Botswana, a conclusion is that food security may be compromised with increased protection (see also Sumner, 2000). Table 10.5 presents the ratio of food imports to total exports, indicating the trade effort that a country must make to finance its food imports. This ratio, which shows large variation across countries, is a better indicator of trade vulnerability than the ratio of food imports to exports that has been utilized in the WTO to define the Net Food Importing Developing Countries (see Diaz-Bonilla et al., 2000). For instance, Venezuela, an NFIDC that is a large net importer of food items, has none the less a low ratio of about 5 per cent (below the average for developing countries), mostly because it is a large exporter of oil, while Peru, also an NFIDC, has more than double that ratio (about 12 per cent). The simulations show that protection reduces the ratio in all countries (some changes are too small to be reflected in just one decimal point), but as argued before, this happens because imports, and consumption, are compressed, while a more desirable outcome for food security would be a decline in that ratio resulting from increased total exports.

218

E. Diaz-Bonilla et al. Table 10.5. Ratio of food imports/total exports (%).

Indonesia Philippines Thailand Vietnam Bangladesh India Sri Lanka Rest of Southern Asia Central America and Caribbean Peru Venezuela Rest of Andean Pact Turkey Rest of Middle East Morocco Rest of North Africa Botswana Mozambique Tanzania, United Republic of Zambia Zimbabwe Rest of Southern Africa Uganda Rest of sub-Saharan Africa Rest of World All developing countries All developed countries World

Base

Protection

R&D

5.5 5.9 3.4 4.0 16.3 4.2 11.0 16.2 11.0 12.5 5.3 8.8 6.1 6.9 10.2 15.5 6.7 18.9 18.3 1.7 4.8 7.3 5.1 8.1 9.6 6.6 4.2 5.0

5.5 5.8 3.4 4.0 16.1 4.2 10.7 16.2 10.9 12.1 5.2 8.7 6.0 6.7 9.8 15.2 6.6 18.8 17.9 1.7 4.7 7.3 4.9 8.0 9.6 6.6 4.2 5.0

5.5 5.9 3.4 4.0 16.3 4.2 10.9 16.2 11.0 12.4 5.3 8.8 6.1 6.9 10.2 15.5 6.7 18.9 18.3 1.7 4.8 7.3 5.1 8.1 9.6 6.6 4.2 5.0

Tables 10.1 to 10.5 show the comparative static results of transforming the implicit consumption tax of protection (mostly collected by the private sector) into an explicit tax that the government allocates to R&D in agriculture. World welfare increases by about US$2.2 bn, and welfare in DB/FSB countries increases by about US$1.7 bn, mostly due to changes in the Northern Africa and Middle East region. Still, the rest of the DB/FSB countries also benefit from an increase in welfare of more than US$300 m (Table 10.1). Tables 10.2 and 10.3 show increases, although small, in GDP and employment. Food consumption also increases, with somewhat larger gains in Sri Lanka, Rest of Middle East and Rest of North Africa, followed by Philippines, Botswana and Morocco (Table 10.4), but the ratio of food imports to total exports does not change much from the base year (Table 10.5). All in all, investments in increased productivity show obvious gains in production, employment and food security. As has already been indicated, another concern raised by several developing countries has been the impact of increased trade protection because of the DB/FSB on South–South trade. Table 10.6 shows the results

Food Security Boxes versus Investments in the Green Box

219

Table 10.6. Effect on value of total agricultural exports (change from the base in million US dollars). Protection Indonesia Philippines Thailand Vietnam Bangladesh India Sri Lanka Rest of Southern Asia Central America and Caribbean Peru Venezuela Rest of Andean Pact Turkey Rest of Middle East Morocco Rest of North Africa Botswana Mozambique Tanzania, United Republic of Zambia Zimbabwe Rest of Southern Africa Uganda Rest of sub-Saharan Africa Development Box Countries All developing countries All developed countries World


1.1
5.5
12.2
21.0
0.5
31.6
2.2
16.0
6.6
3.6
0.5
1.8
5.1
22.0
2.9
3.9
0.1
0.1
1.0 0.0
5.2
0.1
0.5
3.2
146.4
305.8
575.9
881.7

of the simulations: agricultural trade among developing countries, including those applying the higher levels of protection, declines by about US$300 m, and overall agricultural trade falls by some US$880 m. The larger declines are for developed countries (about 580 million), and for other developing countries like Argentina, China and South Africa (not shown here), whose exports drop by more than US$200 m. But even those countries applying higher protection lose trade by about US$150 m – mainly grains exporters in the Asian region such as Thailand, Vietnam and India, affected by protection in other markets (demand side effect), but there are also declines in exports of other agricultural products because land has been reallocated to the protected crops (supply side effect). The simulations show that permanent increases of protection for special products (as opposed to the need for temporary policy instruments for developing countries to confront sudden negative trade shocks and unfair trade practices) may well end up negatively affecting poverty (through the

220

E. Diaz-Bonilla et al.

employment effect) and reducing food security (due to reduced availability and higher prices of food products), while at the same time reducing trade opportunities among developing countries.

Some Implications for Policies and Negotiations Developing countries need to carefully analyse their own agricultural policies and possible alternatives, and base their requests for SDT on that analysis. For years, many of them have discriminated against agriculture, and currently, although the most obvious macroeconomic biases may be gone (Jensen et al., 2003), a large percentage still do not invest enough in agriculture and rural development. At the same time, several developing countries have indicated concerns during the current WTO agricultural negotiations that further trade liberalization could create problems for their large and predominantly poor agricultural populations. Poor countries have argued for a slower pace in reducing their tariffs on the premise that industrialized countries should first eliminate their higher levels of protection and subsidization. A related concern is how to manage sudden negative impacts, such as subsidized exports or import surges. Poor producers may see their livelihoods irreparably damaged by unfair trade competition and drastic shocks if, for instance, they are forced to sell productive assets or take children from school (see Lipton and Ravallion, 1995). The concerns raised by developing countries regarding the presence of significant distortions in world markets, and the need to protect vulnerable groups from negative shocks, are important issues that must be addressed. The need to limit protectionism and subsidies in developed countries, the requests for longer transition periods, and the design of some policy instruments in the WTO framework that are better tailored to poor countries’ capabilities to manage unfair trade practices and shocks, all seem compelling. However, there have also been arguments for even further agricultural protection in developing countries to ease poverty and promote food security. As discussed before, sometimes this suggestion is accompanied by the argument that protection ‘does not cost money’ and is easier to implement in poor countries than alternative policies such as investments in agricultural technology, extension and infrastructure. According to this argument developing countries need ‘flexibility’ to change levels of protection. The HP, the revisions in Annex A of the draft Cancún Ministerial Text that remained un-adopted, the July Package of 2004 – and the developments since then – all suggest that the notions of a new safeguard mechanism and of special products – linked to food security, rural development or livelihood security considerations – will be a feature of the final agreement. But, depending on how the operation of instruments for contingent protection are defined, there is the risk that they may be utilized in ways that lead to high and persistent protection. In that case, as the

Food Security Boxes versus Investments in the Green Box

221

simulations of the first scenario suggest, development, poverty and food security goals in the countries applying them may be affected negatively, while at the same time negating welfare-enhancing trade opportunities for other developing countries. However, in assessing both sets of simulations presented above it must be borne in mind that the results refer basically to increases in permanent protection, and do not necessarily invalidate the need for temporary policy instruments for developing countries to confront sudden negative trade shocks and unfair trade practices. Important technical work remains to be done, to ensure that the new instruments really address unfair trade practices and import surges, without becoming excuses for high and persistent protection. Experience teaches that trade protection has concrete costs, acting as a tax on food with an obvious negative impact on poor households, which in many developing countries spend more than 50 per cent of their incomes on feeding themselves (FAO, 1993), and is mainly received by bigger agricultural producers, which have larger quantities of products to sell. The problems faced by poor farmers and poor consumers are better addressed through policies and investments targeted to them directly. The focus should mostly be on vulnerable groups rather than on crops. Also, it is not necessarily true that the institutional requirements to run efficient and honest customs administrations that can adequately manage the stepped-up border measures suggested by some, are less exacting than organizing, for example, an efficient system of agricultural research and extension. Whatever the institutional requirements, it is obvious that the interventions allowed under the AoA without restrictions, such as research, extension, infrastructure and irrigation, to name a few, are the real foundations for increases in production, productivity and competitiveness, and for the reduction in poverty (IFAD, 2001). A related issue is the argument for increased flexibility. In trade, and in other negotiations, a party will usually try to limit other peoples’ options while attempting to retain flexibility for itself. But it seems dubious that developing countries be granted ample flexibility, while industrialized countries renounce theirs. Of course, in any balanced negotiation, all parties would become committed to some mutually agreed rules, even if they have different provisions for different parties. Developing countries, as weaker players in the global arena, need an international legal system that limits the ability of larger countries to act unilaterally. Moreover, there are arguments why some lack of flexibility may be beneficial to developing countries (Oyejide, 2000). First, the implementation of internationally negotiated rules may limit the power of special interests and arbitrary government measures within developing countries, helping to strengthen domestic legal and institutional frameworks (Diaz-Bonilla, 2000). Second, it has been shown that investment is in part related to the stability and certainty of the policy framework (Campos et al., 1999). A legal framework, internationally sanctioned, that limits flexibility and uncertainty, should help investment. The best approach for developing countries’ own policies is to eliminate biases against the agricultural sector in the general policy framework and to

222

E. Diaz-Bonilla et al.

maintain a neutral trade policy reducing protection overtime, while fully using transition periods negotiated in the WTO to increase investments in human capital, land tenure, water access, technology, infrastructure, nonagricultural rural enterprises, organizations of small farmers and other forms of social capital and political participation for the poor and vulnerable. None of these policies is constrained under the WTO Agreement on Agriculture, and current negotiations seem poised to give even more policy room to developing countries in those areas. The claims for more protection out of concern for small farmers, while under-investing in rural development and poverty alleviation, would otherwise ring hollow.

Notes 1

Under the GATT before, and the WTO now, there is the recognition in the legal texts that developing countries have special needs, and required favourable treatment more than other WTO Members. Under the current WTO framework there are more that 145 SDT provisions in the different agreements and other legal documents (WTO, 2000: WT/COMTD/W/77, 25 October 2000). Some developing countries have argued, caustically, that the AoA and other WTO texts amount to substantial SDT for rich countries. See for instance WT/GC/W/442, September 2001. WTO negotiating documents will be cited with the full number or description in the main text, and will not be repeated in the references. 2 Agricultural trade among developing countries, at close to US$70 bn per year during 1995–2001, represented more than 40 per cent of all agricultural exports by developing countries, and about 20 per cent of total trade in agricultural products (not counting intra-European Union trade) (Diaz-Bonilla et al., 2003b). Those flows are also projected to continue to grow over time (Rosegrant et al., 2001). 3 Other requests for general exemptions were based on specific policy objectives (see, for instance, G/AG/NG/W/96, December 2000 by Mauritius), or when related to small farmers (see, for instance, G/AG/NG/W/130, February 2001 by Nigeria). A more detailed discussion of these proposals is in Diaz-Bonilla et al. (2003b). 4 For instance, exemption for developing country members to provide any minimum market access (G/AG/NG/W/102, January 2001, India); maintain current levels of bound rates on key staples (G/AG/NG/W/142, March 2001, African Group); possibility of rebalancing and raising low tariff bound during the Uruguay Round (G/AG/NG/W/102, January 2001, India). 5 For the new special safeguard see G/AG/NG/W/102, January 2001, India; G/AG/NG/W/13, June 2000 by 11 developing countries, and G/AG/NG/W/142, March 2001, African Group. 6 See ‘Special and Differential Treatment for Developing Countries: Transitional Instruments to Expeditiously Countervail Subsidized Imports (SDCM)’, Special Session of the Committee on Agriculture. Informal Meeting, 24–26 September 2001, by Argentina, Bolivia, Paraguay, the Philippines and Thailand. 7 The HP is not discussed in detail here. For a more in-depth presentation see Ruffer and Swinbank (2003), FOI (2003), Diaz-Bonilla et al. (2003b) and Chapter 7, this volume. 8 The emergence of this group of over 20 developing countries can be considered one of the main political developments during the Cancún Ministerial. Originally, it included Argentina, Bolivia, Brazil, Chile, China, Colombia, Costa Rica, Cuba, Ecuador, El Salvador, Guatemala, India, Mexico, Pakistan, Paraguay, Peru, the

Food Security Boxes versus Investments in the Green Box

223

Philippines, South Africa, Thailand and Venezuela. It later added other countries such as Egypt, but lost others, such as El Salvador and Costa Rica. 9 Instead of making countries designate in advance the SP products, other possibilities include setting some criteria that must be fulfilled, such as number of small farmers, importance in local diet, and so on; or limiting the total number of products that can be protected by the new safeguard at the same point in time, even though the general designation is left open, with or without some specific criteria as indicated. 10 The current Special Safeguard has a price and a volume trigger that allows the imposition of additional import taxes as a proportion of the current applied levels. For the volume trigger the higher import taxes apply only until the end of the year when it has been invoked. The price trigger operates on specific shipments, and the additional import tax can only be imposed on that shipment. 11 The HP also offered different new SDT provisions for domestic subsidies including changes in the Green Box (Annex 2 of the AoA) for public stockholding for food security purposes, and new Green Box measures for support for producers of staple crops, for small-scale/family farms, for payments for losses and land retirement, among others. The HP maintained Article 6.2 of the AoA, that allows the use of some domestic support for low-income and resource-poor producers (LIRP), and expands its coverage with additional provisions for diversification from products harmful for human health, for concessional loans, transportation subsidies to remote areas, on-farm employment subsidies for families of LIRP, conservation measures, marketing support programmes, quality and sanitary and phytosanitary regulations, competitiveness and marketing capabilities of low-income and resource-poor producers, establishment and operation of agricultural cooperatives, risk management of agricultural producers and savings instruments to reduce year-to-year variations in farm incomes. 12 For broader discussions of DB/FSB and related issues see Diaz-Bonilla et al. (2003a) and Ruffer et al. (2002). 13 Countries: Australia and New Zealand, Japan and Korea, Other Asian developed countries, USA, Canada, European Union, China, Indonesia, Malaysia, the Philippines, Thailand, Vietnam, Bangladesh, India, Sri Lanka, Rest of Southern Asia, Mexico, Central America and Caribbean, Colombia, Peru, Venezuela, Rest of Andean Pact, Argentina, Brazil, Chile, Uruguay, Rest of South America, Hungary, Poland, Eastern European countries that will join EC, Former Soviet Union, Turkey, Rest of Middle East, Morocco, Rest of North Africa, Botswana, Rest of SACU, Malawi, Mozambique, Tanzania, Zambia, Zimbabwe, Rest of Southern Africa, Uganda, Rest of Sub-Saharan Africa, Rest of World. Products: Paddy rice, wheat, cereal grains etc.; vegetables, fruit, nuts; oil seeds; sugar cane/beet; plant-based fibres, crops etc.; cattle, sheep and goats, horses, animal products etc.; raw milk; wool and silk; forestry, fishing, bovine meat products, meat products etc.; vegetable oils and fats; dairy products; processed rice; sugar products etc.; beverages and tobacco products; coal, oil and gas, minerals etc.; textiles, wearing apparel, leather products; wood and paper, petroleum, coal products; chemical, rubber and plastic products; mineral and metal products; motor vehicles and parts, transport equipment etc.; electronic equipment, machinery and equipment etc.; utilities, construction, services, government. 14 Nominal wages divided by the consumer price index. 15 Inputs are combined with value added in a Leontieff technology. 16 In practice, for the values of the IRRs considered, about 20 years of flows of benefits suffice for the numbers to converge to the formula utilized in Alston et al. (2000). Several studies have found considerable evidence of lagged impacts of research that expand up to 30 years, with peak values of impact between 10 and 20 years (see, for instance, Cox et al., 1996).

224

E. Diaz-Bonilla et al. 17 Under those assumptions the change in total economic surplus (∆TS) is the change in the producer surplus (∆PS), calculated as ∆PS = Pw  Qo  K  (1 + 0.5  K  ε) where Pw is the world price, Qo is the quantity produced before the change in technology, K is the vertical down shift of the supply curve (resulting from the technological innovation) expressed as a proportion of the initial price, and ε is the elasticity of supply. The vertical shift in the supply curve K (the reduction in cost for the same quantity produced) can be transformed into the percentage increase in production (Q1
Qo)/Qo, called J, under neutral technical change with fixed proportion and linear supply curves, as K = J/ε. Following the usual normalization of world prices to 1, assuming values for ε as in Alston et al. (1995), and combining the equation for ∆PS with the definition of IRR, we obtain g as function of IRR and the ratio of government expenditures to agricultural production. 18 For changes in agricultural productivity what is important is the accumulation of expenditures in research (stocks), by opposition simply to considering annual flows (Pardey and Beintema, 2001). 19 This implies, to make both scenarios comparable, that the higher import taxes in the first simulation are a permanent increase in protection and not a temporary safeguard. In that regard the simulations must be interpreted as comparing protection versus investment in R&D in a ‘thought experiment’, rather than evaluating specific proposals of the DB/FSB (which in several cases refer to temporary safeguards or slower reduction of protection over time). 20 About 10 years. 21 Some could argue that those sectors should also be protected and so on, sector after sector. But many productive sectors are non-tradable and then must bear the cost of the protected sectors. Another general equilibrium effect is that protection of import-competing sectors, which obviously reduces imports, ends up damaging exports through different channels (such as appreciation of the real exchange rate).

References Alston, J., Norton, G. and Pardey, P. (1995) Science under Scarcity: Principles and Practice for Agricultural Research Evaluation and Priority Setting. Cornell University Press, Ithaca, New York. Alston, J., Chan-Kang, C., Marra, M., Pardey, P. and Wyatt, T.J. (2000) A Meta-Analysis of Rates of Return to Agricultural R&D: Ex Pede Herculem? IFPRI Research Report 113, International Food Policy Research Institute, Washington DC. Balassa, B. (1986). Outward Orientation. World Bank Development Research Department Discussion Paper No. 148. World Bank, Washington DC. CAFOD, Action Aid, Oxfam and IATP (2002) An Introduction to the Development Box. A joint paper by CAFOD, Action Aid, Oxfam and IATP. (http://www.cafod.org.uk/ archive/policy/default.asp).

Campos, E.J., Lien, D. and Pradhan, S. (1999) The impact of corruption on investment: predictability matters. World Development 27(6), 1059–1067. Cox, T., Mullen, J. and Hu, W. (1996) Nonparametric Measures of the Impacts of Public Research Expenditures on Australian Broadacre Agriculture: Preliminary Results. Agricultural and Applied Economics, Staff Paper Series No. 399, August 1996. University of Wisconsin-Madison, Wisconsin. Diao, X., Diaz-Bonilla, E. and Robinson, S. (2002) Scenarios for Trade Integration in the Americas. Trade and Macroeconomics Division discussion paper No. 90. International Food Policy Research Institute, Washington DC. Diaz-Bonilla, E. (2000) Developing country perspectives on trade negotiating round. In: Diaz-Bonilla, E. (ed.) A Program to End

Food Security Boxes versus Investments in the Green Box Hunger: Hunger 2000. Tenth Annual Report on the State of World Hunger, Bread for the World Institute, Maryland. Diaz-Bonilla, E., Thomas, M., Robinson, S. and Cattaneo, A. (2000) Food Security and Trade Negotiations in the World Trade Organization: A Cluster Analysis of Country Groups. Trade and Macroeconomics discussion paper No. 59. International Food Policy Research Institute, Washington DC. Diaz-Bonilla, E., Thomas, M. and Robinson, S. (2003a) Trade, food security and the WTO Negotiations: some reflections on boxes and their contents. Chapter in Agricultural Trade and Poverty. Making Policy Analysis Count. OECD, Paris. (http.www.oldbook shop.org/oecd/display.asp Diaz-Bonilla, E., Robinson, S. and Gulati, A. (2003b) Heterogeneity, Convergences and Divergences: Developing Countries and the WTO Agricultural Negotiations. Paper prepared for the UN Trade and Finance Task Force of the Millennium Development Goals, International Food Policy Research Institute, Washington DC. Dimaranan, B.V. and McDougall, R.A. (2002) Global Trade, Assistance, and Production: The GTAP 5 Data Base. Center for Global Trade Analysis, Purdue University, West Lafayette, Indiana. FAO (Food and Agriculture Organization of the United Nations) (1993) Asia. Vol. 1: Compendium of Food Consumption Statistics from Household Surveys in Developing Countries. Food and Agriculture Organization of the United Nations, Rome. FOI (Danish Research Institute of Food Economics) (2003) Note on the Harbinson Draft on Modalities in the WTO Agriculture Negotiations. Agricultural Policy Research Division, Frederiksberg, Denmark. Green, D. and Priyadarshi, S. (2001) Proposal for a ‘Development Box’ in the WTO Agreement on Agriculture, CAFOD and South Centre. (http.www.caford.org.uk). Institute for Agriculture and Trade Policy (IATP) (2003) New WTO Agriculture Text Ignores Export Dumping and Developing Country Proposals. Press Release from the Institute for Agriculture and Trade Policy. (http.www.iatp.org.uk). International Center for Trade and Sustainable

225

Development (ICTSD) (2005) Agriculture Negotiations at the WTO: the July Package and Beyond. International Center for Trade and Sustainable Development, Geneva, Switzerland. International Fund for Agricultural Development (IFAD) (2001) The Challenge of Ending Rural Poverty. Rural poverty report 2001. Oxford University Press, New York. Jensen, H., Robinson, S. and Tarp, F. (2003) General Equilibrium Measures of Agricultural Policy Bias in Fifteen Developing Countries. Trade and Macroeconomics Division Discussion Paper No. 105, International Food Policy Research Institute, Washington DC. Lewis, W.A. (1954) Economic Development with Unlimited Supplies of Labour. Manchester School of Economic and Social Studies, 22. (www.blackwell-synergy.com/ toc/name/72/6). Lipton, M. and Ravallion, M. (1995) Poverty and policy. In: Behrman, J. and Srinivasan, T.N. (eds) Handbook of Development Economics, Vol. 3, North-Holland, Amsterdam. Little, I., Scitovsky, T. and Scott, M. (1970) Industry and Trade in Some Developing Countries. Organization for Economic Cooperation and Development. Oxford University Press, Paris. Oyejide, A.T. (2000) Interests and Options of Developing and Least-developed Countries in a New Round of Multilateral Trade Negotiations. G-24 discussion paper No. 2, prepared for the intergovernmental group of twenty-four on international monetary affairs, United Nations Center for International Development, Harvard University, UNDTAD/GDS/MDPB/G24/2, New York and Geneva. Pardey, P. and Beintema, N. (2001) Slow Magic: Agricultural R&D a Century After Mendel. IFPRI Food Policy Report, International Food Policy Research Institute, Washington DC. Rosegrant, M., Paisner, M.S., Meijer, S. and Witcover, J. (2001) Global Food Projections to 2020: Emerging Trends and Alternative Futures. 2020 Vision Technical Report, International Food Policy Research Institute, Washington DC. Ruffer, T. and Swinbank, A. (2003) Stock-Take

226

E. Diaz-Bonilla et al.

of the WTO Agriculture Negotiations Implications for Developing Countries. Oxford Policy Management. (http.www. opml.co.uk). Ruffer, T., Jones, S. and Akroyd, S. (2002) Development Box Proposals and their Potential Effect on Developing Countries. Oxford Policy Management. (http.www. opml.co.uk). Smith, L.C. and Haddad, L. (2000) Explaining Child Malnutrition in Developing Countries: A Cross-country Analysis. IFPRI Research Report No. 111, International Food Policy Research Institute, Washington DC. Solagral (1999) Paper presented at the WTO Round and Food Security for USAID Partner Countries: An economic growth and agricultural development training workshop, 1–2 November 1999, Washington DC. Sumner, D.A. (2000) Agricultural trade policy and food security. Quarterly Journal of International Agriculture 39(4), 395–409.

UNCTAD (United Nations Conference on Trade and Development) (2000) Impact of the Reform Process in Agriculture on LDCs and Net Food-Importing Developing Countries and Ways to Address their Concerns in Multilateral Trade Negotiations. Background note by the UNCTAD secretariat, TD/B/COM.1/EM.11/2, Geneva. WTO (World Trade Organization) (2000) Implementation of Special and Differential Treatment Provisions in WTO Agreements and Decisions. Note by the Secretariat, WT/COMTD/W/77, World Trade Organization, Committee on Trade and Development, Geneva. WTO (World Trade Organization) (2004) WTO Agricultural Negotiations: The Issues, and Where we are Now. (http://www.wto. org/english/tratop_e/agric_e/agnegs_bkgrnd_ e.doc) Accessed December 2004.

ANNEX. What Boxes? The notions of a Development Box (DB) or a Food Security Box (FSB) emphasize different SDT provisions to protect and subsidize agricultural production in developing countries (‘playing defence’), but some proposals focus on the policies of developed countries (‘playing offence’), as well as some general, or systemic, issues. They are discussed below, in that order. Proposals regarding policies of developing countries (‘defence’)1 Many of the SDT provisions usually considered as part of the DB or the FSB refer to exemptions for domestic policies of developing countries. They can be divided into: (i) those that advocate general flexibilities for developing countries or a subset of them; and (ii) those that suggest changes in specific clauses in one or more of the three areas of market access, domestic support and export subsidies (the so-called ‘three pillars’ of the AoA) and/or other related topics. 1

The proposals discussed are followed by one or more examples of developing countries that raised the issue, but those countries mentioned are not necessarily the only ones that presented or supported the proposal mentioned in each case.

Food Security Boxes versus Investments in the Green Box

227

General flexibility or exemptions The first proposal for a Development Box mentioned above (G/AG/NG/W/13, presented by 11 developing countries2) suggested a general flexibility related to specific agricultural products (‘key staples’). Developing countries would be allowed to apply a ‘positive list approach’ declaring which agricultural products or sectors they would like included in the AoA provisions, and subject to AoA commitments.3 Others considered additional bases for flexibility in commitments. One was when developing countries were pursuing specific policy objectives, such as food security, poverty alleviation, rural development, rural employment and diversification of agriculture. The suggestion is that in those cases all measures taken by developing countries will be exempted from disciplines (see, for instance, G/AG/NG/W/96, December 2000, by Mauritius and poverty reduction measures). Another proposals for general flexibility emphasized measures protecting and subsidizing small farmers, (see, for instance, G/AG/NG/W/130, February 2001, by Nigeria, which requests ‘flexibility regarding import restraint and domestic subsidy for the protection of, and support to, household subsistence farming and smallscale farming in countries where such farming is very important’). Domestic support During the current negotiations several developing countries have suggested additional expansions of the current Special and Differential Treatment for domestic support in addition to those already existing under the Agreement on Agriculture. Some of the main suggestions include: ● ●

●

2

Increasing the current de minimis for developing countries from 10 per cent to 20 per cent (see, for instance, G/AG/NG/W/124 by Sri Lanka). Maintaining and expanding Article 6.2 (for instance, excluding productspecific support given to low-income and resource-poor farmers from the AMS), and ensuring that the support given according to that Article is protected from countervailing duties and other measures as discussed in Article 13 (Due Restraint Clause, also called ‘Peace Clause’) (see among others G/AG/NG/W/130, February 2001, by Nigeria and G/AG/NG/W/55, November 2000, by ASEAN countries). Exempting from reduction commitments measures taken by the developing countries for poverty alleviation, rural development, rural employment and diversification of agriculture (G/AG/NG/W/102, January 2001, by India) or for food security reasons or to help small farmers (G/AG/NG/W/130, February 2001, by Nigeria); more

Cuba, Dominican Republic, Honduras, Pakistan, Haiti, Nicaragua, Kenya, Uganda, Zimbabwe, Sri Lanka and El Salvador. 3 Conceivably, a negative list could also be utilized, with products being excluded from disciplines.

228

E. Diaz-Bonilla et al.

●

● ●

generally, the expenditures for development programmes may be excluded from being counted in the AMS (G/AG/NG/W/142, March 2001, by the African Group). Exempting from the AMS subsidies by developing countries for the purchase of food products for public stocking (G/AG/NG/W/130 February 2001, by Nigeria). Allowing the compensation in the AMS of products with positive and negative specific support (G/AG/NG/W/102 January 2001, India). The possibility of notifying domestic support in a ‘stable currency/basket of currencies’ to take ‘into account the incidence of inflation and exchange rate variations’ (G/AG/NG/W/102 January 2001, India).

Market access A broad exemption has been already discussed under general flexibilities, in the form of a ‘positive list’ of products (G/AG/NG/W/13, June 2000, by 11 developing countries) declared by a country to be subject to AoA commitments (while those not mentioned would be excluded). Other developing countries have linked further reductions in their own tariffs to reduction or elimination of distortions in international markets, due to protection and subsidies in industrialized countries (G/AG/NG/W/ 102 January 2001, by India). Also, several developing countries indicated that future tariff reductions should be based on final bound rates, and not on applied ones (see, for instance, G/AG/NG/ W/142, March 2001, by African Group). Other specific SDT provisions include: ●

●

●

●

Developing country members should be exempt from any obligation to provide any minimum market access (G/AG/NG/W/102 January 2001, India). Developing countries should be given the option of maintaining the current level of bound rates on key staples (G/AG/NG/W/142, March 2001, African Group). Longer time frames for phasing in changes than those negotiated during the Uruguay Round, considering that the periods selected then did not necessarily reflect the gap between developing and developed in the time needed to prepare for those changes (Special and Differential Provision, Non-Paper by African Group, Cuba, Dominican Republic, El Salvador, Honduras, Kenya, Pakistan and Sri Lanka, Special Session of the Committee on Agriculture, 4–6 February 2002). The possibility of rebalancing low tariff bindings in developing countries, including raising ceiling bindings for similar categories of products that were bound during the Uruguay Round (G/AG/NG/ W/102, January 2001, India).

Of special interest are two initiatives to streamline both current safeguards (against import surges) and countervailing mechanisms (against subsidized exports). Several countries suggested the need for a separate

Food Security Boxes versus Investments in the Green Box

229

safeguard, along the lines of the Special Safeguard provisions (Article 5 of AoA), which would be ‘available to all developing countries irrespective of tariffication in the event of a surge in the imports or decline in prices and to ensure food and livelihood security of their people’ (G/AG/NG/W/ 102, January 2001, India; see also G/AG/NG/W/13, June 2000, by 11 developing countries and G/AG/NG/W/142, March 2001, by the African Group). The other proposal is discussed below. Export subsidies Developing countries have proposed ways of dealing with export subsidies (and other subsidies) from developed countries, in a more expeditious way and without the restraints of the Peace Clause (Special and Differential Treatment for Developing Countries: Transitional Instruments to Expeditiously Countervail Subsidized Imports (SDCM), Special Session of the Committee on Agriculture; informal meeting, 24–26 September 2001, by Argentina, Bolivia, Paraguay, the Philippines and Thailand). In this proposal developing countries will be entitled to apply a countervailing measure on a simplified procedure to exports of products originating in developed countries without the need to prove either injury or a causal link between the subsidized import and injury. This measure would only be applied by developing countries to exports of products originating in developed countries that use or are entitled to use trade-distorting domestic subsidies or export subsidies. All trade-distorting domestic subsidies, including the AMS, the ‘de minimis’, and those under Article 6.5 of the AoA will be deemed, under a rebuttable presumption, to be trade distorting and specific. Subsidies falling under policy-specific criteria and conditions enumerated in Annex II of the AoA other than: (i) direct payments to producers; (ii) decoupled income support; and (iii) government financial participation in income insurance and income safety-net programmes (respectively, numerals 5, 6 and 7 of Annex II), will be exempted from the trade-distorting rebuttable presumption. The investigation indicated in Article 11 of the Agreement on Subsidies and Countervailing Measures will only need to establish the existence of a tradedistorting domestic subsidy or export subsidy. Developed members that, according to the their country schedules, are entitled to apply export subsidies, trade distorting domestic subsidies, including the de minimis, those under article 6.5, and/or those under numerals 5, 6 and 7 of Annex II (mentioned above), or that provide non-product specific domestic support will be deemed, under a rebuttable presumption, to be subsidizing the product to be countervailed. The proposal indicates that those provisions will be transitional, remaining in force for the duration of the reform process while developed members remain entitled to use trade-distorting domestic subsidies or export subsidies. At the same time, some developing countries have proposed: ●

The continuation of existing flexibility with respect to export subsidies in Article 9.4 (G/AG/NG/W/55, November 2000, ASEAN and G/AG/NG/W/102, January 2001, India).

230

E. Diaz-Bonilla et al. ●

Allowing the SDT given by Article 27 and Annex vii of the Agreement on Subsidies and Countervailing Measures (ASCM) to Least Developed Countries and to a group of developing countries (provided they do not reach GNP per capita of 1000 dollars per year), which exempts them from the prohibition of using subsidies that are contingent upon export performance (Article 3.1.a. of the ASCM) (G/AG/NG/W/102, January 2001, by India and G/AG/NG/W/107/Rev.1, March 2001, by Egypt).

Other issues and concerns regarding the SDT proposals in the DB/FSB Several of the proposals mentioned above came from Least Developed Countries, other low-income countries and/or developing countries that are net food importers. Other developing countries, mostly those that are part of the Cairns Group, along with other WTO members, raised concerns about the notions of the Development/Food Security Boxes, considering that they would create more distorted markets, going in the opposite direction to the objective of the negotiations, and will end up hurting trading relations among developing countries (whose share in total agricultural trade has been growing over time; see WTO Secretariat paper G/AG/NG/S/6). In addition to the possible content of the boxes, there have also been discussions about their potential users, i.e. what are the countries that may be allowed to use the measures included in the Development/Food Security Boxes. In general, developing countries have opposed extending those exceptions, such as the ones based on food security, to developed countries, even when couched in terms of the multifunctionality of agriculture (see, for instance, India, G/AG/NG/W/102). Within the developing countries some have argued that if a Development Box is created it should apply equally to all developing countries, while others maintain that the issue of the possible users of the boxes should be part of the negotiation (WTO background paper on the negotiations, 2002).

Proposals regarding policies of developed countries (‘offence’) Although the notion of a DB/FSB strictly speaking, covered in the previous sections, is mostly defensive and relates to SDT for developing countries, some proposals have also included requests of changes in the policies of the developed countries under the notion of an expanded Development Box, or simply by emphasizing the need for having a true ‘development round’. In this regard developing countries have been far more unified in what they see as the required adjustments in the agricultural and trade policies of industrialized countries. The use of export subsidies has been widely criticized as unfair and disruptive of international trade. Therefore, many countries in the current negotiations argue that the special treatment of agricultural export subsidies should be eliminated, scraping the Peace Clause (Article 13) and placing them under the Agreement on Subsidies and Countervailing Measures (see among others India, G/AG/NG/W/102).

Food Security Boxes versus Investments in the Green Box

231

A related subject is the operation of state trading enterprises, which may require increasing disciplines and transparency on practices that may be equivalent to subsidies or dumping on the export side, or hidden trade barriers on the import side (G/AG/NG/W/104 + Corr.1, January 2001, by Argentina, Brazil, Paraguay, Uruguay, Bolivia, Chile and Colombia). Also mentioned has been the importance of integrating in a unified framework the disciplines related to the continuum of transactions involving agricultural products, particularly the interface of export subsidies with food aid and export credits (G/AG/NG/W/142, March 2001, by African Group and G/AG/NG/W/140, March 2001, by Jordan). At the same time, several developing countries have an interest in stricter disciplines on export taxes and export controls that may exacerbate price fluctuations in world markets and limit access to food (see for instance Democratic Republic of Congo, G/AG/NG/W/135, March 2001 and G/AG/NG/W/140, March 2001, by Jordan). A second set of issues is the opportunities for expanded market access. Developing countries are requesting that industrialized countries reduce tariffs and tariffs peaks; eliminate tariff escalation; simplify complex tariff structures that include combinations of normal and ad-valorem tariffs (complexity which is compounded by seasonal adjustments in some cases); increase the volume of imports allowed under the current regime of tariffrate quotas (TRQs); and administer those TRQs in a more transparent and equitable manner (G/AG/NG/W/142, March 2001, by African Group; India, G/AG/NG/W/102, January 2001; Namibia, G/AG/NG/W/143, March 2001; Cairns Group, G/AG/NG/W/54, November 2000). Another issue of market access is the continuation or not of the Special Safeguard (SSG) established in the AoA. Several developing countries are asking that the general SSG be eliminated (G/AG/NG/W/107/Rev.1, March 2001, by Egypt), or that its use be prohibited to developed countries but, at the same time, that the SSG be made available to developing countries (G/AG/NG/W/13, June 2000 by 11 developing countries; G/AG/NG/W/142, March 2001, by African Group). In general, the SSG acts as a variable levy, is not transparent and has the potential to be very disruptive of trade. Although, in general, there is agreement among developing countries on the policy changes desired from industrialized countries on market access (with some nuances in the case of the SSG), there is an important area of differences: the treatment of existing special preferences.4 For example, Swaziland (G/AG/NG/W/95, December 2000) has argued that: 4

While the issues discussed refer mostly to existing preferences, some proposals also suggest that future market access be granted to low-income countries. For instance, India suggests that although Tariff Rate Quotas (TRQs) must be eventually abolished, in the meantime developed countries should expand them, make their filling mandatory, and allocate them with stricter application of the MFN principle, but ‘with special preference being given to developing countries having less than $1000 per capita annual income’ (G/AG/NG/W/102, by India).

232

E. Diaz-Bonilla et al. the current preferential market access arrangements enjoyed by small DCs should be protected under the current round of negotiations for agricultural trade reform. The period of protection should be sufficiently long to enable the small DCs improve their welfare by significant measurable margins. The margins can be negotiated in the context of the criteria listed above. One important area of the protection being recommended herein is guaranteed market access at guaranteed prices for sensitive products from small DCs over an agreed period of time. The period should be long enough for meaningful development and adjustment to occur.

(Swaziland further elaborates the issue in the Special Session of 3–4 December 2001 in a non-paper entitled ‘Trade Preferences – a Proposal for Small Developing Countries’; see also Namibia, G/AG/NG/W/143, March 2001). A third set of issues relates to domestic support. Many developing countries feel that industrialized countries have maintained a variety of exemptions for their own policies of domestic support, while they have dismantled or significantly reduced their own domestic support for agricultural producers, mainly because of fiscal constraints and as part of structural adjustment programmes supported by international financial organizations and aid donors. In an attempt to discipline further domestic supports, many developing countries have proposed to tighten the criteria for the Green Box, the reduction of the measure of support by product and the elimination of the exemptions considered under the Blue Box. Another suggestion has been to put a cap on all or a specially defined subset of domestic support measures as a percentage of the total value of agricultural production (see for instance India, G/AG/NG/W/102, January 2001; G/AG/NG/W/14, June 2000, by 11 developing countries). This proposal coincides with the initial US posture. The argument that a uniform limit defined in percentages would contribute to levelling the playing field that is now heavily tilted in favour of industrialized countries (which have the legal room under the WTO and the money to distort production and trade in their favour) seems compelling.

Systemic and other general issues Several developing countries have called for the creation of international financial facilities linked to food security and/or international or regional food emergency stocks. These proposals are linked to concerns about food security, especially regarding LDCs and NFIDCs (for instance, ‘Food Security Non-Paper at the Special Session of the CoA, Informal Meeting, 23–25 July 2001; Cuba, Dominican Republic, El Salvador, Honduras, Kenya, Nicaragua, Nigeria, Pakistan, Peru, Sri Lanka, Venezuela and Zimbabwe).5 5

Not only developing countries have asked for international food security stocks. Japan has also presented a proposal for international food stockholding (A new Framework for international food stockholding, Special Session, 23–25 July 2001, and A possible framework of International Food Stockholding, Special Session, 3–4 December 2001).

Food Security Boxes versus Investments in the Green Box

233

Other developing countries are concerned with the long-term decline in prices of commodities, and consider that during the WTO negotiations this issue should be addressed. Some of them have argued that Article XX of GATT lists international commodity agreements as a legitimate exception to normal MFN treatment, thus granting flexibility to Members to participate in those agreements (Article XX h), and that Article XXXVI, paragraph 4, of GATT explicitly considers the problems of developing countries depending ‘on the exportation of a limited range of primary products’, the need to provide ‘more favourable and acceptable conditions of access toward markets for these products’, and that ‘wherever appropriate to devise measures designed to stabilize and improve conditions of world markets in these products including measures designed to attain stable equitable and remunerative prices’ (Special Session of the Committee on Agriculture. Informal Meeting, 4–6 February 2002, Proposal on Trade in Agricultural Commodities and the Concerns of Single Commodity Exporters (SCEs); Communication from Zimbabwe on behalf of all the Africa Group). Related, or in parallel, to the issue of the decline in prices, several developing countries have asked for rules, possibly similar to those applicable to state trading enterprises, that also apply to private multinational companies with market power in agriculture markets. For instance, Nigeria, under ‘Disciplines on International Market Structure and Competition in Agriculture’ argues that the AoA objectives of addressing ‘distortions in world agricultural markets while taking into account nontrade concerns, including food security… can be realized unless the role of multinationals in these markets could be better understood and regulated’. The suggestions include that: questionnaire approach for state trading enterprises may be expanded to include multinational companies, with a requirement for notification of relevant information; … the WTO Secretariat should also review these companies as part of the Trade Policy Review process of Member countries; … FAO and/or UNCTAD should establish data banks containing comprehensive information on the major players in the world food system.

(G/AG/NG/W/130, February 2001, by Nigeria; see also Special Session of the Committee on Agriculture, Informal Meeting, 4–6 February 2002; Proposal on Trade in Agricultural Commodities and the Concerns of Single Commodity Exporters (SCEs); Communication from Zimbabwe on behalf of all the Africa Group). Another important issue mentioned by several developing countries in the negotiations has been food aid. A general concern is the provision of adequate levels of food aid, which have declined in recent years, and the avoidance of cycles that tend to reinforce, instead of counteract, situations of oversupply and shortages (i.e. the fact that there is excess food aid when world supplies are abundant and prices low, and lack of it when supply conditions are tight and prices high). Therefore, different proposals emphasized the importance of binding obligations on food aid and that those obligations be made counter-cyclical (i.e. they should increase in

234

E. Diaz-Bonilla et al.

periods of high prices) (Special Session of the Committee on Agriculture, informal meeting, 3–4 December 2002. Food Aid. Non-paper, by Cuba, Egypt, Grenada, Mauritius, Nigeria, Sri Lanka and Uganda). Several developing countries also asked that food aid should be made available in grant form; perhaps sourced from other developing countries; focused towards poor countries and social groups; and delivered in ways that do not displace domestic production in the receiving countries (G/AG/NG/W/142, March 2001, by African Group; Food Security. NonPaper at the Special Session of the Committee on Agriculture. Informal Meeting, 23–25 July 2001 Cuba, Dominican Republic, El Salvador, Honduras, Kenya, Nicaragua, Nigeria, Pakistan, Peru, Sri Lanka, Venezuela and Zimbabwe; G/AG/NG/W/143, March 2001, by Namibia). Then there are several requests to strengthen, expand and make binding the commitments, particularly under the Ministerial Decision on Measures in Favor of LDCs and NFIDCs, to provide substantially increased technical assistance (including the one needed to meet SPS requirements in importing countries), facilitate access to technology ‘required to assist the diversification and to build production and processing capacities for agricultural commodities’ and ‘for the development, strengthening and diversification of their production and exports basis’, and, in general, to implement development schemes to insure that countries ‘depending on a single commodity can find suitable means of development’. (Special Session of the Committee on Agriculture. Informal Meeting, 4–6 February 2002; Proposal on Trade in Agricultural Commodities and the Concerns of Single Commodity Exporters (SCEs); Communication from Zimbabwe on behalf of all the Africa Group.)

11

That Was Then but This Is Now: Multifunctionality in Industry and Agriculture

EUGENIO DIAZ-BONILLA* AND JONATHAN TIN† *Executive

Director for Argentina and Haiti, Inter-American Development Bank, 1300 New York Ave., NW, Room NE1137, Washington DC 20577, USA; †2000 Connecticut Avenue, #719, Washington DC 20006-1002, USA

Introduction One of the central debates in world agricultural policy has been how (and, for some, whether) to fully incorporate agriculture within the general framework of the WTO (see Josling et al., 1996). Under the previous GATT, agriculture operated with different rules. This separate treatment was in part attenuated during the Uruguay Round, but current WTO legal texts do not yet reflect a complete integration of agriculture within the general trade rules for goods. There are two different views on how to proceed. Some insist that agriculture should not be treated differently from other sectors, like industry, and therefore current negotiations should complete the integration of agriculture into the WTO framework. One of the main issues in this regard is related to export subsidies. In contrast to industrial goods, for which export subsidies are banned in the WTO legal framework, these subsidies are still allowed for agricultural goods (see the discussion in the section on agriculture in developed countries, below). Another view emphasizes the special role of agriculture, and wants to maintain a separate treatment. Usually, this view is linked to the notion of the ‘multifunctionality’ of agriculture, which has been presented, mainly by industrialized countries, as a new concept that must be considered in the design and implementation of agricultural policies (European Union, 1999; Ministry of Agriculture, Forestry and Fisheries, Japan, 1999). The basic idea is that agriculture, in addition to its direct products, also generates separate positive externalities that justify the intervention of governments to ensure their adequate supply. During the current agricultural negotiations,1 the European Union, Norway, Japan and South Korea, among others, have argued that © CAB International 2006. WTO Negotiations and Agricultural Trade Liberalization: the Effects of Developed Countries’ Policies on Developing Countries (eds E. Diaz-Bonilla et al.)

235

236

E. Diaz-Bonilla and J. Tin

multifunctionality should be part of the non-trade concerns alluded to in Article 20 and in the Preamble of the AoA, which must be taken into account during these coming negotiations. Other countries (basically the members of the Cairns Group and the USA) have opposed granting an independent role for multifunctionality in the conceptual framework of the negotiations (ABARE, 1999; USDA/ERS, 1999). Multifunctionality has already been the subject of one of the longest documents presented in the WTO negotiations by a collection of industrialized and developing countries (WTO, 2000). Many developing countries are still analysing whether the idea has something to offer them in terms of their negotiating positions and policy framework. While multifunctionality has been invoked to support agriculture in developed countries, a similar idea, although not called so at the time, was clearly behind support for industry in developing countries.2 Again in this case, the policy implication was that government intervention was required (through trade protection, subsidies and other special policies) to develop an industrial base that contributed to society more than what market valuations alone would suggest. This chapter, although it does not present a full account of either debate, will discuss some of the parallelisms in their theoretical frameworks, policy implications and economic and social impacts. The main objective is to clarify current policy issues for the agricultural sector in developing countries, highlighting possible consequences for the negotiating position of those countries in the WTO. The rest of the chapter is organized as follows. The following section briefly presents the economic framework under which externalities can be addressed through policy. The next two sections then review the externalities that have been invoked for industry (then) and agriculture (now), the policy debates, and the economic and social implications of the policies followed. The penultimate section discusses briefly the differential treatment under WTO for industry and agriculture, while the final section concludes by analysing some possible implications for the current WTO negotiations, particularly from the perspective of developing countries.

The Economic Framework Economic policies based on multifunctionality and industrialization have both made use of the concept of externalities (positive or negative) arising as by-products of normal market activity (Marshall, 1898) and the theory of public goods (Samuelson, 1954). The defining characteristic of externalities is the lack of markets to transact them, and that those by-products may have beneficial (or costly) impacts on parties other than the economic agents producing them, while their generation is based solely on the private value for the primary actors. Society may therefore be oversupplied with a bad externality (e.g. pollution from a tannery), or undersupplied with a good externality (e.g. workers with practical high-tech skills). Without some

Multifunctionality in Industry and Agriculture

237

form of collective action, there is no guarantee of socially appropriate levels of supply. But collective action does not necessarily mean government action. Here is where the notion of public goods comes in. This is a class of goods or services that are non-rival in consumption (consumption by person A does not take away from person B’s consumption) and nonexcludable (person A cannot prevent person B from consuming that product). Such goods or services are difficult to produce and allocate through markets and, in pure cases, not even through voluntary collective action. Therefore, a case can be made for government intervention. This is precisely what has been argued, both recently in regard to the agricultural sector, mostly in developed countries and, previously, in relation to the industrial sector, mainly in developing countries. According to proponents of support, the externalities emanating from industrial or agricultural production are numerous. Table 11.1 shows those commonly emphasized in the literature on industry and development/modernization (then), and the multifunctional character of agriculture (now). The economic literature addressed most of these issues early on, usually in the context of arguments related to adequate trade and development policies. Economic theory suggests defining the externalities clearly and identifying precisely how they are produced. Policy makers could then look for the intervention that most directly addresses the production (or suppression) of the postulated externality with least distortions to other markets (Bhagwati, 1971). In this approach, the first-best solutions prescribed would almost always be different from the measures, such as trade protection, that have typically been advocated. Assuming that the externality has been clearly identified, then the appropriate policy would be a subsidy (if it is positive) or a tax (if it negative), directly aimed Table 11.1. Some postulated (positive) externalities to agriculture and industry. Industry

Agriculture

Economics and technology Employment Economies of scale Skilled labour creation Promotion of an entrepreneurial spirit Technological development Knowledge externalities Capital accumulation Politics and society Nation-building National security Independence/self-sufficiency Social modernization Poverty alleviation Geography and environment Urbanization

Economics and technology Employment Politics and society Food security Reduction of rural–urban income disparity Cultural heritage Vital rural communities Social cohesion Geography and environment Scenic vistas Environmental protection Biodiversity

238

E. Diaz-Bonilla and J. Tin

at the externality. The farther away from the externality the postulated policy instrument operates, the larger the costs it may impose on other sectors of the economy. If, for instance, the externality is workers with hightech skills, a direct subsidy to firms for educating their workers would be more adequate than a subsidy on production of high-tech goods (that generates a production distortion without necessarily ensuring the desired outcome), which in turn is better than trade protection to firms producing high-tech goods (which creates a production and consumption distortion). More fundamentally, if a firm or industry can internalize and appropriate the externality (for instance, learning-by-doing by workers that can be retained), then there would not be a case for any outside intervention, particularly if there are financial markets that can finance the acquisition of those capabilities (as in the criticism of the infant industry arguments by Baldwin, 1969). Furthermore, the intervention needs a full accounting of benefits, but also of costs, in a general equilibrium sense. Many of those debates, which have been substantially settled in the economic literature on trade and development, have re-emerged, however, around the discussion about multifunctionality in agriculture.

Industry in Developing Countries Externalities and policies Many leaders and intellectuals in the developing world saw industrialization as intrinsically related to nation-building.3 Newly independent countries during the 19th and 20th centuries were hoping to break free from direct political and economic control by the colonial powers. That dependency was considered to be embedded in the productive structure of the developing countries, which supplied primary products to the colonial powers while having to import manufactures. Even in Latin American countries, which had become independent mostly in the 19th century and which, after World War II, had a relatively developed industrial base compared to other regions,4 the concern about unequal relations had strong resonance. Rather than the more obvious issue of direct political control, the argument, as elaborated by Prebisch (1950), and Singer (1950) was an economic one: contrasting market structures in developed countries (characterized by industrial oligopolies and strong unions) with those of developing countries (characterized by smaller firms and surplus labour) meant that the former could retain the benefits of technical progress, while the latter surrendered productivity gains through falling prices of their primary exports (hence the decline in the terms of trade). Besides nation-building and economic and political independence, industrialization was also associated with (and, in the stronger version, would cause) social modernization. Rural populations were supposed to lack entrepreneurial spirit and to appear bound by traditional culture and organization. Urbanization, on the other hand, was linked to progressive

Multifunctionality in Industry and Agriculture

239

attitudes and affiliations. This process was supposed to lead to an open and mobile society, eliminating the assignment of occupations by traditional criteria (sex, ethnicity, family status) (the most complete presentation of these arguments is probably by Kerr et al., 1964).5 Yet, for all the arguments regarding the military, political and social externalities of industrialization, most of the public policy discussion was conducted in economic terms. The main objectives of industrialization were growth, employment and elimination of poverty (see, for instance, Bhagwati 1993, on the sequence of Indian Plans), much as the current notion of multifunctionality emphasizes employment generation in rural areas.6 But certainly, to characterize those objectives of economic policy as externalities of industrialization (or agriculture) would be stretching the notion too far.7 The economic externalities of industrialization involved a different set of issues (Krugman, 1994): the interaction of economies of scale, pecuniary external economies, technological spillovers, backward and forward linkages and strategic complementarities. The combination of these elements suggested the existence of multiple equilibria and the need for some form of coordination, probably, but not only, through government intervention, to move from lower to higher levels of economic activity (Murphy et al., 1989).8 Another issue was macroeconomic stability. Although not specified as an externality, it was also clear that policy makers wanted to make the economy less vulnerable to external shocks and avoid macroeconomic crises through industrialization. It was assumed that, as the number of industrial firms increased, dependence on revenue from primary products would gradually be reduced, which was supposed to insulate the economy from external shocks and to protect against the losses implied by the postulated decline in the terms of trade (CEPAL, 1969). In summary, using the current term, the multifunctionality of industry appeared substantial: nation-building, political and economic independence, national security, modernization, development, technological advance, protection from external shocks, and so on. But, even accepting the multifunctional effects of industry, there was still a question about what were the appropriate policies to attain those effects for the developing countries. Much of the policy thinking, which coalesced around the notion that came to be called Import Substitution Industrialization (ISI), was based on trade protection and subsidies for industry oriented to the domestic market.9 According to the post-war development strategy, the role of agriculture was subordinated to the needs of the industrialization process. Different arguments were utilized to support this view. Quantitative historical analysis (for instance Kuznets, 1966), showed that agriculture declined in importance with the advance of economic development. Also, and especially in Latin America, different authors argued that agricultural production was inelastic to domestic prices, and that international demand was also inelastic with respect to international prices. Therefore, the imposition of taxes on agricultural products would not significantly diminish domestic production, and much of the tax would be paid by importing countries in the form of higher prices (CEPAL, 1969).

240

E. Diaz-Bonilla and J. Tin

Furthermore, over the medium to long term, deteriorating terms of trade suggested the need to diversify the productive structure through industrialization, instead of favouring agriculture. Consequently, from the 1950s to the early 1960s, the prevailing idea was to transfer resources from the agricultural (presumably a low-productivity sector) to the industrial sector (where it was assumed that resources would have higher productivity). The role of agriculture in development was considered one of transferring surpluses of labour, food (‘wage goods’) and raw materials, savings and surpluses of foreign currency, in order to support the development of the industrial sector and public infrastructure (Johnston and Mellor, 1961).

Second thoughts By the mid-1960s several concerns began to be voiced about the adequacy of this development strategy. Protection and subsidies to the industrial sector were damaging other sectors, such as agriculture. Schultz (1964), in an influential book, argued that the farmers in the developing countries were ‘poor but efficient’, reacting with economic rationality to changes in prices and incentives. If the agricultural resources were efficiently utilized, there were no gains to be made by the economy from transferring labour and savings to other sectors. The suggestion was to support the agricultural sector through technological development and human capital formation in rural areas. The Green Revolution of the 1970s and afterwards was based on the idea that there could be a technological solution to the rural problem. Different studies during the 1970s (Little et al., 1970; Balassa, 1971; Krueger, 1978) also criticized the strategy of development based on inwardoriented, import-substituting industrialization in terms of both long-run growth and efficiency aspects. The costs of inefficiency and lack of competitive incentives to productivity growth due to protection were considered higher than the ones associated with problems in international trade. The supply of agricultural products appeared to be reasonably elastic, as was international demand, and the terms of trade between industrial and agricultural products – after adjusting for quality, volume and other corrections necessary for establishing such a ratio – were not deteriorating (for an overview of those debates see Balassa, 1986b). Protected industries appeared to require (and strongly lobbied for) government support long after the intended period of ‘infancy’. Pervasive state intervention into capital markets made investment funds available only to the large, favoured firms and discouraged technical advance in other sectors. On the other hand, developing countries following an export-oriented strategy would benefit from greater flexibility, efficient allocation of resources, technological development, economies of scale and dynamic effects that could not be attained through reliance on the internal market alone (Balassa, 1986b). It was also argued that industrialization fostered through protectionism had generated an industrial structure more capital intensive than the resource

Multifunctionality in Industry and Agriculture

241

endowment of developing countries required. Therefore, poverty alleviation was impaired by policies that protected capital-intensive industrialization and discriminated against agriculture, generating less employment and an income distribution less equal than other development strategies would have allowed. This process of industrialization was accompanied by an uncontrolled process of urbanization, and the continuation and even deepening of poverty in rural areas. The obvious realization – that the poor in developing countries were concentrated mainly in rural areas – led to the conclusion that if poverty alleviation was to be an important objective of economic policy, then greater attention should be given to agricultural and rural development (Chenery et al., 1974). At the macroeconomic level, import-substitution protectionism appeared to have increased inflationary pressures (Krueger, 1981, 1984) and fostered unsustainable fiscal deficits, associated with state interventions, leading to recurrent macroeconomic crises. Contrary to expectations, the countries following inward-oriented policies appeared more vulnerable to external shocks, and more prone to balance of payments crises, which, when they occurred, tended to have a stronger negative impact on the economy (Balassa, 1984, 1986a).10 Import substitution was even criticized in non-economic terms. While in Asia, industrialization took place with domestic firms, in much of Latin America it was related to the expansion of multinational corporations. Critics from the left decried the increasing power of the international capital, and attributed different economic and social problems to the dominance of those international corporations (Frank, 1969, among others). From a very different perspective, what was called ‘neoclassical political economy’ also criticized the notion of government as a benign planner interested in aggregate national welfare (the implicit view of much of the proposals for state-led development). They pointed out the rent-seeking behaviour of actors, which a state-led environment allowed to flourish, with virtually any intervention creating an opportunity for privileges, waste and fraud (Bauer, 1972; Krueger, 1974). Resources were misallocated because decisions were influenced by those rent-seeking groups, whose activities also consumed resources from the private sector that could have been applied to more productive ends. But it appeared that the woes of the ISI strategy did not end there: developing countries seemed plagued by political problems, instability, military coups and human rights abuses. Industrialization was obviously creating a labour class and urban sectors that began to claim for a larger share of economic benefits and for political participation. Public and private sector wage increases related to industrialization and modernization strategies encouraged faster migration from farm to towns, demanding jobs and public services, and causing social unrest. When the economic limits of the ISI strategy (high levels of inflation, balance of payment crises) converged with social unrest, it was not unusual that developing countries suffered military coups against a civilian government accused of being too corrupt or too weak to control the economic and social crisis.11

242

E. Diaz-Bonilla and J. Tin

As Hirschman (1982) noticed, faith in the development consensus was badly damaged by a series of political disasters, ‘ranging from civil wars to the establishment of murderous authoritarian regimes’, and the ‘wholesale loss of civil and human rights’ that were perceived as ‘somehow connected with the stresses and strains accompanying development and “modernization”’. The accumulation of all these (true or alleged) negative impacts on society of the excessive support for industry led to a re-evaluation of the development strategy in many developing countries. It was considered that those countries would benefit by taking advantage of opportunities in international trade, eliminating the distortions created by extreme government intervention, allowing the price system to operate more freely, making sure that technology and investment reflected the endowment of human and other resources (thus avoiding the emphasis on capital-intensive enterprises), and positively reappraising the role of agriculture in the economy (see Balassa, 1971; Little et al., 1970). Countries in Asia, and some in Latin America, building on previous ISI stages, turned towards export-led strategies that generated many of the success stories of recent decades in terms of growth, industrialization, employment and poverty reduction.12 Industrialization strategies over time increased the share of developing countries in world industrial value added (Table 11.2), from 15–18 per cent during the period of ISI to less than 25 per cent in the 1990s (clearly lower than the share of industrial countries in world agriculture, a fact whose implications are discussed later). However, and although it can be argued that industrialization in developing countries remained relatively small in a world context, their export successes began to find trade limits in industrialized countries through some measures not necessarily GATTcompatible, such as voluntary export restraints and quotas (as in the case of textiles).13 At the same time, within GATT, there was a strengthening of the Anti-Dumping Agreement during the Kennedy Round (1964–1967), and of the code for subsidies and countervailing measures during the Tokyo Round (1973–1979). Domestic, and above all export, subsidies for industrial goods were tightly disciplined. Agriculture, however, followed a different route.

Agriculture in Developed Countries Historical perspective An obvious observation is that over the last 50 years or so, world economic policies affecting agricultural policies have differed markedly across countries: those policies tended to tax agriculture in developing countries (as discussed before), while this sector was clearly subsidized in industrialized countries (World Bank, 1986). In the USA, agricultural policies were a reaction to the Great Depression: the need to support farmers’ incomes and the family farms and to stabilize prices and supplies

Multifunctionality in Industry and Agriculture

243

provided the main rationale behind the policies followed (Orden et al., 1999). Although the world had changed substantially since the 1930s, the policies instituted then, with some variations, were in place until the 1995 Farm Bill, when a different approach was followed, only to reappear with a vengeance in the current 2002 Farm Bill (Orden, 2002). Modern agricultural policies in Western Europe and Japan emerged from somewhat different concerns. At the end of World War II, Europe was a devastated territory, suffering food deficits, threatened by internal turmoil and by the presence of the Red Army, stationed at the other side of the Iron Curtain. For the USA and the Western Allies, it was important to develop a strong regional economy to counter the presence of the USSR. At the national political level, farmers were supposed to be a conservative force that would provide some balance to left-leaning urban political parties and labour groups. The Common Agricultural Policy, enshrined in the Treaty of Rome that launched the then European Community in 1957, reflected in part those concerns. In Japan, not only the agricultural policies but also the agricultural structure were the result of the post-war developments. The victors of World War II implemented a sweeping land reform with the vision of a pacifist Japan, in which a political block of numerous small family farms would provide the basic constituency for democratic parties.14 In summary, both Western Europe and Japan were crucial strategic anchors in the Cold War era, the political support of agricultural groups was considered key to the pro-Western alliances at the national level, political stability required that food shortages were avoided, and the economic reconstruction of those war-ravaged economies needed a strong and growing agricultural sector. The strategy that followed, borrowing from the previous discussions, can be called import substitution agriculture (ISA), for which there were also some postulated multifunctional effects, although different from the ones emphasized currently. Policies that were based on reactions to the Great Depression of the 1930s, fears of food shortages in the 1970s and to the Cold War (that ended for all purposes with the breakup of the Soviet Union in 1991) were still in place when the Uruguay Round of the GATT (initiated in 1986) included the agricultural sector as a main priority. But, as with ISI in developing countries, protection embedded in ISA in industrialized countries has been capitalized into economic assets that their owners did not want to see devalued. On the political side, the agricultural sector in the developed countries still has significant clout: it provides the necessary votes that support powerful senior positions in the US Congress, may make the difference in tight political elections in the USA and Europe, and agricultural constituencies are a key component of the main political party in Japan (see, among others, Hayami, 1986, 1990; Rapp, 1988). Therefore, the integration of agriculture in the WTO framework was only partially achieved. The process was supposed to continue under the current Doha Development Round, but it has been resisted. It was against that background of pressures from trade negotiation that the current notion of multifunctionality emerged in developed countries, mostly in Europe and Japan.

244

E. Diaz-Bonilla and J. Tin

Multifunctionality in agriculture The arguments promoting the multifunctional role of agriculture have a similar structure to those utilized for industrialization. Farmers receive money for the food and fibre that they produce. But according to the advocates of multifunctionality, they bring numerous benefits to society as a whole for which they are not compensated. First, they provide rural jobs, considering that farm work is the principal form of employment in their rural areas. But the importance of agriculture goes beyond employment per se, to the extent that farmers provide some sort of social continuity and stewardship over a ‘cultural heritage’. Farmers would also create scenic vistas, offering a refreshing counterpoint to city life for urban dwellers. Farmland may also attract foreign visitors, thus adding to tourist receipts. Another positive externality of agriculture, it has been argued, is environmental and biodiversity protection. Compared to abandoning the land or (some) other alternatives, cultivated land may conserve soil and recharge groundwater, create flood controls, and provide habitats for rare species. Finally, there is the persistent notion that greater domestic food production equals greater food security. The whole notion of multifunctionality also rests on the idea of joint production: the emergence of these positive externalities is considered inseparable from the production of food and fibre. Moreover, it is argued that those externalities do not have markets, and cannot be provided by private collective action. Again, as in the case of industrialization, we have an argument for governments to intervene in order to support and protect the private producers who generate those externalities.

General assessment By now many of the arguments in favour of and against granting multifunctionality a separate role in agricultural policy-making and the negotiations have been widely covered in different publications arguing both sides of the policy debate (for a careful analysis of the economic arguments see Anderson, 2000; OECD, 2001). Most of the discussion so far has centred on whether the benefits of multifunctionality are jointly and inseparably obtained with agricultural production (and therefore if a country wants those benefits it has to support production, possibly distorting trade in the process), or whether the benefits have a separate existence (and thus can be generated through non-distorting measures, possibly those already considered in the Green Box of the WTO) (OECD, 2001). Criticisms point to the assumptions, the policy implications and the instruments utilized. For instance, showing that a productive sector (in this case agriculture, but similarly for others) has positive externalities for the rest of the society does not necessarily imply that production has to be especially encouraged beyond the level that it would normally have

Multifunctionality in Industry and Agriculture

245

attained under no intervention. If the multifunctionality effect is separate from production, then it is better to subsidize that effect directly, rather than production. If for some reason the postulated externalities emerge only as inseparable joint products with production, then there are still other issues to consider. For one, the sector may have negative externalities as well, such as damage to the environment and biodiversity due to widespread use of chemicals, high levels of animal manure originating from intensive livestock production, overgrazing leading to habitat loss and reduced biodiversity, and so on (see, for instance, European Environment Agency, 2001). Moreover, subsidizing a sector to make it expand beyond what would have otherwise been its normal level will increase its use of all types of resources from the economy, competing with other sectors. To the extent that some of those resources are not completely idle, costs of production will increase in the non-subsidized sectors, which may force them to contract. Then a cost–benefit analysis would be needed to assess whether the costs of encouraging a sector beyond its ‘natural’ level (in terms of the main products and of the externalities attached to them) may be larger than the benefits, considering the production and multifunctionality effects that may have been lost in other sectors. Different studies of agricultural policies in industrialized economies have shown the welfare costs they are imposing on the rest of the society (and also the world; see below), although not all claimed multifunctional effects have been factored into those studies. Simulations of reductions of agricultural support in industrialized countries show important benefits for those countries (see, for instance, Sharma et al., 1996 for the Uruguay Round; and Hertel et al., 2000 for the current negotiations). This is in part the result of consumers and taxpayers in rich countries receiving the equivalent of a tax cut in the form of less border protection and reduced fiscal transfers (or of being taxed with less distorting instruments, if import taxes are replaced with other sources of revenues). However, as agriculture represents a smaller fraction of the industrialized countries’ economies, the costs of the ISA now, although important, do not necessarily have the same relative relevance for those countries as the ISI had for developing countries in the past. This fact has important implications for how the costs and benefits of those policies are divided within and across countries (see below). Even if there is undersupply of net positive externalities for the society as a whole (considering the agricultural sector itself and the impact on other sectors), there is still the question of what is the best policy to foster those externalities. As shown before, most economic analysis would argue that the first-best alternative would most likely not be trade protection.15 Other criticisms, as in the case of the ISI, focus on the consequences of widespread government intervention for rent seeking and, eventually, corruption. The European Union’s innovative efforts to monitor payments made under the 1992 CAP reforms display the unavoidable enforcement

246

E. Diaz-Bonilla and J. Tin

problems associated with subsidies in practice (see, for instance, Commission of the European Community, 1999; European Union, 2000; OLAF/European Anti-Fraud Office, 2001). Even in the case of zero corruption, the influence of a well-informed and interested group of farmers, combined with (perhaps equally) interested bureaucrats dispensing the subsidy, will tend to produce a public good beyond its maximum net benefit for the society. Other important issues to evaluate ISA, also discussed in the context of the ISI, are equity and income distribution. One of the objectives of the agricultural policies in many industrialized countries has been to bring agricultural incomes in line with the rest of the population. By and large, this objective seems to have been attained – and perhaps exceeded – in some industrialized countries (OECD, 1999). But those policies are also reinforcing unequal income distribution patterns within agriculture, with agricultural support mostly received by the largest farmers (OECD, 1999). A better way would be to target income transfers to smaller farmers or lowincome farmers, but, both in the EU and the USA, measures intended to cap the amount transferred to large farmers have faced strong resistance.16 So far the discussion centred on the impacts on the same countries that claim the need to support their own agriculture to maintain its multifunctional effects. However, through the linkage of world markets, agriculture is affected globally through individual countries’ policies.

Whose multifunctionality? If the premise that multifunctionality is a joint product with agricultural production is provisionally accepted for the sake of argument, the immediate problem is: whose agricultural production levels are being supported and whose may be hurt in the process? Given some level of demand for food and agricultural products determined by income, prices, population and tastes, any attempt at expanding production in a group of countries on account of the multifunctionality effects would negatively affect agricultural production in countries that do not have the resources to subsidize agriculture. The global implications of ISA, given the larger weight of industrialized countries in world agriculture (about 40 per cent, clearly above the less than 25 per cent share of developing countries in world industry; see Table 11.2), are substantial. Different studies before the Uruguay Round, during the Round, and those now evaluating current negotiations, usually predicted substantial positive effects on developing countries’ income, production and exports of agricultural and agroindustrial products from an eventual reduction of tariffs and other forms of agricultural protection and subsidization in industrialized countries (Valdés and Zietz, 1980; Sharma et al., 1996; Hertel et al., 2000; Diao et al., 2003).17 Tables 11.3 and 11.4 show the percentage change in agricultural production (including primary and agroindustrial food products) that may

Multifunctionality in Industry and Agriculture

247

Table 11.2. Agriculture and industry value added (percentage). (From World Bank, 1986 – 1965 and 1985; World Development Indicators, 2001 – data for the 1990s are from World Bank, World Development Indicators, 2001 (in the case of developing countries) and from OECD, several issues (in the case of industrialized countries).)

Developing countries Agriculture value added share of world agriculture share of developing countries’ GDP Industry value added share of world industry share of developing countries’ GDP Industrialized countries Agriculture value added share of world agriculture share of industrialized countries’ GDPa Industry value added share of world industry share of industrialized countries’ GDPa a High-income

1965

1985

Average 1990s

59 29

61 20

61 14

15 29

18 34

24 24

41 5

39 3

39 2

85 40

82 36

76 30

OECD countries.

result in developing countries and transition economies from the elimination of protection and subsidization in developed countries, as calculated from Diao et al. (2003). Production increases by more than 23 billion dollars (in real terms) compared to the baseline, with the largest changes in Latin America and Asia (Table 11.3). The more disaggregated Table 11.4 shows that all the developing countries and regions increase their agricultural and agroindustrial production once developed countries eliminate their protection and subsidies. To the extent that the notion of multifunctionality has been suggested mainly by industrialized countries, which have the resources to utilize those subsidies, the result of such an approach may be more production and multifunctional effects in richer countries, and less of both in Table 11.3. Annual changes in incomes going to primary agriculture and agroindustrial production if industrialized countries liberalize their agriculture (increase in million dollars). (From Diao et al., 2003.) Region

Change

Sub-Saharan Africa Asia Latin America and the Caribbean Other Developing Countries* All Developing Countries

1,945 6,624 8,258 6,659 23,486

*Other Developing Countries includes transition economies.

248

E. Diaz-Bonilla and J. Tin Table 11.4. Potential impact of industrialized countries’ agricultural trade liberalization on developing countries in value (US$ m.) and percentages. (From Diao et al., 2003.) Annual increase in incomes going to primary agriculture and agroindustrial production

China Indonesia Malaysia Philippines Thailand Vietnam Bangladesh India Sri Lanka Other Southern Asia Mexico Central America/Caribbean Colombia Peru Venezuela Rest of Andean Pact Argentina Brazil Chile Uruguay Rest of South America Middle East Morocco Rest of North Africa Botswana South Africa and Rest of SACU Malawi Mozambique Tanzania Zambia Zimbabwe Other Southern Africa Uganda Rest of sub-Saharan Africa

Value

Percentage

2265.4 593.6 261.3 238.1 1755.0 81.9 43.6 1129.4 26.6 228.7 980.6 1531.8 339.7 363.3 100.8 212.0 1833.0 2258.7 240.8 154.5 242.7 1244.6 236.1 736.9 34.2 459.5 26.7 61.8 87.8 50.9 79.1 175.2 80.6 888.6

1.5 1.2 2.4 1.4 11.0 3.1 0.6 1.1 0.7 1.2 1.9 9.5 2.2 3.3 1.1 3.8 2.8 2.5 2.6 4.8 15.9 4.6 3.3 2.3 14.6 5.1 3.1 5.4 3.1 5.1 4.7 12.0 2.2 2.6

developing countries, which cannot afford such policies. Considering that agriculture and agro-industry are the main economic activities in many developing countries, particularly in poor ones, and that such activities usually have significant growth multipliers for the whole economy, the level of non-realized benefits (including the multifunctional effects) for those economies – as a consequence of the expansion of agriculture in subsidizing countries – may be significant.

Multifunctionality in Industry and Agriculture

249

In summary, the notion of multifunctionality does not solve the issue of distributive effects across countries: it simply adds another dimension. There is always the question of whose production and multifunctional effects expand or contract as a result of protection and subsidization.18

Agriculture and Industry in the GATT and WTO Frameworks Agricultural and industrial goods were treated differently in GATT (Josling et al., 1996), and although some of the most obvious differences have been narrowed during the Uruguay Round negotiations leading to the creation of the WTO, they are still subject to separate disciplines. A point to be noticed is that in note 2 to Article XVI, Section B, primary products are defined as ‘any product of farm, forest or fishery, or any mineral, in its natural form or which has undergone such processing as is customarily required to prepare it for marketing in substantial volume in international trade’. This implies that many of the ‘agricultural’ products treated differently, are in fact agroindustrial goods (meat, sugar, dairy products and so on). In fact, the differential treatment under GATT and the WTO is not only between primary agriculture and industry, but also between those industries based on agricultural raw materials and the rest of the manufacturing sector (Diaz-Bonilla and Reca, 2000). In Box 11.1 a summary is presented of the different provisions for agricultural and non-agricultural goods. It is clear that there are differences in quantitative restrictions (prohibited for industrial goods, allowed under certain circumstances for agricultural goods), safeguards (where a new special safeguard was created for agricultural goods), export subsidies (prohibited and with stricter dispute settlement measures for industrial goods, while allowed for agricultural goods and with some restrictions in the use of counter measures), and domestic subsidies (where the range of allowed subsidies for industrial goods is far more circumscribed than for agriculture and the possibilities for use of countermeasures and the dispute settlement option face less constraints for industrial than for agricultural goods). The differences result both from different substance in the disciplines themselves and, until 2003, from the operation of Article XIII of the Agreement on Agriculture (the ‘Peace Clause’). Without discussing the welfare implications of this asymmetry, at least in legal terms there appears to be an imbalance between what developing countries can do for their industry and what developed countries can do for their agriculture. Some have called the latter special and differential treatment for rich countries. Developed countries have maintained enough legal room under the WTO, and have the financial resources, to implement a variety of policies in agriculture. Developing countries, on the other hand, although also having legal room of manoeuvre on agricultural policies, often lack the needed financial resources to implement those policies, while at the same time face some constraints on how to defend

250

E. Diaz-Bonilla and J. Tin

Box 11.1. Differences in Disciplines. Non-agricultural goods Quantitative restrictions GATT’s original Article XI (General Elimination of Quantitative Restrictions) prohibited quantitative restrictions (quotas, import or export licences or similar measures) for goods in general. However, the same article included, in paragraph 2(c), an exception for ‘agricultural or fisheries product … necessary to the enforcement of governmental programmes for those products or to remove temporary surpluses’. Over time, other types of quantitative restrictions (voluntary export restraints, orderly marketing arrangements or similar export and import measures, referred to as to ‘grey area’ measures) have been increasingly utilized, affecting mostly industrial products. During the Uruguay Round ‘grey measures’ were prohibited in general, and it was agreed that those existing had to be phased out. Safeguards The Uruguay Round clarified the rules for Safeguards, the trade constraints that countries can use to protect any productive sector when it is threatened by an unexpected surge in imports that can cause injury to that sector (Agreement on Safeguards of the WTO). It requires proof of injury to the domestic producers from imports and the granting of compensations (i.e. opportunities for market access in other products that are equivalent in trade value to the trade reduced due to the safeguard). It can be applied for not more than 4 years, but it can be extended up to a maximum of 8 years, and has limits in the levels of tariffs or quantitative restrictions utilized. If quantitative restrictions are imposed, they should not reduce the quantities of imports below the annual average for the last 3 representative years (clear justification is given that a different level is necessary to prevent or remedy serious injury) (Agreement on Safeguards of the WTO). Export subsidies Probably the most important area of differences in treatment between agricultural and industrial goods is in export subsidies. In the GATT framework Article VI (Anti-dumping and Countervailing Duties) and Article XVI Subsidies, Section B, treat differently export subsidies for ‘primary products’ (but note the meaning of ‘primary’). Export subsidies, in general, are considered in the Agreement on Subsidies and Countervailing Measures (ASCM), building on a previous agreement that had been approved during the Tokyo Round. The WTO ASCM prohibits subsidies that are contingent upon export performance and those based on the use of domestic over imported goods. If a WTO member complains and it is found that the country accused is applying those subsidies, they must be immediately withdrawn. If this does not happen within the specified time period, the complaining member can withdraw equivalent trade concessions. The ASCM also exempts least-developed countries and developing countries that have less than 1000 dollars of income per capita from disciplines on prohibited export subsidies. For other developing countries, the export subsidy prohibition applies after 8 years.

Multifunctionality in Industry and Agriculture

251

Other subsidies Within the GATT framework Article XVI considered the possibility of subsidies in one country causing ‘serious prejudice to the interests of any other contracting party’ on export or import markets. In the Agreement on Subsidies and Countervailing Measures (ASCM) three categories of subsidies were defined: prohibited (export subsidies and those based on the use of domestic over imported goods); actionable; and non-actionable or permitted subsidies. The differences regarding export subsidies for agricultural and industrial goods have been discussed in the previous sections. But there is also a differential treatment in the case of the other two general types of subsidies, actionable and non-actionable. According to the ASCM no Member can use subsidies in ways that cause: (i) injury to the domestic industry of another Member; (ii) nullification or impairment of benefits accruing directly or indirectly to other Members; and (iii) serious prejudice to the interests of another Member (Article 5 of ASCM). But that Article immediately adds that it ‘does not apply to subsidies maintained on agricultural products as provided in Article 13 of the Agreement on Agriculture’. Subsidies are defined as financial contribution by the government, or income or price support (Article 1 of ASCM) , that benefit a specific (as different from generic support) industry, groups of firms, or firm (Article 2 of ASCM). Serious prejudice is presumed to exist under certain circumstances defined in the ASCM, particularly when the total ad valorem subsidization of a product exceeds 5 per cent. In those cases, the subsidizing member has to show that the subsidies in question do not cause serious prejudice to the complaining member (i.e. the burden of the proof is inverted). In general, actionable subsidies may lead to the imposition of countervailing duties on subsidized imports by the complaining WTO member when they affect a domestic industry. But in other cases (such as when subsidies are displacing exports of the complaining country in a third market), the disputes are referred to the dispute settlement process, and if it is determined that there are adverse effects, the subsidizing member must withdraw the subsidy or compensate the complaining party with access in other products. The third category of non-actionable subsidies include those that are not specific (as defined in Article 2), and three exceptions defined in Article 8.2(a), (b) and (c ): assistance to industrial research (up to 75 per cent of the costs) and pre-competitive development activity (up to 50 per cent), assistance to disadvantaged regions, and some environmental subsidies (all those exceptions have additional conditions to qualify as such). Still, if another member complains that a non-actionable subsidy is causing serious adverse effects to its domestic industry, the issue enters the process of consultation, negotiations and eventual dispute settlement. The ASCM also gives least-developed countries and developing countries that have less than 1000 dollars of income per capita 8 years to eliminate other prohibited subsidies. For other developing countries, the elimination of other prohibited subsidies after 5 years (7 for transition economies). There are also special provisions regarding the application of countervailing duties, and other remedies when the subsidizing WTO member is a developing country, in the case of privatization, and for countries in transition towards market economies. Continued

252

E. Diaz-Bonilla and J. Tin

Box 11.1. Continued Agricultural goods Quantitative restrictions For agricultural goods, the Uruguay Round negotiations determined that all border measures in agriculture other than ordinary customs duties (such as quantitative import restrictions, variable import levies, minimum import prices, discretionary import licensing, non-tariff measures maintained through state trading enterprises, voluntary export restraints and any other schemes) had to be transformed into tariffs, and then reduced over the 6-year implementation period by 36 per cent (simple average) with a minimum rate of reduction of 15 per cent for each tariff line. Developing countries could apply 2/3 of those rates of reduction. But tariff-rate quotas (a combination of quotas of high tariffs for the out-of-quota volumes) were allowed to guarantee current access and minimum access (where there were no significant imports, minimum access opportunities had to be offered, beginning in the first year of the implementation period with not less than 3 per cent of corresponding domestic consumption in the base period, and expanded to reach 5 per cent of that base figure by the end of the implementation period). Also there was an exception for the tariffication of sensitive products under strictly defined conditions, allowing the use of tariff-rate quotas for some specific products (originally only four countries applied for these provisions: Japan, South Korea, Philippines and Israel). Safeguards Agricultural goods have a another exception: the creation of a ‘special safeguard’ for those products that have complied with the ‘tariffication’ of previous non-tariff barriers. This trade remedy is different from the normal safeguard of Article XIX of the General Agreement. The ‘special safeguard’ for agriculture does not require either proof of injury or compensations: it allows the application of additional duties based on a price trigger (i.e. shipments at prices denominated in domestic currencies below a certain reference level) or a quantity trigger (i.e. imports surging above certain levels depending on the current levels of imports as a proportion of consumption). On the other hand, the special safeguard can be maintained only until the end of the year in which it has been imposed, and the import tax cannot exceed one third of the one applied when the safeguard was imposed (Article 5, Special Safeguard Provisions, of the Agreement on Agriculture). Export subsidies For agricultural goods, and although it was recognized that export subsidies have harmful effects on other countries (Art. XVI, Section B, paragraph 2), GATT contracting parties were only exhorted to ‘seek to avoid the use of subsidies on the export of primary products’, but if they utilized those subsides, countries should not apply them in a manner that ‘results in that contracting party having more than an equitable share of world export trade in that product’ (whose meaning was left undefined, and led to many discussions within GATT; see Josling et al., 1996). During the Uruguay Round, the differential treatment for agriculture was maintained: export subsidies were limited but not completely eliminated. WTO members agreed to cut agricultural export subsidies by 36 per cent in value and 21 per cent in quantity over the 6-year implementation period compared to the 1986–1990 base period level.

Multifunctionality in Industry and Agriculture

253

In the case of developing countries, the reductions are two-thirds of those of developed countries over a 10-year period (with no reductions applying to the least-developed countries). Developing countries are also exempted from cuts on subsidies to reduce the costs of marketing exports of agricultural products or internal transport and freight charges on export shipments, provided that they ‘are not applied in a manner that would circumvent reduction commitments’ (Agreement on Agriculture; Article 9, Export Subsidy Commitments, paragraph 4). There is some limited flexibility to carry unused levels of subsidies between years. While export subsidies for other products were prohibited and had to be eliminated if shown to exist, agricultural export subsides, if they are within the limits agreed in the negotiations (as defined in the country schedules of the AoA), receive a less drastic treatment if challenged by another WTO Member: they can be subject to countervailing duties only ‘upon a determination of injury or threat thereof’, and ‘due restraint shall be shown’ in initiating investigations (Article 13, Due Restraint – also called the ‘Peace Clause’ – point (c)). Also, they are exempted from other possible countermeasures based on nullification or impairment of concessions and/or serious prejudice (as defined in Article XVI of GATT 1994 and in Articles 3, 5 and 6 of the Subsidies Agreement). Other subsidies But while those rules apply to goods in general, agricultural products are again treated differently. Article 13(b) of the Agreement on Agriculture indicates that domestic support measures in the Green Box are: (i) non-actionable for purposes of countervailing duties; (ii) are exempt from actions based on claims of injury, nullification and impairment of concessions, and serious prejudice; and (iii) are also exempt from actions based on nonviolation nullification or impairment of the benefits of tariff concessions (this is a special claim that a country can make against other country, arguing that even though the latter did not violate any specific WTO rule, it is doing something that the first country claims is reducing or eliminating the value of trade concessions). Further, Article 13(c) also exempts domestic support measures under Article 6 of the AoA (which includes the Blue Box, used now mostly by the EU; the de minimis levels of support in developed and developing countries and subsidies given to low-income and resource-poor farmers and some other subsidies permitted for developing countries) from: (i) the imposition of countervailing duties unless there is injury (and due restraint shall be shown in initiating any countervailing duty investigations; (ii) from actions based on claims of injury, nullification and impairment of concessions, and serious prejudice, if the subsidies on a commodity basis did not exceed those budgeted for the 1992 marketing year; and (iii) from actions based on non-violation nullification or impairment of the benefits of tariff concessions, with, again, the limit of 1992 by commodity.

themselves from the agricultural policies of industrial countries. This situation is different with regard to industrial goods, where the policies aimed at subsidizing and expanding industries appear more tightly constrained, while there are more legal instruments to counter possible undesired practices from other countries. Certainly the adequate answer (in welfare terms) to this imbalance is not to ask for developing countries to have more legal access to distorting policy instruments for industrial development,19 but to move agriculture further toward full integration with WTO rules.

254

E. Diaz-Bonilla and J. Tin

That Was Then But This is Now This chapter has presented two cases where governments have decided to subsidize certain sectors based on their expected multiple contributions to public welfare. A notion of multifunctionality was clearly behind support for industry in developing countries, particularly in the first decades after World War II. But also, as in the case of multifunctionality of agriculture now, the problem was that extreme support for a sector had, as a counterpart, important negative effects on other economic activities and social groups, leading in several instances to lower levels of welfare for the society as a whole. Yet, from the point of view of the world economy, given the relative size of developing countries in the world economy, and the importance within their own economies of agriculture and industry (Table 11.2), the costs were borne basically by the countries following the ISI strategy – a key distinction from the current ISA in industrialized countries, which, as argued, has a larger weight in the world economy and important global effects. Now developed countries invoke similar multifunctional arguments for agriculture. And it is also clear that subsidization and protection of agriculture in industrialized countries is imposing costs on their own economies, as inefficient industrialization did in the case of developing countries. But industrialized countries are richer than developing countries and the share of the agricultural sector in their economies is small (Table 11.2): therefore, the costs for industrialized countries of expanding the multifunctionality of agriculture beyond its non-subsidized levels could, in principle, be absorbed by their societies without the relatively larger economic problems that extreme ISI appears to have caused to several developing countries. However, while the industrial sector in developing countries fostered by ISI was relatively small compared to the industrial world economy, the agricultural sector in industrialized countries fostered by ISA represents a larger fraction of world agriculture (Table 11.2) affecting agriculture globally. If all agricultural sectors generate externalities, policies subsidizing agriculture in some countries – and encouraging their expansion beyond what would have been the case without that support – implies that the agricultural sector in other countries may be forced to contract (or grow less than would have otherwise been the case). In this way, some countries would be denied the multifunctional benefits of their agricultural sectors. The old argument regarding the bias against agriculture due to the emphasis in industrialization can then be translated into the most recent discussion regarding subsidized agricultural production in developed countries displacing similar production in developing countries. A point already noted is that the current distinction in the WTO texts (and before in GATT) is not between agriculture and industry in the disciplines applied, but between agricultural primary goods and agroindustrial food products, on the one hand, and the rest of the industry on the other. For instance, beef and sugar are agroindustrial goods, which generate a good portion of the employment in the processing segment. This

Multifunctionality in Industry and Agriculture

255

implies that it is not only the farmer or the rural economy that may be affected differently by this separate treatment, but also workers and entrepreneurs in the agroindustrial sector, some of them located in the urban sector. Therefore the possibility of expanding employment in developing countries both in the rural and urban economies may be hampered by this differential treatment of agricultural and agroindustrial goods, when compared to industrial goods (Diaz-Bonilla and Reca, 2000). But whatever the economic arguments, the limitations under WTO of the instruments to promote the multifunctionality of industry in developing countries, while maintaining the instruments to promote the multifunctionality of agriculture in industrialized countries, would be another case of glaring legal imbalances between the rights and obligations of countries under the WTO. Certainly, as it was argued, the adequate approach considering the welfare implications is to move agriculture further into full integration with WTO rules, rather than developing countries trying to have additional legal access to distorting policy instruments for industrial development; even without factoring the possible negative welfare effects for developing countries themselves of such policies, it seems clear that they cannot possibly win a battle of subsidies against industrialized countries. Overall, the arguments in support of a legal treatment under the WTO that ensures rural and agricultural development in developing countries do not need new and debated notions such as multifunctionality, and can be more effectively based on traditional arguments linked to growth dynamics, poverty alleviation, food security and environmental issues, as they apply to those less favoured countries. Further, the notion of multifunctionality may be not only unnecessary for developing countries to support the policies needed for rural development, but may also be harmful. As argued, this would be the case if it leads mostly to expansion of the production of industrialized countries more than would have been the case without the additional support predicated upon such a notion. In that case, agricultural production in developing countries (and the multifunctional effects linked to it) would contract because of the excess of subsidized production in industrialized countries.

Notes 1

They began in March 2000, as mandated by Article 20 of the Agreement on Agriculture (AoA), and were reaffirmed and made part of a single negotiating undertaking in Doha in November 2001. 2 Of course, the proponents of multifunctionality do not restrict its application to developed countries, although several dimensions of the concept apply mostly to more advanced countries (see below). Vice versa, the notion of government support to industrialization is not confined to developing countries, as illustrated for instance by the debate on industrial policies in the USA especially during the mid-1980s (see, among others, Tyson, 1992).

256

E. Diaz-Bonilla and J. Tin 3

Alexander Hamilton, in his ‘Report on Manufactures’, submitted to the US Congress in 1791, argued that ‘not only the wealth but the independence and security of the country appear to be materially connected with the prosperity of manufactures. Every nation … ought to endeavor to possess within itself, all the essentials of national supply’, that he considered necessary ‘to the safety as well as to the welfare of the society’ (Hamilton, 1791). 4 The Latin American industry had emerged in good measure as a result of growing demand in the region and the natural protection offered by the breakdown in trade and finances during the Great Depression and the two World Wars. 5 Although the Left did not necessarily share this benign view of modernization, some Marxist arguments emphasized the need to move beyond feudalism (which was associated with the agricultural sector) to capitalism (identified with industrialization and the development of the urban proletariat) (Mitrany, 1951). 6 Early development literature appears to assign zero marginal value to labour in the agricultural sector, or at least a value far smaller than in alternative uses (Lewis, 1954). Now multifunctionality appears to assign higher value to employment in agriculture than in alternative fields. 7 There is a similar debate now regarding whether employment belongs to the notion of the multifunctionality of agriculture (see OECD, 2001). 8 The argument was that in any pre-industrial economy, pioneering firms are subject to considerable start-up costs. Without an industrial base (of skilled labour, supplier networks, experienced capital markets and so on), each new industry is at much higher survival risk than if it had started in an already industrialized country. Survival risks also arise from the fact that demand for industrial goods would initially be weak: without an industrial system producing many goods by many wage earners, the first industry is in a highly contingent position, whereas the successive ones are progressively less so. Each new industry would contribute technological spillovers and the development of a skilled labour force to the society as a whole: learning-by-doing would contribute to the hiring firm’s bottom line (for which the firm is compensated), but also to the viability of the industrial sector overall (for which the firm is not compensated). 9 Industrialization could advance in a more or less balanced fashion (although for some, like Rosenstein-Rodan (1943) this would focus basically on light industry, while for others such as Mahalanobis (1955) heavy industry should be incorporated), or instead could be propelled by the fundamental tensions of disequilibria to generate investments, as argued by Hirschman (1958). 10 An important reason for that economic instability was that the ISI strategy created a stop–go dynamic in economic activity: the acceleration of the economy usually led to fewer exports (because a larger percentage of the goods were consumed internally due to growing incomes) and more imported inputs and capital goods (demanded by the expanding industry), generating balance of payment crises when official external reserves reached very low levels. 11 The need to ‘re-establish order’ was the reason generally utilized to try to justify the breakdown of the democratic process (see, for instance, Diaz-Bonilla and Schamis, 2001, on Argentina). 12 Also, the debt crises of the 1980s in many developing countries accelerated the change from the ISI towards an outward-oriented strategy. 13 Of course, some of the trade restrictions in industrialized countries were also directed against other developed countries such as Japan. 14 For the discussion of US strategic interests regarding Europe, and Japan during the Cold War, see, among others, Brown (1983).

Multifunctionality in Industry and Agriculture

257

15

For instance, if the concern is about rural livelihoods, it is not clear, nostalgia aside, that the most beneficial intervention on behalf of rural populations would be to target the agricultural sector per se instead of the geographic area in general with (say) tourism infrastructure and telecommunications. For environmental protection it would be better to subsidize directly the desired environmental practice. In other cases different approaches can be tried, such as buying the land and selling it to trusts that can then administer the land according to best practice (which also goes towards scenic vistas) (OECD, 2001). 16 Equity concerns must also include wider calculations considering consumers and taxpayers in general. Market protection to support agriculture increases the price of food and makes the poor pay a larger share of income in those products (based on the well-known regularity called Engel’s Law). And the equity impact of budget transfers will also depend on the structure of taxes (OECD, 1999). 17 The reference is to production because multifunctionality has been linked to it. Those production effects, however, are different from overall welfare effects, depending on other issues such as changes in terms of trade. 18 Related to the issue of whose multifunctionality it is important to note that the meaning of that concept may be very different for industrialized countries and for the variety of developing countries. For example, the issue of rural employment and vitality of rural communities in industrialized countries, where subsidies are predicated in part on the need to support a choice of lifestyle, is completely different from the situation in developing countries where most of the population is in agriculture, not because that is where they want to be, but because the development process has not offered them other alternatives (WTO, 2001). Also, the notion of food security appears to have different meanings in different countries (Díaz-Bonilla et al., 2000). Environmental problems also differ across countries, appearing mostly as pollution of land and water, due to excess use of agrochemicals in industrialized countries (in part, a consequence of generous production subsidies) and degradation and overuse of natural resources in developing countries (resulting mostly from poverty and lack of financial support to improve technology). Finally the issue of preserving rural landscapes in industrialized countries, as a way of allowing urban dwellers scenic vistas and the possibility of rural relaxation, does not seem to have an obvious equivalent in impoverished developing countries. 19 It should be noted that although the WTO legal system restricts the range of instruments that can be utilized to develop the industrial sector, the policies restricted are usually the most controversial in terms of net welfare effects for a society, while those allowed still include a variety of measures that can be applied within a vision of industrial development based on science and technology (Amsden, 2000) which, utilizing the terminology of the Agreement on Agriculture, are in most cases measures equivalent to those of the Green Box for agriculture.

References Amsden, A. (2000) Industrialization Under New WTO Law. Paper presented at UNCTAD X, 26–27 January, Bangkok, Thailand. Anderson, K. (2000) Agriculture’s ‘multifunctionality’ and the WTO. The Australian Journal of Agricultural Economics 44(3), 475–494.

Australian Bureau of Agricultural and Resource Economics (ABARE) (1999) ‘Multifunctionality’ A Pretext for Protection? Current Issues, 99(3), August. ABARE, Canberra. Balassa, B. and Associates (1971) The Structure of Protection in Developing

258 Countries. The Johns Hopkins University Press, Baltimore, Maryland. Balassa, B. (1984) Adjustment policies in developing countries: a reassessment. World Development, 12 September, pp 955–972. Balassa, B. (1986a) Policy responses to external shocks in developing countries. American Economic Review 76, May, pp 75–78. Balassa, B. (1986b) Outward Orientation. World Bank Development Research Department, Discussion Paper No. 148, World Bank, Washington DC. Baldwin, R. (1969) The case against infant industry protection. Journal of Political Economy 77, 295–305. Bauer, P.T. (1972) Dissent on Development. Harvard University Press, Cambridge, Massachusetts. Bhagwati, J. (1971) The generalized theory of distortions and welfare. In: Bhagwati et al. (eds) Trade Balance of Payments and Growth: Papers in International Economics in Honor of Charles Kindleberger. NorthHolland, Amsterdam. Bhagwati, J. (1993) India in Transition: Freeing the Economy. Oxford University Press, London. Brown, S. (1983) The Faces of Power. Constancy and Change in United States Foreign Policy from Truman to Reagan. Columbia University Press, New York and London. CEPAL (1969) América Latina. El Pensamiento de la CEPAL. Editorial Sudamericana, Santiago, Chile. Chenery, H., Ahluwalia, M., Bell, C., Dulloy, J. and Jolly, R. (1974) Redistribution with Growth. World Bank, Oxford University Press, New York. Commission of the European Community (1999) Protecting the Communities’ Financial Interests and the Fight Against Fraud – Annual Report 1998. European Commission, Brussels. Diao, X., Diaz-Bonilla, E. and Robinson, S. (2003) How Much Does it Hurt? Measuring the Impact of Agricultural Trade Policies on Developing Nations. IFPRI, http://www.ifpri. org/media/trade20030826.htm, Accessed August 2003.

E. Diaz-Bonilla and J. Tin Diaz-Bonilla, E. and Reca, L. (2000) Trade and agro-industrialization in developing countries: trends and policy impacts. Agricultural Economics 23(3), 219–229. Diaz-Bonilla, E. and Schamis, H. (2001) From redistribution to stabilty: the evolution of exchange rate policies in Argentina, 1950–98. In: Frieden, J. and Stein, E. (eds) The Currency Game: Exchange Rate Politics in Latin America. Inter American Development Bank and The Johns Hopkins University Press, Washington DC. Diaz-Bonilla, E., Thomas, M., Cattaneo, A. and Robinson, S. (2000) Food Security and Trade Negotiations in the World Trade Organization: a Cluster Analysis of Country Groups. Trade and Macroeconomics discussion paper No. 59. International Food Policy Research Institute, Washington DC. European Environment Agency (2001) Europe’s Environment – The Third Assessment. EEA, Copenhagen, Denmark. (http://reports.eea.en.int/environmental– assessment_report_2003_10/eu) European Union (1999) Preparation of the Third WTO Ministerial Conference: Council Conclusions, 25 October 1999. http://europa.eu.int/comm/dg06/external/ wto/officdoc/index_en.htm Accessed 8 June 2000. European Union (2000) Fact-Sheets: Reinforcing Control of Agricultural Expenditure. http://europa.eu.int/comm/dg06/publi/ fact/fraud/index_en.htm Frank, A.G. (1969) Capitalism and Underdevelopment in Latin America. Monthly Review Press, New York. Hamilton, A. (1791) Report on the Subject of Manufactures. In: Syrett, H. et al. (eds) The Papers of Alexander Hamilton, 26 vols. Columbia University Press, New York and London. Hayami, Y. (1986) The roots of agricultural protectionism. In: Anderson, K. and Hayami, Y. (eds) The Political Economy of Agricultural Protection. Allen and Unwin, Australia Pty Ltd, Sydney. Hayami, Y. (1990) The why, how, and consequences of agricultural policies: Japan. In: Sanderson, F. (ed.) Agricultural

Multifunctionality in Industry and Agriculture Protectionism in the Industrialized World. Resources for the Future, Washington DC. Hertel, T.W., Anderson, K., Francois, J.F. and Martin, W. (2000) Agriculture and NonAgricultural Liberalization in the Millennium Round. Discussion Paper No. 0016. Centre for International Economic Studies Policy, Adelaide, Australia. Hirschman, A. (1958) The Strategy of Economic Development. Yale University Press, New Haven, Connecticut. Hirschman, A. (1982) The rise and decline of development economics. In: Gersowitz et al. (eds) The Theory and Experience of Economic Development. Yale University Press, New Haven. Johnston, B. and Mellor, J. (1961) the role of agriculture in economic development. American Economic Review 51, 566–593. Josling, T., Tangermann, S. and Warley, T.K. (1996) Agriculture in the GATT. McMillan, London. Kerr, C., Dunlop, J.T., Harbison, F. and Myers, C. (1964) Industrialism and Industrial Man. Oxford University Press, New York. Krueger, A. (1974) the political economy of the rent seeking society. American Economic Review 66, 1–19. Krueger, A. (1978) Liberalization Attempts and Consequences. Ballinger, Cambridge. Krueger, A.O. (1981) Interactions between inflation and trade regime objectives in stabilization programs. In: Cline, W.R. and Weintraub, S. (eds) Economic Stabilization Policy in Developing Countries. Brookings Institute, Washington DC. Krueger, A. (1984) Trade policies in developing countries. In: Handbook of International Economics. North Holland, Amsterdam. Krugman, P. (1994) The fall and rise of development economics. In: Rodwin, L. and Schön, D.A. (eds) Rethinking the Development Experience. Brookings Institution Press, Washington DC. Kuznets, S. (1966) Modern Economic Growth. Yale University Press, New Haven, Connecticut. Lewis, W.A. (1954) Economic Development with Unlimited Supplies of Labour. Manchester School of Economic and Social Studies, Vol. 22(2), pp 139–191.

259 Little, I., Scitovsky, T. and Scott, M. (1970) Industry and Trade in Some Developing Countries. Oxford University Press, Paris. Mahalanobis, P.C. (1955) The approach of operational research to planning in India Sankhya, The Indian Journal of Statistics 16, pp 3–62. Marshall, A. (1898) Principles of Economics, 4th edn. Macmillan, London. Ministry of Agriculture, Forestry and Fisheries, Japan (1999) Annual Report on Food, Agriculture and Rural Areas in Japan. FY 1999 Summary Provisional Translation. http://www.maff.go.jp/hakusyo/kaigai/ehak usyo99.htm Accessed 8 June 2000. Mitrany, D. (1951) Marx Against the Peasant: A Study in Social Dogmatism. George Weidenfeld and Nicolson Ltd, London. Murphy, K., Shleifer, A. and Vishny, R. (1989) Industrialization and the Big Push. Journal of Political Economy 97(5), 1003–1026. OLAF (2001) – the European Anti-Fraud Office. http://europa.eu.int/comm/dgs/olaf/index. htm Orden, D. (2002) Reform’s stunted crop: Congress re-embraces agriculture subsidies. Regulation 25(1), 26–32. Orden, D., Paarlberg, R. and Roe, R. (1999) Policy Reform in American Agriculture: Analysis and Prognosis. The University of Chicago Press, Chicago, Illinois. Organisation for Economic Co-operation and Development (OECD), Directorate for Food, Agriculture and Fisheries (1999) Distributional Effects of Agricultural Support in Selected OECD Countries. Agr/ca(99)8/Final. November 1999. OECD Publications, Paris. Organisation for Economic Co-operation and Development (OECD) (2001) Multifunctionality Towards an Analytical Framework. OECD Publications, Paris. Organisation for Economic Co-operation and Development (OECD), Directorate for Food, Agriculture and Fisheries. (several issues) Agricultural Policies in OECD Countries: Monitoring and Evaluation. OECD Publications, Paris.

260 Prebisch, R. (1950) The Economic Development of Latin America and its Principal Problems. United Nations, New York. Rapp, D. (1988) How the US Got into Agriculture. And Why it Can’t Get Out. Congressional Quarterly, Washington DC. Rosenstein-Rodan, P. (1943) Problems of industrialization of Eastern and SouthEastern Europe. Economic Journal, June–September 1943, pp 202–211. Samuelson, P. (1954) The pure theory of public expenditure. Review of Economic and Statistics 36(4), 387–389. Schultz, T.W. (1964) Transforming Traditional Agriculture. Yale University Press, New Haven, Connecticut. Sharma, R., Konandreas, P. and Greenfield, J. (1996) An overview of the assessments of the impact of the Uruguay Round on agricultural prices and incomes. Food Policy 21(4/5), 351–363. Singer, H.W. (1950) The distribution of gains between investing and borrowing countries. American Economic Review 40 (May), 473–485. Tyson, L.D. (1992) Who’s Bashing Whom? Trade Conflict in High-Technology Industries. Institute for International Economics, Washington DC. USDA/Economic Research Service (1999) The

E. Diaz-Bonilla and J. Tin Use and Abuse of Multifunctionality. USDA, Washington DC. Valdés, A. and Zietz, J.A. (1980) Agricultural Protection in OECD countries: its costs to less-developed countries. IFPRI Research Report Series, No. 21, International Food Policy Research Institute, Washington DC. World Bank (1986) World Development Report. Oxford University Press, New York. World Bank (2001) World Development Indicators. World Bank, Washington DC. WTO (World Trade Organization) (2000) Note on Non-trade Concerns. Submission to the November 2000 special session of the WTO committee on agriculture by Barbados, Burundi, Cyprus, Czech Republic, Dominica, Estonia, the EC, Fiji, Iceland, Israel, Japan, Korea, Latvia, Liechtenstein, Madagascar, Malta, Mauritania, Mauritius, Mongolia, Norway, Poland, Romania, Saint Lucia, Slovak Republic, Slovenia, Switzerland and Trinidad and Tobago. G/AG/NG/W/36Rev.1. Committee on Agriculture. Special Session. 9 November 2000. WTO (World Trade Organization) (2001) Proposals by India in the areas of: (i) food security, (ii) market access, (iii) domestic support, and (iv) export subsidies. Negotiations on WTO Agreement on Agriculture. G/AG/NG/W/102.

12

Trade in Genetically Modified Food: Promises and Pitfalls for the Poor

CHANTAL POHL NIELSEN* AND KAREN E. THIERFELDER† *Danish

Research Institute of Food Economics, Agricultural Policy Research Division, Rolighedsvej 25, 1958 Frederiksberg C, Denmark; †Economics Department, US Naval Academy, 589 McNair Road, Annapolis, MD 21402, USA

Introduction Increased food availability is one of the most obvious preconditions for improving food security in developing countries. This objective is primarily achieved by increasing domestic food production, although imports play an important role as well by reducing the variability of food availability over time. One of the most recent advances in agricultural technology – the development of genetically modified (GM) crops – holds the promise of increasing farm productivity and thereby enabling an increase in domestic food supplies at lower cost. Yet, as history has shown, new technologies sometimes face significant popular opposition that stunts their continued development. This is also true for the use of genetically modified organisms (GMOs) in food production. Skepticism towards GM food has been particularly vociferous in Western Europe,1 although consumers elsewhere also demand information about GM foods. This opposition to GM food may affect the extent to which developing countries reap the potential net benefits from this new technology. In a highly integrated world trading system (and in spite of the persistence of trade barriers of various sorts) any single country cannot afford to ignore the consumer reactions and policy choices made in countries with which they trade. This chapter seeks to shed light on the complexities surrounding the international GMO debate and how they affect developing countries. First, we provide a brief overview of the current status of GM crops in agriculture and explore the potential for growing GM crops in developing countries. Then we describe how the emergence of GMOs has prompted the development of national regulations and international agreements that could raise difficulties in international trade and potentially lead to © CAB International 2006. WTO Negotiations and Agricultural Trade Liberalization: the Effects of Developed Countries’ Policies on Developing Countries (eds E. Diaz-Bonilla et al.)

261

262

C.P. Nielsen and K. Thierfelder

disputes in the World Trade Organization (WTO). The final part of the chapter surveys a number of empirical studies of the potential economic and trade impacts of the new technology under different scenarios, relating to how the world trading system and national markets handle GM commodities.

Background and Current Status2 The most recent research and development (R&D) advances in modern biotechnology have introduced an ever-widening range of genetically modified products to agriculture. While traditional biotechnology improves the quality and yields of plants and animals through, for example, selective breeding, genetic engineering3 is a new biotechnology that enables direct manipulation of genetic material (inserting, removing or altering genes). In this way the new technology speeds up the development process, shaving years off R&D programmes. Proponents argue that genetic engineering entails a more controlled transfer of genes because the transfer is limited to a single or just a few selected genes, whereas traditional breeding risks transferring unwanted genes together with the desired ones. Despite that advantage, those opposed to GM products argue that the side effects in terms of potentially adverse impacts on the environment and human health are unknown. The ethicists among them also worry that there are seemingly no limits to the possibilities of gene transfers, even between plants and animals, which leads to discussions about whether limits to the use of this technology should be drawn based on ethics. According to FAO (2004), the evidence to date regarding GM products currently on the market shows that they are safe to eat, although continuous monitoring is recommended. Moreover, the report concludes that more complex products of the future may require additional food safety procedures in addition to those currently in use. In terms of the potential environmental impacts of GM crops, there is greater disagreement among scientists. There is general agreement on the types of hazards that exist, but disagreement on their likelihood and severity (FAO, 2004). Genetic engineering techniques and their applications have developed rapidly since the introduction of the first genetically modified plants in the 1980s. In 2002 genetically modified crops occupied almost 60 million ha of land – representing a considerable expansion from less than 2 million ha in 1996. This GM production takes place in 16 countries, but is concentrated in just a few principal producers. In 2002 four countries cultivated 99 per cent of the global transgenic crop area. The USA held 66 per cent of the total crop area, followed by Argentina (23 per cent), Canada (6 per cent) and China (4 per cent) (James, 2003). In general, these countries have devoted a large share of total acreage to biotech products. In 2000 the GM shares were: USA 41 per cent, Argentina 84 per cent, Canada 17.6 per cent and Australia 41 per cent (Marra et al., 2002). Cultivation of transgenic crops has so far been most widespread in the production of soybean and maize, accounting for 62

Trade in Genetically Modified Food

263

per cent and 21 per cent of total transgenic crop production in 2002, respectively. Cotton and canola made up 12 per cent and 5 per cent, respectively, of the total. Rapid adoption rates have meant that for the first time, more than half (51 per cent) of the 72 million ha of soybean grown worldwide were GM in 2002. Of the 25 million ha of canola, GM varieties accounted for 12 per cent, while the corresponding GM share for global maize production (140 million ha) was 9 per cent in 2002 (James, 2003). Development of plants with enhanced agronomic traits aims at increasing farmer profitability, typically by reducing input requirements and hence costs, i.e. an increase in factor productivity. To date genetic engineering in agriculture has mainly been used to modify crops so that they have improved agronomic traits such as tolerance of specific chemical herbicides and resistance to pests. In 2002 herbicide tolerance deployed in soybean, maize and cotton occupied 75 per cent of the global GM area. Pest resistance – more specifically, Bt (Bacillus thuringiensis) insect resistance – occupied 17 per cent of total GM area (James, 2003). Genetic modification can also be used to improve the final quality characteristics of a product for the benefit of the consumer, food processing industry or livestock producer. Such traits may include enhanced nutritional content, improved durability and better processing characteristics. Most of these types of modification are still in the research pipeline. Continued expansion in the use of transgenic crops will depend in part on the benefits obtained by farmers cultivating transgenic instead of conventional crops relative to the higher cost for transgenic seeds.4 So far the improvements have been not so much in increased yields per hectare of the crops but rather by reducing costs of production (OECD, 1999). Empirical data on the economic benefits of transgenic crops are still very limited, however. The effects vary from year to year and depend on a range of factors such as crop type, location, magnitude of pest attacks, disease occurrence and weed intensity. In developing countries, one of the main reasons for low crop yields is the prevalence of biotic stresses caused by weeds, pests and diseases. The first generation of improved transgenic crops, into which a single trait such as herbicide tolerance or pesticide resistance has been introduced, can provide protection against several of these biotic stresses. Examples of traits and crops currently being field-tested in developing countries include virus-resistant fruits and vegetables (e.g. melon, papaya, potato, squash, tomato and sweet pepper); insect-resistant grains and legumes (e.g. rice, soybean); diseaseresistant vegetables (e.g. potato); and delayed-ripening vegetables (e.g. chilli pepper) (James and Krattiger, 1999). The development of more complex traits such as drought resistance, which is a trait controlled by several genes, is under way and highly relevant for tropical crops that are often grown under harsh natural and weather conditions and on poor-quality soils. Furthermore, the development of foods with enhanced nutritional value – such as the recent development of ‘Golden Rice’ that is enhanced with Vitamin A and iron – may be a low-cost way of helping deal with widespread malnutrition problems (see e.g. Nielsen and Anderson, 2001b).

264

C.P. Nielsen and K. Thierfelder

In summary, rapid adoption rates in some countries indicate that some farmers find these new crops beneficial. Also, there seems to be potential for further development of this technology. Yet there are others – particularly consumer and environmental groups – that are truly concerned about potentially adverse long-term consequences of widespread cultivation of GM crops and the use of GMOs in food production. These concerns have triggered a range of national and international regulatory actions, to which we now turn.

National and International GMO Regulations5 National GMO regulations Given the novelty of commercial application of genetic engineering techniques, the regulatory frameworks around the world are in the process of being formulated and re-formulated in response to consumer reactions to these new products. Existing regulations differ substantially both in scope and stage of implementation, varying from very restrictive regulations in certain industrial countries to non-existent in certain developing countries. EU regulations related to GMOs have adhered to the precautionary principle, i.e. that considerations of human health and the environment should rank higher than possible economic benefits in circumstances where there is uncertainty about the outcome. In response to the strong resistance by consumer and environmental groups in the EU to both genetically modified foods and the use of GMOs in agricultural production, the EU imposed a de facto moratorium on the authorization of new releases of GMOs in 1998. This action reflected an extreme interpretation of the precautionary principle. The freeze was intended to give the EU time to study the issue more carefully and to develop a system of traceability and labelling. In an effort to remove this moratorium, the EU later established guidelines for the development of strategies and best practices to ensure the co-existence of GM crops with conventional and organic farming. The guidelines set out general principles, pointed out technical and procedural aspects to be taken account of, and provided a list of possible actions that could be tailored for implementation at national, regional and local levels. Such actions could include on-farm measures (isolation distances, buffer zones, pollen barriers such as hedgerows), cooperation between neighbouring farms (e.g. information about sowing plans, coordination of crop choice and planting times), monitoring and notification schemes, training of farmers, exchange of information, and advisory services (European Union in the US, 2003). The aim of the co-existence guidelines is to ensure that farmers are able to choose freely which types of crops to cultivate, be they GM, conventional or organic. Furthermore, the guidelines were not an EU-wide ‘one-size-fits-all’ approach, but rather one that allowed adaptation to specific national, regional and local conditions (European Commission, 2003).

Trade in Genetically Modified Food

265

These regulatory changes represented a step in the direction of once again giving the green light to the commercial growing of GM crops in the EU. Part of the motivation for moving in this direction seemed to have been the fact that the USA, backed by Canada and Argentina, filed a suit with the WTO against the 1998 moratorium, arguing that the ban was scientifically unfounded and violated global trading rules. However, since the regulatory advances mentioned, and the enactment of separate changes in labelling legislation for GMOs, which was also interpreted as another step towards lifting the moratorium, approvals for new GMOs products in the European market have been minimal so far,6 and the WTO case is still pending. Underlying this transatlantic dispute are fundamentally different views of GM products. In the EU, GM and conventional foods are viewed as distinctly different products. US regulators, on the other hand, view genetic engineering as ‘just another technology’, and therefore they view GM foods as ‘substantially equivalent’ to the conventional variety. These different perceptions of GM foods affect labelling requirements. At one extreme, the USA and Canada require labelling only when the nutritional or allergenic composition has been altered through genetic engineering. Otherwise, labelling food containing GM ingredients is voluntary. These countries are nevertheless providing voluntary guidelines for those who wish to apply labels specifying nonGM. At the other extreme, most other OECD countries, the EU, Japan, Australia/New Zealand, Korea, Switzerland, the Czech Republic, Hungary, Norway and Iceland all have some type of mandatory labelling regulations for GM foods. The EU has mandatory labelling for all food and food ingredients containing genetically engineered DNA/proteins above a 0.9 per cent tolerance level through its recently amended and approved rules for traceability and labelling of GMOs. These rules also stipulate a 0.5 per cent threshold for the adventitious presence of GMOs that are unauthorized but have nevertheless been assessed as risk-free (Bridges, 2003). The USA is critical of the new EU labelling and traceability rules, however, and argues that little has been done to remove barriers to US products, which are generally not labelled.

International GMO regulations Given the different national approaches to regulation of genetically modified products, future trade disputes are a distinct possibility, as discussed above. In particular, it is possible that the so-called Cartagena Protocol on Biosafety, which came into force in September 2003, may conflict with existing WTO agreements.7 The Biosafety Protocol has the objective of ensuring safe transboundary movement of living modified organisms resulting from modern biotechnology. In particular, the Protocol allows governments to decide whether or not to accept GMO imports and under what conditions, and reconfirms their rights to set their own domestic regulations. Most importantly, the Protocol stipulates that lack of

266

C.P. Nielsen and K. Thierfelder

scientific evidence regarding potential adverse effects of GMOs on biodiversity, also taking into account risks to human health, shall not prevent a signatory from taking action to restrict the import of such organisms in order to avoid or minimize risks (UNEP, 2000). In essence, this is an explicit acceptance of the guiding influence of the precautionary principle. When used in an international trade context, there is a risk that the precautionary principle can develop into protectionism against any new technology in, e.g. agriculture. It will be extremely difficult to assess whether a measure is there for precautionary reasons or simply as a form of hidden protectionism. The Biosafety Protocol has the general objective of protecting and ensuring sustainable use of biological diversity whilst also taking into account risks to human health. Agreements within the WTO also acknowledge the rights of a country to protect its environment, ensure food safety and to inform consumers. But WTO members are also obliged to adhere to agreements that restrict the way trade-related measures are used to achieve these goals. More specifically related to food safety and animal and plant health are the Agreement on Sanitary and Phytosanitary Measures (SPS) and the Agreement on Technical Barriers to Trade (TBT). These agreements allow member states to impose certain restrictions on trade if the purpose of the measure is to protect human, animal or plant life and health. The TBT agreement also covers technical measures aimed at protecting the environment and other objectives. At the same time the agreements aim to ensure that applied measures and technical regulations do not pose unnecessary or arbitrary barriers to trade and that they are no more trade-restrictive than necessary to fulfill the legitimate objective (WTO, 1998a,b). Moreover, use of such measures must be consistent with the key principles of the WTO, i.e. nondiscrimination among member states, ‘national treatment’ of imports once having entered the domestic market and transparency of customs procedures. Hence, some of the current WTO agreements may prove to be in conflict with the rights to restrict trade in living modified organisms provided for in the Biosafety Protocol. Both the SPS and the TBT agreements of the WTO encourage the use of international standards, guidelines and recommendations where they exist, such as in the realms of the Codex Alimentarius. Codex adopted in June/July 2003 three risk analysis standards for GM foods. The standards include references to product tracing and labelling as risk management tools. Many observers viewed this as a breakthrough in international negotiations on the use of traceability systems and at least partially vindicated the EU’s insistence on introducing a labelling and traceability system for GM products. The USA, on the other hand, had a different perception of the Codex standards. In particular, it has been argued that ‘traceability’ and ‘product tracing’ are two different concepts: ‘traceability’ would refer to the whole production chain, while ‘product tracing’ would be limited to ‘one step forward and one step back’ (Bridges, 2003). Since then Codex members have been unable to agree on rules for biotechnology

Trade in Genetically Modified Food

267

labelling, both in 2004 and 2005 meetings. So, although in principle the establishment of international standards ought to be helpful in reducing the likelihood of future trade disputes, this first example of Codex standards in the area of GMOs shows how complex and politically sensitive the area is. In terms of potential trade disputes relating to GMOs, there is another important aspect of the Biosafety Protocol, which seems open to interpretation and hence potential dispute. This is the relationship between the Protocol and the WTO agreements. The text (UNEP, 2000) states that the ‘Protocol shall not be interpreted as implying a change in the rights and obligations of a Party under any existing international agreements’ while, at the same time, the latter statement is ‘not intended to subordinate [the] Protocol to other international agreements’. Given the existing dispute settlement mechanism of the WTO, it is likely, however, that potential disputes will be taken up in this forum. Summarizing, the emergence of GMOs in agricultural and food production has introduced a variety of new issues within the WTO, only some of which are amenable to confront under current rules.8

Empirical Economic and Trade Impact Assessments9 As the previous discussion has clearly illustrated, trade in genetically modified foods is a controversial topic. This section surveys five recent analyses of the potential economic and trade impact of the new biotechnologies in agriculture, under different scenarios concerning changes in consumer preferences and how the national and international regulatory systems and markets handle GM commodities10. Given that the GM technologies are both very new and evolving rapidly, there are major knowledge gaps about their potential impact on production and distribution costs. Moreover, the world trading system is just starting to adapt. Hence, the economic studies reviewed here must all be viewed as tentative, based on uncertain premises and incomplete data. Much of the work has had to rely on sensitivity analysis – identifying important parameters – and consideration of potential scenarios, rather than on econometric analysis of historical data. The studies have nevertheless produced a few robust conclusions, as well as identifying crucial assumptions and important information gaps. An appropriate tool to analyse complex trade issues is a global computable general equilibrium (CGE) model. Given the key importance of the agricultural sector in the developing countries’ economies, both in terms of income and employment generation, it is useful to analyse agricultural technology changes in a general equilibrium framework. Such models describe both the vertical and horizontal linkages between all product markets, both within the individual countries as well as between countries via their bilateral trade flows. Furthermore, CGE models take full account of resource constraints (land, labour and capital) and enable macro-

268

C.P. Nielsen and K. Thierfelder

level analysis of the welfare implications of a given technology or policy change. Indeed, all the studies surveyed here use global CGE models, although they differ by country and sectoral focus through their choices concerning data aggregation levels. They also incorporate different assumptions about: (i) how GM technology affects productivity; (ii) the transfer of GM technology across countries; (iii) market segmentation; and (iv) the nature of the shift in consumer preferences in developed countries. Although they use different theoretical models, all five analyses use world production, trade and protection data from the Global Trade Analysis Project (GTAP) based at Purdue University, Indiana (Hertel, 1997). As discussed under ‘background’, China is one of the four principal GM producers in the world. China has been investing heavily in biotech R&D since the 1980s. Anderson and Yao (2001) evaluated the effect of China’s development of GM technology using the GTAP model and simple assumptions about the productivity impact of GM crops. The authors simulated GM technology improving productivity in four crops: rice, cotton, maize and soybean. They also took into account that some countries – North America, South America (Argentina, Chile and Uruguay) and South-east Asia – already used GM technology in these crops. The authors concluded that China would gain from GM technology if there were no environmental externalities and no adverse consumer reactions. They found that the welfare gains from GM adoption would depend on retaining market access abroad. Market access is important for direct sale of GM products and indirect links via the sale of products such as textiles and apparel, which use a GM-potential input in production. However, one must also account for the costs of R&D development that are not included in the welfare measure they reported. Furthermore, adverse consumer reactions will reduce the benefits of adopting GM technology in China. This country, therefore, has incentives to ensure that the GM debate does not lead to limitations of market access for GM products. Huang et al. (2002) also evaluated the effects of GM technology on China’s production and trade. They used productivity estimates that were based on empirical micro-level data for the cotton sector and experimental data for the rice sector. They also used the GTAP model, with baseline projections for 2010. This allowed them to incorporate future changes in China that are important for its trade relations: i.e. China’s accession to the WTO between 2002 and 2005; global phase-out of the Multifiber Agreement by January 2005; and EU enlargement. Huang et al. considered four scenarios. First they considered the effects of factor-biased productivity growth in cotton, based on empirical estimates. They found that the price of cotton would decline by 10.9 per cent. This affects trade in textiles because textile production costs would decline. Output and exports of textiles would increase by 0.7 per cent and 0.9 per cent, respectively. When rice also benefits from a productivity gain due to GM technology, the price would decline by 12 per cent in 2010. The welfare effects of GM technology (in both cotton and rice) are substantial: annual income would increase by US$5 bn in 2010, or about US$3.5 per person.

Trade in Genetically Modified Food

269

Next, they considered the implications of a negative consumer response in China’s major rice export markets: Japan, Korea, the enlarged EU and South-east Asia. They found that exports of GM rice from China would drop substantially (although this is from a low base since rice exports are only 1.2 per cent of production). There would be a slight decline in output growth for China. Finally, they considered the effect of labelling GM rice, to be consistent with China’s demand that GM soybean imports from the USA are labelled as such.11 They modelled the effect of labelling as an increase in the cost of services required for rice production, assuming a total increase in production costs due to labelling to be 3 per cent. They found that labelling is costly in terms of welfare. The welfare loss to China would be US$1.3 bn because labelling would raise the domestic price of GM rice, hurting consumers. The next three studies took a global view of the GMO debate. Nielsen and Anderson (2001a) used the GTAP model to evaluate the effect of GMO productivity increases in maize and soybean production in a range of countries. They assumed a GM-driven productivity growth of 5 per cent in coarse grain (excluding wheat and rice) and oilseeds in North and South America, Mexico, India, China, Rest of East Asia (excluding Japan and the East Asian newly industrialized countries, or NICs) and South Africa. Other countries are assumed to refrain from the use of GM crops in their production systems – the EU by choice and sub-Saharan Africa because they are assumed not to be able to gain access to the technology. Nielsen and Anderson (2001a) focused on policy choices and consumer reactions to GM foods. The first scenario (Scenario 1) is a base case where GM technology is introduced as described above, but with no policy or consumer reactions to this. A 5 per cent reduction in overall production costs in these sectors leads to increases in coarse grain production of between 0.4 per cent and 2.1 per cent, and increases in oilseed production of between 1.1 per cent and 4.6 per cent, in the GM-adopting regions. The production responses are generally larger for oilseeds as compared to coarse grain. This is because a larger share of oilseed production as compared with coarse grain production is destined for export markets in all the reported regions, and hence oilseed production is not limited to the same extent by domestic demand. Increased oilseed production leads to lower market prices and hence cheaper costs of production in the vegetable oils and fats sectors, expanding output there. This expansion is particularly marked in South America, which exports a large share of production. In North America, maize is also used as livestock feed, and hence the lower feed prices lead to an expansion of the livestock and meat processing sectors there. Due to the very large world market shares of oilseed from North and South America and coarse grain from North America, the increased supply from these regions causes world prices for coarse grain and oilseeds to decline by 4.0 per cent and 4.5 per cent, respectively. As a consequence of the more intense competition from abroad, production of coarse grain and oilseeds declines in the non-adopting regions. This is particularly so in Western Europe, a major net importer of particularly oilseed, of which about half comes from North America. In the developing countries too, production

270

C.P. Nielsen and K. Thierfelder

of coarse grain and oilseeds is reduced slightly. The changes in India, however, are relatively small compared with e.g. China and South America. This is explained by the domestic market orientation of these sales. Global economic welfare (as traditionally measured in terms of equivalent variations of income, ignoring any externalities) is boosted in this first scenario by US$9.9 bn per year, two-thirds of which are enjoyed by the adopting regions. All regions (both adopting and non-adopting) gain in terms of economic welfare, except sub-Saharan Africa (a non-adopter). Most of this gain stems directly from the technology boost. The net-exporting GM-adopters experience worsened terms of trade due to increased competition on world markets, but this adverse welfare effect is outweighed by the positive effect of the technological boost. Western Europe gains from the productivity increase in the other regions only in part because of cheaper imports; mostly it gains because increased competition from abroad shifts domestic resources out of relatively highly assisted segments of EU agriculture. Next, Nielsen and Anderson (2001a) considered the scenario (Scenario 2) in which Western Europe not only refrains from using GM crops in its own domestic production systems, but the region is also assumed to reject imports of genetically modified oilseed and coarse grain from GM-adopting regions. Note that – as in the two previous analyses – the distinction between GM-inclusive and GM-free products is simplified to one that relates directly to the country of origin, and labelling costs are ignored.12 This import-ban scenario reflects the most extreme application of the precautionary principle within the framework of the Biosafety Protocol. A Western European ban on the imports of genetically modified coarse grain and oilseed changes the situation in Scenario 1 dramatically, especially for the oilseed sector in North America, which is highly dependent on the EU market. The result of the European ban is not only a decline in total North American oilseed exports by almost 30 per cent, but also a production decline of 10 per cent. For coarse grain, by contrast, only 18 per cent of North American production is exported and just 8 per cent of those exports are destined for Western Europe. Therefore, the ban does not affect North American production and exports of maize to the same extent as for soybean, although the downward pressure on the international price of maize none the less dampens significantly the production-enhancing effect of the technological boost. Similar effects are evident in the other GMadopting regions, except for India – once again because its production of these crops is virtually all sold domestically and so is not greatly affected by market developments abroad. For sub-Saharan Africa, which by assumption is unable to adopt the new GM technology, access to the Western European markets when other competitors are excluded expands. Oilseed exports from this region, for example, rise by enough to increase domestic production by 4 per cent.13 Aggregate welfare implications of this scenario are substantially different from those of Scenario 1. Two-thirds of the global gain from the new GM technology as measured in Scenario 1 would be eroded by an

Trade in Genetically Modified Food

271

import ban imposed by Western Europe: it falls from US$9.9 bn per year to just US$3.4 bn, with almost the entire erosion in economic welfare borne in Western Europe. This assumes that consumers are indifferent to the choice between GM-free and GM-inclusive foods and that the import ban therefore restricts them from benefiting from lower international prices. To the extent that some Western Europeans in fact value a ban on GM products in their domestic markets, that would partially offset the loss in economic welfare. The rest of the welfare loss is borne by the net-exporting adopters (mainly North and South America). Since the non-adopting regions generally purchase most of their imported coarse grain and oilseed from the North American region, they benefit even more than in Scenario 1 from lower import prices: their welfare is estimated to be greater – by almost one-fifth in the case of a Western European import ban as compared with the case of no European reaction. The key exporters of the GM products, North and South America and China, all show a smaller welfare gain in this scenario compared with the scenario in which there is no European policy response. Net importers of maize and soybean (e.g. ‘other high-income’, which is mostly East Asia), by contrast, are slightly better off in this scenario than in Scenario 1. Meanwhile, the countries in sub-Saharan Africa are affected in a slightly positive instead of slightly negative way, gaining from better terms of trade. In particular, a higher price is obtained for their oilseed exports to Western European markets compared with Scenario 1. As an alternative to the scenario of a policy response by governments, Nielsen and Anderson (2001a) also considered a shift in private European preferences away from imported coarse grain and oilseed and in favour of domestically produced crops. The scenario is implemented as an exogenous 25 per cent reduction in final consumer and intermediate demand for all imported oilseed and coarse grain (that is, not only those which can be identified as coming from GM-adopting regions). Some European consumers and firms are assumed to choose to completely avoid products that are produced outside Western Europe. Western European producers and suppliers are assumed to be able to signal – at no additional cost – that their products are GM-free by e.g. labelling their products by country of origin. This is possible because it is assumed that no producers in Western Europe adopt GM crops (perhaps due to government regulation), and hence such a label would be perceived as a sufficient guarantee of the absence of GMOs. Having consumers express their preferences through market mechanisms rather than through a government-implemented import ban has a much less damaging effect on production in the GM-adopting countries. In particular, instead of declines in oilseed production – as in Scenario 2 – there are slight increases in this scenario, and production responses in coarse grain are slightly larger. As expected, domestic oilseed production in Western Europe must increase somewhat to accommodate the shift in preferences, but not nearly to the same extent as in Scenario 1. Furthermore, there are in fact minor price reductions for agri-food products in Western Europe in part because (by assumption) the shift in preferences is only partial, and so some

272

C.P. Nielsen and K. Thierfelder

consumers and firms do benefit from lower import prices. In other words, in contrast to the previous scenario, a certain link between EU prices and world prices is retained. The output growth in sub-Saharan Africa in Scenario 2, by taking the opportunity of serving European consumers and firms while other suppliers were excluded, is replaced in this scenario by declines: sub-Saharan Africa loses export shares to the GM-adopting regions. Van Meijl and van Tongeren (2001) also used the GTAP model to evaluate the effects of using GM technology in maize and soybean. Unlike the other studies, they had a different type of productivity shock for each crop. For maize, studies indicate that GM technology increases yield. They model this as a Hicks-neutral productivity growth at 5 per cent, following Nielsen and Anderson (2001a). For soybean, there are savings on inputs of chemicals and labour – this is modelled as a 5 per cent chemical cum labour augmenting technical change in that sector. Their analysis also differs from the other studies in two other ways. First, they model the spread of GM technology as an endogenous knowledge spillover linked to trade flows and country characteristics. They begin from the premise that knowledge is embodied in traded goods and that is how it travels between countries. ‘The international diffusion of these technologies is not perfect but dependent on trade linkages, absorption, capacity, size of farms and whether a technology is socially acceptable.’ There is no monetary compensation for the technology transfer. In contrast, Anderson and Yao (2001) note that a country must pay for research and development to acquire GM technology. Van Meijl and van Tongeren also differ from other studies because they include the EU Common Agricultural Policy (CAP). The CAP shields EU farmers from world market price developments. So, if world prices for grain declines, as a result of GM technology in the major exporters, the lower world prices do not necessarily trigger a substitution of EU demand towards imported varieties. From the base scenario, GM adoption by North America (NAM), van Meijl and van Tongeren show the potential knowledge spillovers (potential because social acceptance is not accounted for), shown as the percentage productivity shock (NAM received 5 per cent) they can apply to the GM sectors. Australia/New Zealand potentially receives full spillovers because their farm size and education level exceed selected threshold levels. Argentina and the EU potentially receive 70 per cent or 60 per cent of the productivity gain; they have relatively highly educated farmers (like NAM), but average farm size is lower. Potential spillovers to developing countries are smaller because they trade fewer chemicals with NAM, their farm size is too small and/or education level is too low to adopt the new GM technologies profitably. Next, they consider production and farm income impact of alternative EU policy responses to GMOs. With the CAP, the EU is isolated from the downward pressures on prices brought about by the global productivity boost. When the EU does not accept GMO production technologies, modelled as an import ban, there is no internal productivity gain. Production and farm income do not change. They note that this contrasts with Nielsen and

Trade in Genetically Modified Food

273

Anderson (2001a), who find a sharp reduction in coarse grain output. They argue that Nielsen and Anderson (2001a) do not represent the CAP and so overstate the negative production effect of not adopting GM technology in the EU. However, van Meijl and Van Tongeren do not discuss the effect of GM technology on the programme cost of the CAP. Depending on the nature of the productivity shock and price decline in GM-adopting countries, the CAP expenditures may become too expensive to maintain.14 The final study to be reviewed in this chapter distinguishes itself from the other four by developing a model that enables all countries to produce both GM and non-GM crops. In the GTAP model applications reviewed above, a country produces either the GM or non-GM variety exclusively. Although also using the GTAP database, Nielsen et al. (2001) use an alternative, but very similar, global CGE model to conduct their analysis. More specifically, the model developed by Nielsen et al. (referred to henceforth as NTR) segments the coarse grain and oilseed sectors into GM and non-GM lines of production. The NTR model allows a country to produce both varieties in response to market conditions. This segregation is introduced based on a notion that there may be a viable market for guaranteed GMO-free products alongside the new GMO-inclusive varieties if the GMO-critical consumers are willing to pay a price premium. Furthermore, the NTR model distinguishes itself by retaining the GM/non-GM distinction throughout the entire food processing chain, i.e. with identity preservation from primary crop to intermediate input and through to the final product.15 In the base data used for this model analysis, it is assumed that all regions initially produce some of both the GM and non-GM varieties of oilseed and coarse grain. The assumed GM shares of production, based on estimates provided in James and Krattiger (1999), are just 10 per cent in all but three regions. The exceptions are the Americas and developing Asian countries, where it is assumed 40 per cent of coarse grain and 60 per cent of oilseed (90 per cent in South America) use GM technology.16 The GM-adopting cereal grain and oilseed sectors are assumed to make more productive use of the primary factors of production as compared with the non-GM sectors. NTR assume the GM oilseed and GM coarse grain sectors in all regions have a 10 per cent higher level of primary factor productivity as compared with their non-GM (conventional) counterparts. This shock is slightly different from the shock imposed in the four GTAP applications: it is twice the size, but it is applied only to primary factor inputs and not to intermediate input use. In addition, there is some evidence that cultivating GM varieties substantially reduces the use of chemical pesticides and herbicides (e.g. Pray et al., 2000). Based on this work, the use of chemicals in the GM oilseed and GM cereal grain production is assumed to be reduced by 30 per cent. The NTR study focuses on the impact of consumer responses to GM food. The starting point for the consumer preference experiments is that food products come in two varieties, distinguished by their method of production: GM and non-GM. The model has a representative consumer who views these two varieties as imperfect substitutes. Three different

274

C.P. Nielsen and K. Thierfelder

consumer response scenarios are examined. In the base case, consumers in all countries are relatively indifferent with respect to the introduction of GM techniques in food production, and so find GM and non-GM food varieties to be highly substitutable. The next two scenarios attempt to capture the fact that citizens in Western Europe and high-income Asia dislike the idea of genetically modified foods. In the second scenario this dislike is captured by lowering the elasticities of substitution between the GM and non-GM varieties for consumers in these two regions. As consumers perceive GM and non-GM varieties as bad substitutes, the GM-variety must sell at more of a discount to induce consumers to purchase it. Citizens in all other regions are basically indifferent, and hence the two varieties remain fairly substitutable in consumption. In some countries, irrespective of how cheap these products may become (relative to non-GM foods), some consumers may simply not want to consume GM foods. Hence in the third scenario, consumers are assumed to change the ratio of GM to non-GM foods demanded at initial prices – a structural shift in their expenditure pattern. Consumers spend the same amount on consumption of food, but the composition is changed in favour of non-GM varieties, even with no change in prices. In this scenario, the GM share of foods in consumption in Western Europe and high-income Asia is reduced to 2 per cent. In all scenarios, output of GM varieties increases and output of non-GM varieties declines, due to the productivity shock to the GM varieties. However, the magnitude of the change depends on consumer preferences in Western Europe and Japan, two important export markets for the GMproducing regions. For example, North America is very sensitive to changes in preferences toward GMOs because it is the world’s largest exporter of both oilseed and coarse grain, and it is particularly dependent on the GMcritical markets for these exports. Total exports of the GM varieties decline as GM and non-GM substitutability worsens in the GM-critical regions, especially for oilseed, because almost 80 per cent of North American oilseed exports are initially sold in these markets. The consumer preference shift also affects developing countries. For example, GM oilseed exports from South America (an extensive GM adopter) and sub-Saharan Africa (a region with a low share of GM varieties in total production) initially increase due to the factor productivity shock. Total GM oilseed exports decline as preferences in high-income Asia and Western Europe turn against GMOs. Furthermore, GM exports are redirected from the GM-critical regions and spread evenly over the other importing regions. Of South America’s total oilseed exports, 84 per cent are initially sold on GM-critical markets as compared with 58 per cent of oilseed exports from sub-Saharan Africa. The adjustment in total GM oilseed exports is therefore relatively larger for South America. As expected, the exports of non-GM oilseeds from these two regions are generally being diverted toward the GM-critical regions and away from other regions. Both South America and sub-Saharan Africa depend on imports for almost one-tenth of their total cereal grain absorption. However, in terms of

Trade in Genetically Modified Food

275

sources, South America depends almost entirely on North America for its imports, while imports into sub-Saharan Africa come from North America (50 per cent), Western Europe (16 per cent), and the Rest of World (28 per cent). Because citizens of South America and sub-Saharan Africa are assumed to be uncritical of GMO content, total GM cereal grain imports increase as preferences in Western Europe and high-income Asia turn against GMOs. This is because GM exports are now increasingly directed to non-critical markets (i.e. fewer markets), and so the import price declines even further than the price decline due to the factor productivity shock. Given competition from increased supplies of GM crops, prices of non-GM crops also fall, and so South America and sub-Saharan Africa also face declining prices on non-GM imports from the GM-critical regions as preferences shift. As with the import ban and preference shift scenarios in Nielsen and Anderson (2001a), the results of the NTR study also show that the ‘costs’ of the preference changes are borne mainly by the GM-critical regions themselves. In particular, the welfare increases (measured in total absorption) in all the developing country regions are not affected by the preference changes in the GM-critical regions. Low-income Asia is the major beneficiary in absolute terms, being both a net importer of the two crops and basically indifferent as to GM content. Hence, the region benefits from substantially lower import prices on GM crops. Despite the high dependence on the GM-critical regions for its exports of oilseed, the increase in total absorption in South America is unaffected by the preference changes there because bilateral trade flows adjust well – trade diversion offsets the effects of demand shifts in the GM-critical regions. In sub-Saharan Africa the gains are small in absolute terms, mainly due to the small share of these particular crops in production and trade, but they are also unaffected by preference changes in GM-critical regions. In the final scenario, where consumers in Western Europe and highincome Asia turn against genetically modified foods, the changes are much more dramatic compared with the scenario of assuming only reduced price sensitivity. The price of GM varieties in the GMO-critical countries declines further because of the almost complete rejection of these products, whereas the price of non-GM foods increases. This leads to substantially larger price wedges in the GM-critical regions as compared with the previous scenarios. The larger price wedges between GM and non-GM primary crops follow through the entire food processing chain. The price increase for non-GM foods is, however, moderated by the fact that there are markets for non-GM products in all regions in the model, so consumers are not closing themselves off to necessary goods, nor are they required to produce all the non-GM goods themselves. This is because the model allows all countries to produce both varieties and hence supply both GMO-indifferent and GMOcritical consumers, and assumes that they are able to credibly verify these characteristics to importers. Clearly, this is a simplification of reality, and one can easily imagine that for some regions, living up to the principles of identity preservation and verifying this is very costly, thereby putting them at a cost disadvantage. Such effects are not captured in this model.

276

C.P. Nielsen and K. Thierfelder

An interesting question is whether these changing preferences in Western Europe and high-income Asia can create opportunities for developing countries to export non-GM varieties of cereal grain and oilseed to these regions. Sub-Saharan Africa has some production of oilseed, for example, and although exports of these crops do not account for a significant share of total production value at present, they might increase if niche markets for non-GM crops develop in Western Europe. For a region like sub-Saharan Africa, with strong ties to Western Europe, changing consumer attitudes toward GM foods may be an important determinant of future decisions regarding genetic engineering in food production. Similarly, low-income Asian countries might look into expanding their production of, e.g. non-GM oilseeds, if nearby niche markets in highincome Asian countries develop.

Conclusions and Future Perspectives The use of modern biotechnology to create GMOs through agricultural production holds great promise but also presents important challenges. Developing countries have a particularly difficult decision to make. On the one hand, it may be argued that they cannot really afford to be critical about the introduction of new technologies and production processes that increase food supplies and raise the nutritional intake of their currently impoverished populations. On the other hand, developing countries must consider the potential for trade in GM products. Developing countries often have a very vulnerable trade structure, concentrated in a few products and export markets. This makes them highly dependent on consumer reactions and policy decisions related to GMOs in their key export markets. To be successful in ‘reaping the promise while leaping the pitfalls’,17 farmers, consumers and policy makers in developing countries must weigh the benefits of GM production for domestic use against the potential difficulties that may arise in export markets. What has been learned from the empirical analyses surveyed here? The world trade models surveyed differ in country and sector focus. They also incorporate different assumptions about how GM technology affects markets, with alternative treatments of market segmentation. Most models assume country specialization in either GM or non-GM commodities, while a few others assume countries can produce both and separate them in the supply chain. None of the models incorporate labelling and market segmentation costs, given the lack of good data at this point as to what those costs might be. The models also differ in terms of the assumed productivity gains associated with adoption of GM technology – again, there is as yet little quantitative information available.18 The models also vary in how they specify shifts in consumer attitudes toward GM commodities, which reflects different views about how these changes are going to play out in various countries. Finally, the models differ in how they capture the dissemination of the GM technology across countries, again reflecting different views about

Trade in Genetically Modified Food

277

how the process might occur in the future. At this point, the model applications are essentially prospective views, based on some empirical evidence (largely from partial studies) and educated speculation about how these trends might evolve in the future. As better information becomes available, the models will evolve to incorporate it. These caveats notwithstanding, a few robust results can be drawn from the various studies: ●

●

●

In general, adopters of the more productive GM technologies gain, while non-adopters or GM-critical regions either do not gain at all or only gain slightly (through changes in world prices), depending on how strongly they segment their markets. In particular, developing countries will benefit if they can adopt the new technologies, and get mixed results if they are non-adopters. Assuming – perhaps heroically – that the costs of labelling and market segmentation are not large, world markets can adjust relatively easily to the introduction of the new technologies. The price changes and changes in production and trade flows, while significant, are not dramatic. Developing countries are more sensitive to the issue, given their higher shares of agriculture in national product and of food in household consumption. Large developing countries, like India and China, gain from the new technology and are not much affected by changes in trade regime, since their domestic markets are large. Countries in subSaharan Africa, on the other hand, are very dependent on EU markets and would potentially be strongly affected by EU trade restrictions as more products relevant to their particular export structure have the potential of benefiting from GM technology. Also, countries like Brazil and other Mercosur partners face a delicate policy problem of which GM strategy to adopt in light of Europe’s hesitation.

In terms of future research, there are a number of knowledge gaps that require further study. Issues that are particularly important to any analysis of the impacts of GM technology on trade and on developing countries include: ●

● ● ●

●

What are the potential benefits of the next generation of GM goods, which might incorporate improved nutritional content (e.g. golden rice) or adaptability to particular climate conditions (e.g. drought tolerance) in developing countries? What are the productivity and cost gains from available and potential GM technologies? What are the likely outcomes of the policy debates regarding labelling requirements and market segmentation? What are the actual and potential costs associated with labelling requirements, identification of preservation and segmentation of domestic and world markets for GM and non-GM products? What are the implications of different intellectual property right regimes for the dissemination of the new technologies and for the distribution of the benefits?

278

C.P. Nielsen and K. Thierfelder ●

●

●

What will be the nature of consumer attitudes about commodities that incorporate GM technologies? What will be the impact on market demand? What will be the impact of these new technologies on market structure and degree of international competition? Will the new market structure follow the pattern of the pharmaceutical industry, with an oligopolistic structure requiring a period of protected monopoly markets in order to recoup on large, up-front research costs? How will the institutions of the world trading system, including the WTO, adapt to the evolution of these new technologies? Given these trends, is there a need for various new forms of ‘special and differential treatment’ for developing countries in the world trading system?

Clearly, modern biotechnology is not the entire solution to the food supply constraints and food insecurity problems in the developing world, but may well make a significant contribution if accompanied by appropriate policies. To date, transgenic crops have been developed and commercialized by a few large, private (multinational) corporations based in industrial countries. There is an urgent need to develop productivityenhancing and nutritionally improved crops that are specific to problems facing developing countries. In order for these to be made available and affordable for poor, small-scale and subsistence farmers – and thereby also contributing to alleviating food insecurity at the household level – the public sector may need to subsidize investment in relevant research. This is indeed also one of the main messages of the FAO (2004), i.e. biotechnology is capable of benefiting small, resource-poor farmers, but achieving this requires directing research and farm-level applications towards the particular needs and possibilities of these people. It is here that the changing locus of agricultural research from the public sector to the private multinational sector over the past decades becomes relevant. Private biotechnology firms in developing countries will have enough difficulty reaping rewards from investing in transgenic research on crops that are sold in markets, let alone from developing improved subsistence crops. This means that additional public financial resources may need to be mobilized for research either within the public sector and/or in collaboration with private companies. In fact FAO (2004) suggests that ‘the public sector may have to purchase the right to use private-sector technology on behalf of the poor’. However tractable this may sound in theory, it must be remembered that even conventional technologies that are on the shelf today have not yet reached the poorest farmers’ fields, and so there is no reason to believe that the new GM technologies will fare any better (FAO, 2004). The development community must develop a better understanding of what constrains small farmers’ access and use of new technology before investments in GM technologies directed towards their particular needs lead to widespread benefits for those producers. Apart from development of the technology itself and ensuring that farmers have access to it, there are serious regulatory challenges facing the developing countries. The risks of possible adverse effects on human and

Trade in Genetically Modified Food

279

animal health, on the environment and on biodiversity need to be managed better, by boosting the capacity of developing countries to formulate and enforce relevant food safety and biosafety regulations. Other regulatory challenges associated with the introduction of modern biotechnological products include anti-trust legislation and the securing of intellectual property rights and farmers’ rights. Management of intellectual property rights is crucial for encouraging private sector investment as well as for enabling access to basic technologies and their applications. Developing countries will, without doubt, require financial and technical support from donor countries to develop their capacities to tackle the challenges of modern biotechnology.

Notes 1

More generally, recent regulatory failures in the EU have raised consumer concerns about food safety (e.g. mad cow disease, dioxin-contaminated animal feed and foot and mouth disease), although none of these are related to GM foods. 2 This section draws on and updates information provided in Nielsen and Anderson (2000). 3 Definitions of genetic engineering vary across countries and regulatory agencies. In this context a broad definition is used, in which a genetically modified organism is one that has been modified through the use of modern biotechnology, such as recombinant DNA techniques. In the following, the terms ‘genetically engineered’, ‘genetically modified’ and ‘transgenic’ will be synonymous. 4 As long as private companies uphold patents on their transgenic seeds they will be able to extract monopoly rents through price premiums or technology fees. 5 This section draws on and updates information provided in Nielsen and Anderson (2000). 6 Until June 2005 there were three approvals: two maize products (one by Syngenta and other by Monsanto) and a rapeseed by Monsanto. 7 To date more than 70 countries have ratified the Cartagena Protocol on Biosafety. It is worth noting that this protocol is a part of the 1992 Convention on Biological Diversity to which, in particular, the USA is not a signatory. Nevertheless, it is expected that the USA will abide by the rules of this Protocol so as to facilitate trade. 8 See Anderson and Nielsen (2001) for a more thorough discussion of the potentially contentious issues as they relate to the WTO agreements. 9 This section draws on parts of a survey by Nielsen et al. (2002). 10 Scientific studies of the potential impacts on the environment, biodiversity or health are not reviewed here. 11 China found its exports of processed foods to the UK being restricted in 2000 because they may have contained traces of GM soybean imported by China from the USA. In turn, China placed a moratorium both on soybean imports from the US and on the use of GM varieties by its own farmers for food and feed (but not on cotton) production (Nielsen and Anderson, 2001b). 12 By distinguishing between GMO-inclusive and GMO-free products by country of origin, one concern may be that GM-adopting regions channel their exports to the country or region imposing the import ban (here, Western Europe) through Third World countries that are indifferent as to the content of GMOs and that do not adopt GM technology in their own production systems. The possibility of such trans-shipments is abstracted from in this analysis.

280

C.P. Nielsen and K. Thierfelder 13

This increase may in fact be larger if Western Europe’s possibilities of increasing its own production of oilseed are restricted by the Blair House agreement, signed in 1992 by the USA and the EU during the Uruguay Round and which was instrumental in reaching the 1994 final agreement for the WTO agricultural negotiations. 14 Burfisher et al. (2002) analysed a similar case of interdependent trade and domestic policy in Mexico, where the government sought to insulate maize producers from world market prices. When faced with cheap imports from the US (due to tariff elimination as part of NAFTA), Mexico’s programme costs became exorbitant. 15 This identity preservation feature represents a model extension compared to the authors’ previous work (Nielsen et al., 2001) where the GM/non-GM distinction only applied to the primary crop level. In that study GM and non-GM varieties were substitutable in intermediate and final demand. See Stone et al. (2002) for a similar treatment using the GTAP model. 16 The numbers for South America reflect GM use in Argentina. For Brazil, the adoption rate is much lower. 17 The phrase was the subtitle of a conference on the future of GMOs convened by the Biotechnology Impacts Center at the University of California, Riverside in October 2003. 18 Marra et al. (2002) make some effort to generalize from the individual studies at the farm level in the USA.

References Anderson, K. and Nielsen, C.P. (2001) GMOs, the SPS Agreement, and the WTO. In: Anderson, K. and McRae, C. (eds) The Economics of Quarantine. Centre for International Economic Studies, Adelaide and Biosecurity Australia, Canberra. Anderson, K. and Yao, S. (2001) China, GMOs and World Trade in Agricultural and Textile Products. Paper prepared for the Fourth Annual Conference on Global Economic Analysis, Purdue University, West Lafayette, Indiana, June 2001. Burfisher, M., Robinson, S. and Thierfelder, K. (2002) Developing countries and the gains from regionalism. American Journal of Agricultural Economics 84(3), 736–748. European Commission (2003) GMOs: Guidelines on Co-existence, MEMO/03/187. (EC website, accessed 29 September 2003). European Union in the US (2003) ‘GMOs: Commission Publishers Recommendations to Ensure Co-Existence of GM and nonGM Crops, News Release No. 46/03. (EU website, accessed 23 July 2003). FAO (2004) The State of Food and Agriculture:

2003–2004. Agricultural Biotechnology: Meeting the Needs of the Poor? Rome. Hertel, T.W. (1997) Global Trade Analysis: Modelling and Applications. Cambridge University Press, Cambridge, UK. Huang, J., Ruifa, H., van Meijl, H. and van Tongeren, F. (2002) Global Trade and Biotechnology Boosts to Crop Productivity in China. Paper prepared at Fifth Annual Conference on Global Economic Analysis, Taipei, Taiwan, 5–6 June 2002. ICTSD, (2003) Various newsletters on Biotechnology issues, available at the home page of the International Centre for Trade and Sustainable Development (ICTSD), http://www.ictsd.org/issarea/environment/ news/biotech.htm James, C. (2003) Global Status of Commercialized Transgenic Crops: 2002, ISAAA Brief No. 27. ISAAA, Ithaca, New York. James, C. and Krattiger, A. (1999) The role of the private sector. In: Persley, G.J. (ed.) Biotechnology for Developing/Country Agriculture: Problems and Opportunities: Focus 2: A 2020 Vision for Food,

Trade in Genetically Modified Food Agriculture, and the Environment. International Food Policy Research Institute, Washington DC. Marra, M.C., Pardey, P.G. and Alston, J.M. (2002) The Payoffs to Agricultural Biotechnology: An Assessment of the Evidence, International Food Policy Research Institute (IFPRI), Environment and Production Technology Division (EPTD) Working Paper No. 87, Washington DC. Nielsen, C.P. and Anderson, K. (2000) GMOs, Trade Policy, and Welfare in Rich and Poor Countries. CIES Discussion Paper No. 0021: Centre for International Economic Studies, University of Adelaide and Working Paper No. 3/2000: Danish Research Institute of Food Economics (FØI), Copenhagen. Nielsen, C. and Anderson, K. (2001a) Global market effects of alternative european responses to GMOs. Weltwirtschaftliches Archiv (Review of World Economies) 137(2), 320–346. Nielsen, C.P. and Anderson, K. (2001b) Genetically Modified Foods, Trade, and Developing Countries: Is Golden Rice Special? Invited paper prepared for a seminar entitled Evaluating the Impact of Genetically Engineered Food on the Health and Economy of Rural Populations, University of Michigan, Michigan, USA, 12 November 2001. Nielsen, C.P., Robinson, S. and Thierfelder, K. (2001) Genetic engineering and trade: panacea or dilemma for developing countries? World Development 29(8), 1307–1324. Nielsen, C.P., Robinson, S. and Thierfelder, K. (2002) Trade in genetically modified food: a survey of empirical studies. Paper prepared for presentation at an OECD Working Group on Cereals, Animal Feeds and Sugar of the Working Party on

281 Agricultural Policies and Markets, 15–16 April 2002. Nielsen, C.P., Thierfelder, K. and Robinson, S. (2003) Consumer preferences and trade in genetically modified food. Journal of Policy Modeling 25(8), 777–794. OECD (1999) Modern Biotechnology and Agricultural Markets: a Discussion of Selected Issues and the Impact on Supply and Markets. Directorate for Food, Agriculture and Fisheries. Committee for Agriculture. AGR/CA/APM/CFS/MD (2000) 2, OECD, Paris. Pray, C.E., Ma, D., Huang, J. and Qiao, F. (2000) Impact of Bt Cotton in China. Paper presented at a seminar at the International Food Policy Research Institute (IFPRI), 9 May 2000. Stone, S., Matysek, A. and Dolling, A. (2002) The Implications of GMOs for Australian Trade, Paper prepared for the Fifth Conference on Global Economic Analysis, Taipei, Taiwan. UNEP (2000) Cartagena Protocol on Biosafety to the Convention on Biological Diversity. http://www.biodiv.org/biosafe/biosafetyprotocol.htm van Meijl, H. and van Tongeren, F. (2001) International Diffusion of Gains from Biotechnology and the European Union’s Common Agricultural Policy, Paper prepared for Fourth Annual Conference on Global Economic Analysis, Purdue University, West Lafayette, Indiana, 27–29 June 2001. WTO (1998a) Understanding the WTO Agreement on Sanitary and Phytosanitary (SPS) Measures. May. WTO, Geneva. WTO (1998b) Goods: rules on NTMs. In: WTO: A Training Package. Module 3. 15 December. WTO, Geneva.

13

Is the Everything But Arms Initiative the Way to Go for Least Developed Countries in the WTO Negotiations?*

WUSHENG YU AND TRINE V. JENSEN Danish Research Institute of Food Economics, Rolighedsvej 25, 1958 Frederiksberg C, Denmark

Introduction The Everything But Arms initiative (EBA) of the European Union (EU) has been adopted for the purpose of further improving Least Developed Countries’ (LDCs) access to the EU market. Under the EBA, all restrictions on imports except arms from the LDCs to the EU are to be removed on a non-reciprocal basis (European Community, 2001a). This initiative is a continuation and enhancement of the non-reciprocal preferences granted by the EU under the ACP (Africa, Caribbean and Pacific) preferences and the Generalized System of Preferences (GSP) schemes. It is considered by the EU as a very significant step in fulfilling the Doha Development Agenda (WTO, 2001a). In fact, it has been included in the EU’s WTO agricultural negotiation proposal (European Commission, 2002). Reactions to the EBA by other countries vary. For example, a negotiation proposal released by the US Department of Agriculture in 2002 does not contain such an offer, whereas many of the African countries continue to ask for special and differential treatment similar to the EBA initiative (WTO, 2001b). Despite the different reactions to the EBA, the idea of granting full market access in agricultural goods to LDCs, since it was proposed by the EU, has been incorporated into different negotiating proposals, including the one presented by the then Chairman of the Committee in Agriculture, Stuart Harbinson (WTO, 2003). This study is part of the more general issue of how the granting or the erosion of preferences may affect the economies of developing countries, and also the negotiations of the Doha Round. The specific objectives of this study are twofold. First, we try to provide an economic assessment of the 282

© CAB International 2006. WTO Negotiations and Agricultural Trade Liberalization: the Effects of Developed Countries’ Policies on Developing Countries (eds E. Diaz-Bonilla et al.)

The Everything But Arms Initiative

283

likely impact of the EBA on the LDCs. Several earlier studies have examined this issue but consensus on its quantitative effects has not yet been achieved.1 The different quantitative results may be attributed to the fact that the tariff cuts employed and the bases from which these cuts are applied vary in those studies.2 As such, a further evaluation of the EBA, by accurately taking account of existing preferences, is warranted. The second objective is to examine how the ongoing WTO negotiations affect potential benefits from the EBA. WTO negotiations aim at achieving a fair and market-oriented trading system by strengthening rules and reducing support and protection on the one hand, and assisting economic development in developing countries on the other. Offering preferential market access opportunities to the LDCs and allowing them to enjoy reduced reform schedules, or even exempting them from such requirements, are the primary tools used to achieve the latter. However, benefits from this ‘differential and preferential treatment’ will be directly affected by any actions taken at the multilateral level. For the EBA initiative, this implies the possibility of preference ‘erosion’ when further multilateral liberalization takes place. As the EBA is merely a continuous step toward duty-free access for LDCs, a point shown below, its real benefits may well be wiped out by any ambitious multilateral trade reform. Therefore, it is important to empirically investigate this possibility by examining the impact of the EBA both with and without possible multilateral trade liberalization under the WTO. This rest of the chapter is structured as follows. The next section offers a detailed description of the EBA. The third section briefly discusses the model and data used for our quantitative assessment of the EBA, and the following section starts with the design of a series of policy scenarios, followed by simulation results from these scenarios, including analyses of the impact of the EBA itself, and the joint impact of the EBA and a series of possible EU reform scenarios. A summary of the welfare costs associated with keeping the EBA ‘meaningful’ concludes this section. Concluding remarks are offered in the final section.

The ‘Everything But Arms’ Initiative in the EU’s Preferential Trading Framework The EU maintains a complex system of regional and preferential trading agreements (PTA), which range from custom unions such as the European Union itself, free trade areas such as the European Economic Area and the Mediterranean Partnerships, to various non-reciprocal trade preferences. This ‘multi-layer’ system extends to virtually all its trading partners, except for six countries/regions, which enjoy the Most Favoured Nation (MFN) status (i.e. the basic, non-preferential treatment under the WTO).3 Among these agreements, the ACP preferences and the GSP schemes are the two sets of non-reciprocal trade preferences concerning the LDCs.4 They are the predecessors of the EBA.

284

W. Yu and T.V. Jensen

The ACP preferences: from Lomé to Cotonou The first formal agreement between the EU and the ACP countries that granted non-reciprocal trade preferences to the latter was established in 1975 under the first Lomé Convention. The structure and coverage of these preferences have changed through several generations of the Lomé Convention, the latest of which covered 77 countries, including 40 LDCs designated by the UN (European Commission, 2003a),5 and granted them almost complete duty-free access to the EU market. Exceptions in the form of tariffs and/or quantitative or seasonal restrictions were maintained mainly on agricultural products. The so-called ‘sensitive products’ such as sugar, beef and veal, bananas and rum imports have been, however, regulated through separate protocols. Through several generations of Lomé Conventions, the impact of ACP preferences has not been impressive. Even though for most ACP countries the EU has been their main trading partner, with an aggregate export share of 29 per cent in 2000, their share in EU’s total imports declined from 13.3 per cent in 1976 to 3.7 per cent in 2000. Moreover, exports from the ACP to the EU have been concentrated in raw materials and agricultural products. This unimpressive performance led to a redesign and renegotiation of the Lomé Convention, which resulted in replacing Lomé IV by the Cotonou agreement. At present, however, the trade component of the Cotonou agreement is only a framework for the EU and the ACP countries to renegotiate the so-called Regional Economic Partnership Agreements (REPA), with the aim of liberalizing all bilateral trade between them on a reciprocal basis, thereby making the agreements fully WTO-compatible. In the meantime the present trading regime will be maintained during the preparatory period (from 2002 to 2008 at the latest). After that, the REPAs will enter into force gradually, with a transition period of 12 years. Until now, only the EU and South Africa have successfully negotiated a REPA.

The generalized system of preferences The GSP scheme has been allowed under GATT first, and now under the WTO, through the Enabling Clause, which states that preferential treatment for developing countries under the GSP has to be nondiscriminatory, non-reciprocal and autonomous. The non-discriminatory requirement is exempted for LDCs, which can therefore receive better treatment than other developing countries. The EU was the first to implement such a scheme in 1971. Currently, 146 countries and 25 territories, including all the LDCs,6 benefit from the European GSP scheme (European Community, 2001b). The scheme operates at two levels: (i) the general arrangement provides basic trade preferences based on traditional objectives of economic development; and (ii) special incentive clauses regarding

The Everything But Arms Initiative

285

sustainable development. The former provides tariff preferences depending on the sensitivity of the product. According to EU regulations (European Community, 2001b), preferences are granted according to two differentiated product categories: for non-sensitive products, duty-free access is granted; for sensitive products, the reduction of the ad valorem duty is 3.5 percentage points of the corresponding MFN rate, whereas specific tariffs are to be reduced by 30 per cent. The Common Custom Tariff Schedule of the EU covers approximately 10,500 HS-8 tariff lines in total, and the MFN rates are zero for at least 2000 of these tariff lines. The GSP scheme covers around 7000 of the dutiable products, of which 3300 are classified as non-sensitive and 3700 are classified as sensitive (European Commission, 2003b). The special incentive clauses of the GSP are related to the respect of social rights and protection of the environment. With these clauses, the preferences as specified in the general arrangement can be either increased or reduced. For example, countries that respect social and environmental standards in relevant international agreements receive more favourable treatment, often in forms of doubling the normal preferences. In addition, the EU is also allowed to reduce or withdraw the preferences for individual countries or specific sectors if they reach a certain level of competitiveness. The benefit of the GSP scheme may also be suspended if certain safeguard clauses are triggered. In summary, together with stringent rules of origin, these clauses give the EU the flexibility to adjust the magnitude and coverage of the preferences, thereby increasing the uncertainty for the countries participating in the system. The LDCs have been part of a special arrangement in the GSP scheme for many years. These countries have traditionally received more favourable treatment and have enjoyed duty-free access for a large part of industrial and agricultural products. Before the EBA, there were only slightly over 900 products (HS-8 tariff lines) that were subject to duties. Furthermore, since 1998 the EU has tried to extend similar preferences as contained in the ACP preferences to the nine LDCs that are part of the GSP scheme but not part of the ACP.

A detailed description of the EBA The initiative to further improve market access for LDCs was followed up in February 2001 when the EU formally adopted the ‘Everything But Arms’ regulation (European Community, 2001a). As a direct extension of the GSP,7 the EBA amended the GSP scheme in several ways. First, the EBA offers the LDCs a higher degree of market access. For virtually all products this means quota- and duty-free access. Second, it grants all LDCs duty- and quota-free access to the EU market, whereas the GSP scheme provides different treatments by country and product. Third, unlike the GSP scheme, the EBA has no time limit. However, the EU has the right to review the performance of the EBA and may introduce amendments.

286

W. Yu and T.V. Jensen

Country and products coverage All the 49 LDC countries as designated by the United Nations are included in the EBA. Under the EBA, all restrictions, including tariff and Tariff Rate Quotas (TRQs), on virtually all exports (except for arms) from LDC to the EU are to be removed, without the reciprocal action by the LDCs on exports from the EU. As the existing preferences (e.g. the ACP preferences and the GSP scheme) have already led to quota- and duty-free access for most products in LDCs, the EBA affects only 919 HS-8 tariff lines. Upon fully implementing the EBA, restrictions on exports from the LDCs remain for just 25 HS-8 tariff lines related to the trade with arms. Table 13.1 illustrates the distribution of EBA products across product categories according to the classification of the Global Trade Analysis Project database (GTAP8 for short). It can be seen that virtually all the EBAaffected products relate to agriculture and food. In terms of HS-8 tariff lines (Columns 4 and 5 of Table 13.1), meat products, processed food products and dairy contain the most tariff lines (each with over 10 per cent of total EBA lines). The categories of beverage, vegetables and fruit and bovine meats are also important components of the EBA. Sensitive products in the EBA Although the duty- and quota-free access has been granted for 876 of the 919 products with immediate effect, this treatment for the remaining items (43 lines) will take place in three progressive stages. These items belong to the three sensitive product groups, namely, sugar, rice and bananas. Specifically, for sugar and rice, duties will be reduced in three stages and eliminated by 2009. For bananas, duties will be reduced by 20 per cent annually, starting on 1 January 2002 and ending on 1 January 2006. To compensate for these delays, the EU will offer immediate and real market access through the creation of duty-free quotas for sugar and rice, based on the best figures for LDC exports during the 1990s plus 15 per cent. These quotas will be increased by 15 per cent each year during the interim period, eventually reaching 6,696 and 197,335 t, respectively, in 2009 (European Community, 2001a). A close look at concessions offered in the EBA To assess the significance of the EBA, it is important to consider the trade and protection situation of the EBA-affected products prior to the implementation of this preferential agreement. For this purpose, we first examine trade data on exports from the LDCs to the EU (in particular, EBA-affected exports), before calculating the tariff concessions implied by the EBA. Figure 13.1 illustrates the export values of major products from the LDCs to the EU in 1999. Only four aggregated products have export values exceeding US$100 m. These are processed food products, other crops, plant fibres and vegetables and fruit. Exports of fish, sugar, oilseed, animal

Number of HS items GTAP categories Sugar Veg., fruit and nuts Cereal grains etc. Vegetable oils and fats Food products etc. Beverages Processed rice Meat products etc. Bovine meat products Animal products etc. Paddy rice Dairy products Bovine cattle, sheep and goats, horses Crops etc. Plant-based fibres Chem., rubber and plastic Fishing Oilseeds Wool, silk cocoons Wheat Sugar cane and beet Total

Total (HS-6)a

EBA (HS-6)b

EBA (HS-8)c

% of all EBAd

7 89 10 47 248 30 2 43 29 48 2 24 8 62 8 987 40 15 6 2 2

7 19 9 7 102 8 2 33 19 9 2 23 3 0 0 5 0 0 0 2 2

9 35 12 10 290 108 17 153 48 19 16 165 14 0 0 17 0 0 0 3 3

1.0 3.8 1.3 1.1 31.6 11.8 1.8 16.6 5.2 2.1 1.7 18.0 1.5 0.0 0.0 1.8 0.0 0.0 0.0 0.3 0.3

1,709

252

919

75–103 2.3 37.1 0.2 2.1–2.5 1.2 87.4 9.7–19.2 9.7–19.2 2.8–4.5 61.6 51.2 2.8–4.5 0.0 0.0 – 0.0 0.0 0.0 – –

Export of EBA products (US$1,000)f

Total LDC exports (US$1,000)g

Share of EBA exports (%)

67,618 17,203 11,283 1,667 475 432 399 378 217 149 38 21 16 0 0 0 0.0 0 0.0 0 0

67,618 124,124 11,283 88,401 680,237 1,379 399 940 217 39,055 38 36 16 946,585 245,746 156,942 85,439 65,298 9,531 0 0

100.0 13.9 100.0 1.9 0.1 31.3 100.0 40.2 100.0 0.4 100.0 58.3 100.0 0.0 0.0 0.0 0.0 0.0 0.0 n/a n/a

99,896

2,523,284

4.0

287

Notes: a Number of total HS-6 lines for each GTAP category. b Number of HS-6 lines affected by the EBA (the Trains database). c Number of HS-8 lines affected by the EBA. d % share in all EBA-affected HS-8 lines. e Tariff rates adopted from Bora et al. (2002). f Total exports of EBA products from the LDCs to the EU in 1999 (Trains). g Total exports from the LDCs to the EU in 1999 (Trains).

100

Ave. preEBA tariff (%)e

The Everything But Arms Initiative

Table 13.1. The EBA: an illustration of product coverage, tariff rates and LDC exports.

W. Yu and T.V. Jensen

946.6

1000.0 900.0 800.0 700.0 600.0 500.0 400.0 300.0 200.0 100.0 0.0

680.2

245.7 65.3

39.1

lg

ra

in

uc ts

11.2

m

ea er

O

th e

ra

ni

C

al

pr

O

od

ils ee

ga

r

d

67.6

Su

oi et ab

le

fru d an s

le et ab

Ve g

85.4

Ve g

fib as ed

ls

its

s re

d fo o t-b an Pl

Pr

oc es se d

rc ro th e O

88.4

Fi sh

124.1

ps

Million US$

288

Fig. 13.1. Exports of agricultural products from LDCs to the EU, 1999. Source: Authors’ aggregation based on data from the TRAINS database*.

products and cereal grain are between 10 and 100 million dollars. Total agricultural and food exports from the LDCs to the EU amounted to just over US$2.5 bn in 1999 (representing a quarter of all merchandise exports from the LDCs to the EU). Exports from the LDCs constitute a small fraction of total imports of the EU, suggesting that the impact of the EBA might be small for the EU. The fact that the EBA involves only a small fraction of the EU’s tariff lines (919 HS-8 lines, to be exact) can be best seen in Table 13.1, where the product coverage of the EBA, the associated tariff concessions (in terms of the differences between pre-EBA average tariff rates and zero) and export values are tabulated. Starting with sugar, which has the highest export value of over US$67 m. in 1999, the first three columns show that all the seven HS-6 sugar items are covered by the EBA and therefore are all subject to tariff reduction. Pre-EBA tariff rates for the sugar items range from 75 to 103 per cent, indicating substantial potential gains from fully implementing the EBA. However, the reservations by the EU, in terms of gradual removal of the import restrictions on sugar, will probably delay such gains. The second row in Table 13.1 concerns vegetables and fruit. Although the export value of the EBA component of this category (mainly bananas) accounts only for 14 per cent of its total trade, it still reaches US$17 m. However, the fact that the pre-EBA tariff rates were only 2.3 per cent indicates that potential gains to the LDC are likely to be much smaller, as compared to sugar. The category of cereal grains is the third largest EBA category, covering nine out of the possible ten HS-6 lines and all of ts export value. The pre-EBA tariff rates of 37 per cent imply that there is much to be gained from the EBA. The table shows that there may be some potential for rice as well. Pre-EBA tariff rates for rice were over 87 per cent and its export *

The TRAINS database, or Trade Analysis and Information System, is compiled by UNCTAD.

The Everything But Arms Initiative

289

value in 1999 was about US$0.4 m. Again, rice is listed as a sensitive product and is subject to a special phasing-out clause in the EBA. Other than the above-mentioned products, the EBA seems to have minor potential effects, either due to the very low pre-EBA export values, or due to the very low pre-EBA tariff rates. Overall, the export value of all the EBA-affected products amounts to just below US$100 m. in 1999, representing only a 4 per cent share in total agricultural and food exports from the LDCs to the EU. For sugar, cereal grains and rice, there seems to be some scope of further export expansion for the LDCs after implementing the EBA. However, the special arrangement regarding sugar, rice and bananas will probably delay such gains. For other EBA products, the potential of export expansion seems to be limited. Safeguard measures in the EBA The safeguard measures specified in the GSP are largely retained in the EBA, with some amendments. Most notable among the amendments is the addition of the situation of ‘massive imports into the EU market’ as a trigger for withdrawing the preferences. With regard to the three sensitive products (sugar, bananas and rice), the EU is allowed to suspend the preferences entirely if imports cause ‘serious disruptions’ to the EU’s mechanisms that regulate these products. In addition, the rules of origin specified in the GSP also apply to the EBA initiative. These safeguard measures and rules of origin are clearly reminiscent of the GSP. Whether or not the objective of the EBA can be achieved in the longer run is certainly contingent on how these measures are applied in reality. For this reason, some caution should be exercised when evaluating the potential effects of the EBA.9

Methodology and Data We now turn to the quantitative assessment of the impact of the EBA. This is done by simulating various relevant policy scenarios and by examining the simulated results from these scenarios. Due to the multi-regional and multi-sectoral nature of the EBA initiative, we conduct such simulations within a global computable general equilibrium modelling framework. The GTAP model (Hertel, 1997) and database are utilized for the simulations. The GTAP model is a standard global model that features inter-sectoral linkages through nested Constant Elasticity of Substitution production functions and international trade linkages through the so-called Armington specifications (Armington, 1969). The demand side of the model is featured by a Constant Difference Elasticity demand system. Standard neoclassical assumptions such as constant returns to scale, perfect competition, profit and utility maximization are applied. With this modelling structure, explicit welfare analysis and decomposition are possible.

290

W. Yu and T.V. Jensen

This study applies an aggregated version of the most recent GTAP database (Dimaranan and McDougall, 2002), covering 19 countries and aggregated regions and 20 aggregated sectors. As our focus is on the LDCs in Africa, the aggregation we choose includes five individual LDC countries, namely, Malawi, Mozambique, Tanzania, Zambia and Uganda. Other LDCs10 are largely included in the aggregated rest of sub-Saharan African (SSA) regions.11 The other countries/regions included in the aggregation are: the USA, the EU, Japan, Canada, Australia and New Zealand, Eastern Asia, South-east Asia, Southern Asia, Eastern and Central Europe, Latin America and the Caribbean, Middle East, Rest of Africa and Rest of the World. Most of the agriculture and food sectors in the GTAP database are incorporated as they appear individually (i.e. without being aggregated further). They include paddy rice, wheat, cereal grains, vegetable and fruit,12 oilseed, plant fibres, other crops, other animal products, beef, other meats, vegetable oil, dairy, processed rice, sugar and other processed food products. In addition to these agricultural commodities, the model also includes non-agricultural products that are aggregated into natural resources, textile and clothing, manufacturing and services. Agricultural protections are built in the database in the forms of import tariffs, export subsidies and various domestic support measures (output subsidies, intermediate input subsidies, factor-based payments and others). Since the existing preferential arrangements between the EU and the LDCs are not adequately represented in the database, we need to adjust the tariff rates in the original GTAP database so that the preferential tariff rates actually being applied are accounted for. After this adjustment, the starting point of the subsequent simulations contains the same tariff rates as listed in Column 6 of Table 13.1.13

Policy Scenarios and Results Using the GTAP model and the adjusted database, three sets of policy scenarios are simulated. The first scenario (called the EBA Scenario) is a simple comparative static scenario in which the EU’s import tariff rates for all commodities imported from the six LDCs in our aggregation are eliminated. Note that the tariff cuts are not identical across the EBA products. According to Table 13.1, the biggest cuts are for sugar, cereal grains, rice, dairy and livestock products. The second set of scenarios (Scenarios 2, 3 and 4) is intended to illustrate how the EBA preferences are ‘eroded’ in the presence of a series of ‘plausible’ reform scenarios, in which the EU reforms its own trade policies in the three major negotiation areas on a Most Favoured Nation (MFN) basis14. Specifically, in Scenario 2 (Market Access Scenario), on top of the EBA shocks (as simulated separately in Scenario 1), tariff rates on imports from non-LDC countries to the EU are reduced by 50 per cent. In Scenario 3 (Export Subsidy Scenario), on top of the EBA shocks, all subsidies associated with the EU’s exports (to all destinations) are removed. In

The Everything But Arms Initiative

291

Scenario 4 (Domestic Support Scenario), on top of the EBA shocks, all domestic support measures of the EU are reduced by 35 per cent.15 Results from each of these scenarios show the joint impacts of the EBA and the reforms in the corresponding negotiation area. In the last scenario (Scenario 5), in addition to carrying out the EBA reform, the EU is also assumed to simultaneously conduct the same policy reforms in the three negotiation areas as simulated separately in Scenarios 2, 3 and 4 (i.e. reducing import tariff rates imposed on agricultural and food products from non-LDC regions by 50 per cent; removing agricultural export subsidies; and reducing all agricultural domestic supports by 35 per cent). Therefore, results of this scenario can be used to summarize the combined impacts of the EBA and the hypothetical EU reforms in all three negotiation areas. It is worth noting that in the last four scenarios mentioned above (i.e. Scenarios 2, 3, 4 and 5), the EU is assumed to reform its policies unilaterally and the non-LDC countries/regions are assumed to maintain their own protection measures. In doing so, we intend only to capture the ‘erosion’ the preferences granted to the LDCs by the EU.

Potential gains from the EBA (Scenario 1) In Scenario 1, we ‘implement’ the EBA by removing all tariff and quantitative restrictions associated with 919 HS-8 tariff lines. This is necessarily a simplistic treatment of the actual initiative, especially with regard to the three sensitive products. Therefore, the results reflect a more optimistic picture of the reforms. Exports from the LDCs will benefit from such a move, especially for the sectors facing deeper tariff cuts. Output and production patterns of the LDCs will also be affected due to these countries’ relative specialization in certain sectors. In addition, this reform will also carry some macro implications for the LDCs. Effects on trade and production Table 13.2 reports the changes in LDCs’ exports to the EU and the corresponding post-EBA export values, as resulted from implementation of the EBA initiative (Scenario 1). Not surprisingly, in terms of percentage changes, exports increase the most for the sectors with the highest tariff cuts (i.e. wheat, cereal grains, meats, dairy, rice and sugar). The most notable changes are for sugar, whose exports reach US$117 million, 183 million, 165 million and 472 million in Malawi, Tanzania, Zambia and rest of sub-Saharan Africa, respectively. However, due to the low pre-EBA exports of wheat, cereal grains and dairy, high percentage changes (shown in the upper panel of Table 13.2) do not translate into significant increases in dollar terms. As can be seen from the middle panel of Table 13.2, post-EBA export values for these products remain quite small. In contrast, exports of the products with smaller tariff cuts (oilseed, plant fibres, other crops and vegetable oil) decrease slightly.

292

Table 13.2. EBA’s impact on exports from LDCs and developing countries to the EU (Scenario 1).

Changes in export quantities, percentage from base data Malawi 240 238
5
12 Mozambique 281 277 6
3 Tanzania 257 242
4
14 Zambia 301 227
7
23 Uganda 296 292 9
1 SSA* 290 286 7
3 Australia & New Zealand
0.1
0.8
0.3
0.1 Japan
0.1
0.8
0.2 0 ASEAN
0.1
0.8
0.2 0 S. Asia
0.1
0.8
0.3 0 E. Asia
0.1
0.8
0.2 0 USA
0.1
0.8
0.2 0 Canada
0.1
0.8
0.2 0


10
2
8
15
1
2 0.4 0.5 0.4 0.5 0.5 0.5 0.5


11
4
12
19
1
2 0.4 0.4 0.3 0.4 0.4 0.4 0.4


1 13
17
6 18 19
0.2
0.2
0.2
0.2
0.2
0.2
0.2

Post-EBA export values, US$ million Malawi 0 Mozambique 0 Tanzania 0 Zambia 0.1 Uganda 0 SSA 3.6

4 5.3 10.7 2.4 2.3 160.2

5.5 22 125.5 10.4 18.6 1173.5

389.1 5.2 212.3 25.6 441.3 3988.8


4.1
1.2
1.4 1.6 0
0.1


9.6
2.4
7.7
5.9
0.4
1.1


8.4 0.1
3.1
0.4
0.6
0.8

3.9 27.8 72.5 7.5 11.1 766.4

Changes in volumes of output, percentage from base data Malawi 0.3 0.4 0.1 Mozamb.
0.5 0.2 0 Tanzania
1.9 1.2
2.1 Zambia 3.5 0.1
1.1 Uganda 0.1 0.1 0 SSA 0.3 0.4 0.7

Source: simulation results from Scenario 1. All changes are based on the base data. * SSA refers to Rest of sub-Saharan Africa.

Other meats

Veg. oil

Dairy

Rice

24 43 35 29 114 75
0.1
0.1
0.1
0.1
0.1
0.1
0.1

25 43 35 28 114 75
0.1
0.1
0.1
0.1
0.1
0.1
0.1


15
3
9
15 0
1
0.1
0.2
0.1
0.1
0.1
0.1
0.1

417 492 435 449 506 500
0.1
0.1
0.1
0.1
0.1
0.1
0.1

1120 1288 1146 1197 1345 1320
8
8
8
8
8
8
8

0.6 0.3 13 0.8 8.2 113

0 1.4 2.2 0.3 0.6 6.3

0.1 0.1 5.2 0.6 0 5.9

0.5 0.5 5.2 0.4 0 242.6

0.3 1.5 0.2 0.1 2.4 7.8

0.6 0.1 4.8 0.5 0.7 128.9

116.7 20.3 183.4 164.9 1.4 471.9

3.3 0.5
0.2 0.3 0.3 0.3


0.3
0.9 0.7
0.3 0.9 0.3

0.1 0.1 0.7
0.1 0.1 0

0
0.4
5.4
0.2
0.3
0.2

0.8
1.6 0.7 0.2 0.5 0.1


0.7 0 6.4 0.5 1.3 0.9

397.1 16.3 39.3 268.4 1.4 14.8

Sugar 679 859 1562 517 1145 895
15.6
15.6
15.6
15.6
15.6
15.6
15.6

W. Yu and T.V. Jensen

Other crops

1.4 3.7 22.3 1.5 1.8 33.6

Oilseed

Other animal Bovine products meats

Plant fibres

Wheat

Cereal Veg. and grains fruit

The Everything But Arms Initiative

293

Production patterns in the LDCs are similarly impacted by the EBA. As shown in the lower panel of Table 13.2, output of sugar increases dramatically – in Malawi and Zambia it is more than doubled, whereas in Mozambique, Tanzania and sub-Saharan Africa, it is increased by far lower percentages. Outputs of rice and cereal grains also increase in most LDCs. On the other hand, decreases in some other products are also observed, especially for plant fibres and other crops. This is because the sharp decreases in their exports dampen total demand for these products. Also, expansion of outputs in other sectors leads to an outflow of resources from these sectors. Aside from the above-mentioned products, output changes in other products vary across countries. Most of these changes are quite small and can be attributed to the adjustment on the export side. These results demonstrate that the LDCs do have some potential in producing certain products. Moreover, as exports from the LDCs constitute only a small fraction of the total imports of the EU, the EBA’s impact on the EU and developing countries appear to be quite insignificant. As shown in Table 13.2, exports from non-LDC countries to the EU generally decrease slightly (in the range of 0.1 to 0.8 per cent), and for a few products such as plant fibres and other crops, bilateral exports from non-LDC countries actually increase slightly, due to the decreases in exports from the LDCs. Rice and sugar are the lone exceptions, where notable decreases are observed (e.g. 15 per cent decrease in sugar exports from developing countries). Macroeconomic and welfare effects Implementing the EBA will lead to welfare improvement for the LDCs, as shown in Table 13.3. GDP grows by 4.7 per cent, in Malawi, 2.4 per cent in Zambia, 1.1 per cent in Mozambique and 0.5 per cent in the rest of subSaharan Africa. Changes in utility level range from 1.2 per cent in Malawi to 0.1 per cent in sub-Saharan Africa. In monetary terms (equivalent variation), welfare gains are the highest in sub-Saharan Africa (US$169 m.), followed by Tanzania (US$59 m.), Malawi (US$30 m.), Zambia (US$11 m.), Mozambique Table 13.3. Macroeconomic and welfare effects of the EBA (Scenario 1). Welfare (EV in US$ m.)

EU Malawi Mozambique Tanzania Zambia Uganda SSA World

GDP (%)

Utility (%)

Total

Efficiency

0 4.7 1.1 3.5 2.4 0.3 0.5 n/a

0 1.2 0.3 0.9 0.3 0 0.1 n/a


184 30.4 8.5 59.7 11.3 2.4 169.2 38.2


2.2 9.5 1.3 15.2
13.5 0.2 29.4 39.7

Source: simulation results from Scenario 1. Note: SSA = sub-Saharan Africa.

Terms of trade
178.5 22.7 3.8 32.2 26.1 1.5 117.5
1.3

294

W. Yu and T.V. Jensen

(US$8 m.) and Uganda (US$2.4 m.). Although these gains in dollar terms are quite limited, for certain smaller LDCs, such as Malawi and Tanzania, their macro-impacts are non-negligible. These changes in welfare can be attributed primarily to allocation efficiency effects and terms-of-trade effects. Allocation efficiency is improved for the LDCs due to the reallocation of resources to the products with more trading potential, while terms of trade improve with higher relative exporting prices. Both of these effects are directly linked to the unilateral removal of import tariffs by the EU on a number of agricultural and food products. Table 13.3 shows that all LDCs benefit from a terms of trade improvement and in most LDCs this effect dominates the total welfare results. For instance, over US$117 m. of the total US$169 m. gains for rest of sub-Saharan Africa are due to improved terms of trade. Allocation efficiency also improves for all LDCs except Zambia, which suffers an efficiency loss of over US$13 m.16 For the EU, a total welfare loss of US$184 m. occurs, almost entirely due to the unfavourable terms-of-trade effects. However, this loss is negligible for the EU and, in fact, there are no noticeable losses in either its utility or GDP. Most non-LDC regions (not shown in the table) experience very minor losses in welfare but no noticeable decline in either utility or GDP, signaling that the trade diversion effects of the EBA are quite small. The aggregated welfare effects of US$38 m. for the whole world appear to be positive but very small. Role of the sensitive products As highlighted in the previous section III, the three sensitive products (sugar, bananas and rice) are of special importance to the LDCs. To see how the delayed liberalization of these products affects the overall impact of the EBA, we simulate a supplementary scenario where the pre-EBA tariff rates for these products are maintained while those for other products are eliminated. Under such a scenario, total welfare gains are reduced markedly (presented in Table 13.4). For sub-Saharan Africa, the welfare gains of US$41 m. are less than a quarter of the gains available from a ‘full’ EBA scenario, whereas for Malawi and Tanzania, the gains are less than 10 per cent of those under the EBA scenario. A decomposition of the welfare results shows that except for sub-Saharan Africa, there are no significant allocation efficiency gains and the terms-of-trade effects are also minimal. In terms of utility and GDP, all the six LDCs experience less than 1 per cent increases. In contrast, the EU suffers a far smaller loss (US$31 m. as compared to a US$184 m. loss in Scenario 1). These results show that the EBA initiative is not likely to generate any significant welfare gains for the LDCs because of its limited product coverage and its limited tariff concessions (given the preferences that had been granted in the past). Furthermore, a great deal of the gains will probably come from the three sensitive products (especially sugar) that are

The Everything But Arms Initiative

295

Table 13.4. Macroeconomic and welfare effects of the EBA without sensitive products. Welfare (EV in US$ m.)

EU Malawi Mozambique Tanzania Zambia Uganda SSA World

GDP (%)

Utility (%)

Total

Efficiency

Terms of trade

0 0.1 0.7 0.4 0 0.2 0.1 n/a

0 0 0.2 0.1 0 0 0 n/a


30.8 0.3 5.4 5.9 0 1.3 41.8 10.2

0.7 0.1 0.8 1.1
0.1 0.1 8.9 10.3


30.3 0.3 2.5 3.6 0.1 0.9 27.6 0

Source: simulation results. Note: SSA = Sub-Saharan Africa.

subject to delayed implementation. Finally, the negative impact on the EU and non-LDC countries seems to be quite small, indicating that granting duty- and quota-free access to the LDCs is of no financial burden to the EU and the non-LDC countries.

Preference erosion and impacts of possible EU reform scenarios (Scenarios 2–4) The attractiveness of the EBA initiative and its predecessors has been made possible largely by the protection measures that the EU maintains against other exporters. In fact, the multi-instrument Common Agricultural Policy (CAP17) of the EU pushes up internal EU prices much higher than the prevailing world market prices. Traditionally, the LDCs have had some success in exploiting the higher internal EU prices, thereby gaining preferential advantages over low-cost producers in the rest of the world. Take sugar, one of the most important products in the EBA initiative, as an example. The EU sugar regime18 is among the most controversial instruments of the CAP. With this regime, the EU has become a major exporter of sugar, whereas some LDCs are exempted from some import restrictions through the ACP trade preferences. The ongoing WTO agricultural negotiation and the imminent enlargement of the EU exert serious pressure on the EU to reform the CAP. On the one hand, many WTO members exert pressure on the EU to eliminate these distortions (see FOI, 2003, for a summary of negotiation positions of key WTO members). On the other hand, extending the same support as enjoyed by the existing EU members to the newly admitted ones would put serious pressure on the EU budget. Against this backdrop, it is conceivable that the EU will – at least partially – reform the CAP, thereby effectively reducing the preference margins available to the LDCs. For this reason, we conducted further simulations to investigate how the preferences as granted by the EBA would diminish under a series of ‘plausible’ EU reform scenarios. These scenarios contain conjectured EU

296

W. Yu and T.V. Jensen

reforms during the next round of WTO trade liberalization. The combined results of these scenarios will demonstrate the costs associated with keeping the preferences meaningful by maintaining the protection measures of the EU. Two sets of results are provided for each of the four scenarios (Scenarios 2–5). The first of these concerns changes in imports and outputs in the EU (reported in Table 13.5), whereas the second reports welfare changes obtained from these experiments (Table 13.6). When presenting the results, we discuss mainly the combined results from the EU reforms and the EBA, whereby illustrating the ‘erosion’ of the EBA preferences. Where needed, we also draw out the qualitative results on the separate impacts of the EU reforms alone, to explain how the presence of these reforms may reduce/erode the gains from the EBA.

Impacts of market access reform in the EU (Scenario 2) Reducing tariff rates on imports from all trading partners except the LDCs by 50 per cent will reduce the domestic prices of imports in the EU. This impact is far greater than that of the reforms under the EBA, due to the dominant share of imports from non-LDC regions in total EU imports. This leads to an increase in total imports into the EU. However, preferential margins available to the LDCs are reduced significantly and exports from these countries face more competition from the non-LDC countries. The competition is fiercest for vegetables/fruit, other food, plant fibres, other crops, other animal products, beef and vegetable oil. Due to the fact that the LDCs have already enjoyed tariff-free access to the EU markets for many of these products and that the non-LDC countries receive quite significant concessions from the EU’s market access reform, the LDCs lose some of their export market shares for these products, thereby reducing the benefits from the EBA. Table 13.5 shows the combined effects of the market access reform and the EBA (results of Scenario 2). For many products, exports from the LDCs to the EU actually decrease. For example, sub-Saharan Africa’s exports of vegetables, fruit, and other crops decrease respectively by 2 and 4 per cent, indicating a less favourable situation as compared to the pre-EBA case. For other agricultural products such as wheat, cereal grains, dairy and sugar, even though the preference margins offered by the EBA remain positive, other major exports to the EU are able to increase as well. As a result, exports from the LDCs increase at much lower magnitudes than in the EBA scenario. To see this, one can compare the percentage changes in the EBA (Scenario 1, shown in Table 13.2) and Scenario 2 (Table 13.5). In the case of sugar, exports originating from the SSA region increase by less than 600 per cent in Scenario 2, as compared to a nearly 900 per cent increase in Scenario 1. In terms of welfare, all LDCs are worse off from the EU’s market access reform, either due to losses in allocation efficiency or deteriorations in terms of trade, or both. These results are shown in Table 13.6. Take subSaharan Africa as an example. The total welfare change in this region is a

Wheat

Cereal grains

Veg. fruit

Market Access scenario (Scenario 2) Exports from LDCs Mwi 171 185
11 Moz 200 215
2 Tan 178 186
11 Uga 203 219
1 Zam 197 177
13 SSA 200 214
2 EU total Import 4.1 0.8 1.3 Output
7.6
5.5
2.8 Export Subsidy scenario (Scenario 3) Exports from LDCs Mwi 231 225
6 Moz 266 261 4 Tan 245 227
5 Uga 284 276 7 Zam 289 216
8 SSA 273 268 5 EU total Import
1.2
1.5
0.4 Output
3.3
7.1 0.4 Export
7.4
17.7 0.6

Oilseed


8
1
10 1
17
1
2.4 1.2


12
3
14
2
23
4
0.6 0.9 2.4

Plant fibre


8
1
7 0
11
1
2.9 0


9
2
8 0
15
2
0.2 0.9 2

Other crops


12
4
12
3
17
4
1.4 0.2


12
6
13
2
20
4
0.9 0.7 1.9

Other animal prod.

Bovine prod.


7 5
23 6
12 6


9 3
4 49
8 22

21 37 28 98 22 62

25.1
6.5

3.6
1

24 42 34 113 28 73

24 41 34 112 28 73

0.6
2.8


3 11
19 16
8 15
1.1
0.4 1.4


0.5
1.5
10.9

Other meats


0.2
0.4
2.7

Veg. oil


18
8
14
7
18
8 2.1
1.5


15
3
10
1
15
2
0.5
0.1 0.2

Dairy

Rice

Sugar

369 427 379 431 387 424

641 734 652 732 669 730

8.2
2.6

20.4
18.3

412 477 420 497 443 476

1122 1278 1142 1337 1198 1311

662 832 1521 1111 507 862


0.7
2.7
13.3

2.7
6.5
14.8

7.7
11.4
49.4

461 564 1065 756 368 582

The Everything But Arms Initiative

Table 13.5. Changes in import and output in the EU under various reform scenarios, percentage from base data.

36.3
8.7

Continued 297

298

Table 13.5. Continued. Changes in import and output in the EU under various reform scenarios, percentage from base data.

Wheat

Cereal grains

Veg. fruit

Domestic Support scenario (Scenario 4) Exports from LDCs Mwi 413 418
11 Moz 454 462
4 Tan 420 404
12 Uga 499 509 3 Zam 474 383
16 SSA 466 476
2 EU total Import 3.9 3.4
1.1 Output
8.1
8.4 3.5


16
10
18
7
20
11
0.3 0.8

12 19 7 29
5 22 9.1
16.2

17 21 12 30 4 23 5.5
14.2

Plant fibre


8
3
9 2
16
2
0.9 1.9


6
1
7 2
12
1
3.6 2.1

Other crops


16
11
18
4
26
9
2.4 4.2


17
11
18
7
23
11
4 4.2

Other animal prod.

Bovine prod.

Other meats

20 34
3 46 9 40

36 54 44 136 36 87

39 56 46 141 38 90

2.4
1.7

1
0.8

0.3
0.7


2 9 2 62
3 28

34 48 38 121 31 74

26.5
9

3.7
2.3

12 22
11 28 1 23 2.1
5.2

Veg. oil


7 4
3 10
10 6 1.2
1.2


10
3
9 1
13
3 2.9
2.8

Dairy

429 496 435 535 446 502

Rice

1147 1292 1144 1377 1187 1324

Sugar

697 866 1566 1203 518 902


0.1 0

2.9
2.9

374 418 368 446 380 407

660 737 654 749 670 733

461 553 1042 768 360 565

7.1
5.4

20
21.7

32.3
17.3

11
2.4

Source: simulation results. Note: Mwi, Moz, Tan, Uga, Zam and SSA refer to Malawi, Mozambique, Uganda, Zambia and rest of Sub-Sahara Africa, respectively.

W. Yu and T.V. Jensen

Combined scenario (Scenario 5) Exports from LDCs Mwi 274 300 Moz 297 330 Tan 270 287 Uga 318 353 Zam 291 277 SSA 295 327 EU total Import 7.8 2.9 Output
19.1
19.5

Oil seed

Market Access scenario (Scenario 2) Region

Efficiency

TOT

Total

Aus. and New Zealand
51.6 356.7 293.4 Japan 155.3 1,131.2 801.5 ASEAN 179.1 708.4 874.6 SAARC 95.2 262.7 360.2 Eastern Asia 480.7 1,359.3 1,674.2 USA 5.7 1,134.1 1,372 Canada 16.7 133.9 99.8 Latin America and 217.3 645 922.9 Caribbean EU 9,103.3
9,490.6
69 Eastern Europe 511.5 1,940.5 2,556.4 and FSU Middle East 318 930.2 1,269.2 Malawi 5.6 12.7 16.9 Mozambique
1.1
2.7
4.3 Tanzania 8.9 18.5 34.9 Zambia
9.8 13.9 3.2 Uganda
0.4
2.3
3.6 Sub-Saharan
14.6
3.8
24.9 Africa Rest of Africa 255 564.2 805.7 World 10,794.7
46.1 10,750

Export Subsidy scenario (Scenario 3) Efficiency

TOT

Domestic Support scenario (Scenario 4)

Total

Efficiency


30 136.3
269.9
270
21
87.2
26.6
0.9
66.9
130.3
214.4
40.5
76.8 6.4
52
22.1

104.7
527.8
115.9
30.7
207.5
325.9
68.9
79.6


34.9
40.6 3.4 6.4 12.1
228.9
44
246.5

1,639.1 1,934.6
386.8
503.8

3,629.1
896.9

6,792.4 2,910.5
139.8
362.3

9,694.1 16,952.8
4,267.3 12,995.4
512.2
36.3 1,046.3 1,093.2


608
567.9
1,178.9 9.2 21.9 29.3 1.2 3.5 8.1 14 30.5 56.4
13.4 25 10.5
0.1 0.5 0.9
44
48.6
71.1


192.7
347.1 7.7 17.6 1.2 2.4 13.8 29.8
12.6 25.1
0.7
6.9
2.6 22.7


549.6 23.8 5.2 54.7 11.4
9.7 27.5


152.1
281.6
695.9
0.1


75
220.3 5,820.2
2.4


292.5 26.8 5,817.7 14,849.3


434.7
994.7

TOT

Total

Combined scenario (Scenario 5)


48.8
81.6
746.8
568.5
27.6
25.2
110.1
107.5
187
139.8
554.7
1,002.7 82.3 63.5
246.2
548.5

Efficiency

TOT

Total


118.3
165.1 159.4 79.6 422.1
453.6
109.2
85.9

456.3 154.1 594.3 164.7 1,028.8 576 236 408.4

328.7
230.2 741 242.9 1,336.2 95.7 106.4 332.9


506.1 3.8
1.3 6.5
8.8
1.6
119.3


22.8 7.7
4.5 15 12.1
10.4
257.1

The Everything But Arms Initiative

Table 13.6. Welfare results under various EU reform scenarios (US$ m., percentage from base data).


522.5 10.5
7.9 27.5 2.5
15.5
396.5

67.7 83.3
24.3 14,826

Source: simulation results. Note: the first three columns of each scenario show the allocation efficiency effects, terms of trade effects and total welfare effects, respectively. Please note that numbers in first two columns do not add up to the corresponding number in the third. 299

300

W. Yu and T.V. Jensen

loss of US$25 m., down from a gain of US$169 m. in the EBA Scenario. For the other LDC countries, welfare changes are either negative or marginally positive, as compared to the universal gains from the EBA scenario. To understand these welfare results, we focus on the negative export price effects, which dominate the total terms-of-trade effects. These negative export price effects occur for two reasons. On the one hand, MFN tariff cuts by the EU lead to a substitution away from imports originating from the LDCs and relatively smaller increases in imports from the LDCs. This point can be seen from the smaller increases shown in Table 13.5, as compared to the corresponding changes in Table 13.2. On the other hand, the preferential access granted to the LDCs possibly ‘traps’ the exports from the LDCs and prevents them from shifting to other markets, thereby dampening the prices of the LDCs’ exports. This is supported by the simulation results (not shown here), that increases in the LDCs’ exports into other markets are quite limited. For the EU, the effects of the market access reform alone far exceed that of the EBA initiative. Thus, the EU experiences quite significant import increases in those sectors that are currently protected by high MFN tariff rates. For example, total imports for beef, sugar and rice all increase by over 20 per cent (shown in Table 13.5). Accordingly, outputs of these products decrease in the EU. Table 13.5 shows that outputs for rice and sugar decrease by 18.3 and 8.7 per cent, respectively. In terms of welfare, allocation efficiency gains for the EU are completely cancelled out by terms of trade losses, resulting in negative but marginal total welfare changes for the EU: a loss of US$69 m. dollars, which is still an improved scenario when compared to the effects of the EBA alone (a loss of US$184 m.). The effects on non-LDC countries are positive and many times greater than the potential losses to the LDCs. In fact, the global welfare gain is estimated to be more than US$10 bn. This number clearly suggests an overwhelming benefit of the EU’s market access reform to the world, despite the possible negative impact on the LDCs by eroding the trade preferences. Impacts of eliminating export subsidies in the EU (Scenario 3) Eliminating export subsidies in the EU (in Scenario 3) leads to an increase in the EU’s export prices, especially for products with significant subsidies, namely, wheat, cereal grain, rice, beef, other meats, dairy and sugar. This in turn reduces EU exports of these products to the world market and at the same time pushes up the corresponding world market prices, as the EU is a ‘large’ trader in the world agricultural market. Accordingly, outputs of these products in the EU are reduced. In contrast, for products with few or no subsidies, this reform is actually favourable because it expands their outputs by absorbing resources that move out of the previously subsidized sectors. Consequently, exports of these products actually increase. As the impact of the EBA is much smaller than that of the EU’s MFN reforms, the above-mentioned effects generally dominate the combined results of

The Everything But Arms Initiative

301

Scenario 3. Table 13.5 illustrates the opposite movements of outputs of different products in the EU. For example, outputs for oilseed and plant fibres increase by just below 1 per cent, while sugar production decreases by more than 10 per cent. Overall, total exports and outputs in the EU decrease, which pushes up world market prices. One interesting but puzzling result from Scenario 3 (shown in Table 13.5) is that the percentage output decline of many products in the EU is far below that of the corresponding exports from the EU. This implies that a significant portion of the outputs must be sold within the EU, possibly due to substitution of imported products with domestically produced ones. Due to the Armington specification used in the GTAP model, which defines imported and domestic products as imperfect substitutes, this is indeed the logical result of the model. When world market prices rise because of fewer exports from the EU, the ‘home bias’ structure results in fewer imports into the EU, thereby generating this seemingly paradoxical result, i.e. trading partners of the EU do not gain directly from the EU’s removal of export subsidies. From Table 13.5, we can see that except for sugar and rice, total imports of all other products into the EU decline by up to 1.5 per cent.19 Although their exports to the EU are adversely impacted by the removal of the EU’s export subsidies, net food-exporting countries are generally able to redirect their exports to markets other than the EU and in some cases increase exports to these markets. However, non-LDC net foodimporting countries generally lose in terms of economic welfare due to higher world market prices. The EU itself enjoys positive welfare gains of nearly US$3.6 bn (shown in Table 13.6) as a result of removing its export subsidies, due to gains from both allocation efficiency and terms of trade. As expected, the negative impact of the EBA on the EU appears to be quite insignificant, in comparison to the gains from removing export subsidies. The impact on the LDCs is quite different from that on the EU. Higher import prices for agricultural products lead to terms of trade losses for the LDCs. Again, take sub-Saharan African as an example. Table 13.6 shows that the gains from the EBA are not enough to offset the negative welfare effects caused by the removal of the EU’s export subsidies. In fact, the combined effects of the EBA and the removal of the EU’s export subsidies (i.e. Scenario 3) are negative for the SSA region. Due to their trade patterns with the EU and the world, the other LDC countries are affected marginally by this reform. For these countries, the combined welfare effects under Scenario 3 are positive but slightly smaller than that of the EBA alone. To summarize, eliminating export subsidies benefits the EU the most due to budget savings and efficiency gains. However, this move does not give rise to welfare gains for the rest of the world. Food-importing countries no doubt have to face high import prices. For LDCs, there are no obvious benefits as most of them fall into the category of foodimporting countries. In addition, fewer exports from the EU imply more fierce competition in the EU market, which does not help the LDCs’ exports, either.

302

W. Yu and T.V. Jensen

Impact of domestic support reforms in the EU (Scenario 4) Reducing different types of domestic support measures has different effects on producers in the EU. For example, reducing output subsidies for a given product generally raises its price, which in turn puts downward pressure on its outputs. A reduction of intermediate input subsidies tends to cause significant substitution among different inputs and in general also impacts its final outputs. Land- and capital-based payments are among the most important policy instruments within the framework of the CAP. Reductions of these payments will push down the rental prices of land and capital and cause significant reallocations of these factors. The exact effects of this reform depend on both the budget outlays of the existing support measures and their distributions across different instruments and products. These impacts on the EU will have repercussions on the world market through changes in trade pattern and terms of trade. The above considerations are confirmed by the results from Scenario 4. On the input side, it is found that resources generally move out of agriculture, and that prices of the factors that are specific to agriculture, such as land, drop significantly. On the output side, the net effects of reducing these domestic support measures vary across products. Prices of wheat, grain, oilseed and meats and animal products increase significantly, whereas prices of vegetables and fruit, plant fibres and other crops decrease. Prices of rice, sugar and dairy also decrease, but to a lesser degree. Outputs of the first product group (see Table 13.5) decrease, whereas those for the second group increase. The most notable output decreases are found in oilseed (over 16 per cent, according to Table 13.5), followed by wheat and grain (over 8 per cent), whereas outputs of vegetables and fruit and other crops increase by 3.5 and 4.2 per cent, respectively. World market prices change accordingly, although not as significantly as with domestic prices in the EU, due to the fact that other major exporters adjust their production and trade so as to absorb some of the shocks to the world market prices. For the LDCs, as the world market prices of some key exporting goods (such as vegetables and fruit and other crops) drop, the gains from the EBA diminish. This is especially the case for rest of sub-Saharan Africa, which is by far the largest exporter among the six LDCs examined in this study. Terms of trade gains from the EBA are eroded considerably, particularly due to the unfavourable price effects in wheat, grain and other crops. In monetary terms, domestic support reform in the EU will lower the welfare gains of sub-Saharan Africa to only US$27 m., even with the benefits from the EBA considered. For most other LDCs, the negative impacts of the domestic support reform seem to be limited and are not enough to offset the gains from the EBA. Thus, the welfare results for these countries generally remain positive, even though they are quite small. The EU is understandably gaining significantly from the domestic support reform through budgetary savings and by using resources more efficiently. Its terms of trade also improve, thereby adding to a total welfare gain of nearly US$10 bn – even with the losses resulted from implementing

The Everything But Arms Initiative

303

the EBA being considered (Table 13.6). Other regions are affected negatively, mainly due to terms of trade losses caused by higher world prices. Overall, the world would enjoy a welfare gain of US$5.78 bn from the EU’s domestic reform alone, which far outstrips the welfare gains from the EBA alone. When the effects of the EBA are included (as in Scenario 4), the welfare gain worldwide is about US$5.82 bn. These results support the hypothesis that domestic reform by the EU will reduce high EU prices in some key commodities that are particulary relevant to the LDCs. Consequently this reduces the terms of trade gains from the EBA for the LDCs. Still, these losses to the LDCs appear to be quite small compared to the overall gains achieved for the world. The EU stands to gain significantly from such a reform and is in the position to provide assistance to the LDCs in order to offset the negative impacts on them.

Welfare costs associated with keeping the EBA meaningful for the LDCs (Scenario 5) The above results demonstrate that the benefits of the EBA to the LDCs would diminish, or even totally disappear, should the EU carry out a series of ‘plausible’ reforms on a MFN basis as part of the new round of multilateral trade liberalization. As a summary of the three individual experiments (Scenarios 2–4), the last panel of Table 13.6 reports the welfare results of implementing the three reforms simultaneously, together with implementing the EBA. It is shown that simultaneously conducting these reforms (Scenario 5) leads to welfare losses for all the LDCs concerned in the study, especially for the rest of the sub-Saharan Africa region, which will suffer a loss of over half a billion US dollars,20 thereby turning the gains obtained from the EBA scenario (US$169 m., see Table 13.3) to a loss of nearly US$400 m. Therefore, our results provide some justification for the concerns about preference erosion and the fear of marginalizing the interests of the LDCs in the current negotiations. As the EBA initiative aims at eliminating tariff and quantitative restrictions imposed on LDCs exports, there will be nothing more for the EU to offer those countries in terms of preferential treatment. Certainly, maintaining the existing protection measures in the EU unchanged would be an option for keeping these preferences meaningful. However, this option entails an enormous cost as it distorts world trade and harms the economic welfare of the whole world. Our results (Table 13.6) show that the global welfare gains from the possible EU reform scenarios (the three reforms combined) amount to nearly US$15 bn. In fact, most regions covered in this study gain from such a move by the EU, with the EU itself being the biggest beneficiary with a gain of nearly US$13 bn. The LDCs, the Middle East and North African region and Japan are the only losers under such a scenario. As discussed in details in the previous subsections, about two-third of the total welfare gains come from the market access reform by the EU, whereas the domestic support reform provides much of the remainder.

304

W. Yu and T.V. Jensen

Removing export subsidies does not seem to generate significant gains for the rest of the world. However it does add to the welfare gains of the EU. Since these gains cannot be achieved if the EU wants to keep the preferences meaningful to the LDCs, the world would have to forgo these welfare gains (US$15 bn) for the sake of keeping the preferences meaningful. Our results have shown that the additional benefits offered through the EBA will amount only to a small fraction of the abovementioned gains. For the EU to forgo the benefits at such a scale for itself and for the world at large would be difficult to justify from a purely economic perspective. While the study shows that the global welfare effects from the EU’s trade liberalization initiatives far exceed the potential damage to the LDCs, the question that remains unanswered from the LDCs’ perspective is how to protect the interests of these poor countries when the rest of the world stands to gain further from the new round of trade liberalization. The notion that just because the losses of the LDCs are small they can simply be neglected and forgotten is certainly hard to accept. On the contrary, this issue has received ample attention during the Doha meeting. In fact, as agreed by the WTO members in Doha, the interest of the LDCs within the overall objective of multilateral trade liberalization has been explicitly declared as a focus. The difficulty raised in this chapter is clearly that multilateral liberalization might diminish the benefits of special and differential treatment, while the latter could slow down the progress of the former. Facing this dilemma, one has to realize that trade policy is not the only policy option for assisting economic development in the LDCs. In many cases, a set of policies should be implemented simultaneously. For example, the EU, with all the potential welfare gains from reforming its protection measures, can team up with other developed countries to compensate the LDCs for any possible losses. This compensation can be offered as, among other forms, development aid that targets various supply-side constraints that prevent the LDCs from effectively competing in the world market. For example, in the case of the sub-Saharan African region, our analysis shows that there is potential for these countries to export sugar, vegetables and fruit, other crops, rice and plant fibres. Therefore, the EU and other developed countries can target their development aid to improve these countries’ infrastructure and boost their productivity so that the real comparative advantages of these countries can be exploited in the world market.

Concluding Remarks This only shows that the EBA initiative affects only a limited number of products, mainly food and agricultural goods (including sensitive products such as sugar, rice and bananas), due to the preferences that have already been granted to the LDCs prior to the EBA. In addition, current exports of

The Everything But Arms Initiative

305

the EBA-affected products from the LDCs to the EU are quite limited. Therefore, we expect only limited positive effects for the EBA products in the concerned LDC countries. Given the limited effective coverage of the agreement due to the preferences granted in the past, it is unlikely that this initiative will generate sizeable gains for the LDCs at the macro level. The quantitative analysis conducted in this study shows that total welfare gains from such an initiative are less than US$300 m. for all the LDCs concerned in the study. Furthermore, a great deal of the gains will come from the three sensitive products that are subject to a gradual, lengthy liberalization process, especially sugar. Lastly, the negative impact on the EU and other countries seems to be quite small, indicating that granting quota- and duty-free access to the LDCs does not cost the EU very much and does not pose a big threat to exports from nonLDC countries. It should be further noted that the quantitative assessment of the benefits of the EBA is based on the assumption that the safeguard provisions and the rules of origin attached to the EBA initiative would not play the role of restricting exports from the LDCs. This is none the less a strong assumption, especially considering the results for the trade and production of sugar. If, indeed, the LDCs are able to take full advantage of this initiative and increase their exports of sugar substantially, the safeguard clause may well be triggered. Therefore, even the modest benefits of the EBA as shown in this study should be treated with caution. Multilateral trade negotiations under the auspices of the WTO exert pressure on the EU to reform its agricultural policy. It is the protection measures contained in the CAP that support the high EU domestic market prices and keep the preferences attractive to the LDCs. Further analysis in this study demonstrates that the preference margin – as well as gains from these preferences – will be diminished under a series of ‘plausible’ reform scenarios in which the EU cuts its market access barriers, export subsidies and domestic support measures. In fact, most LDCs will suffer welfare losses under such scenarios. These results suggest that further trade liberalization may actually harm the LDCs. However, using this result as an argument for maintaining the EU’s (and other countries’) protections and for opposing the WTO route toward trade liberalization would cause much bigger potential welfare losses to the whole world and ultimately would harm the LDCs in the long run. Indeed, the reforms carried out by the EU are shown to generate sizeable gains for the world as a whole, which far exceed the potential losses to the LDCs. A coordinated effort by all WTO members in liberalizing the world trading system will surely generate much larger welfare gains. Based on the quantitative analysis, two sets of policy implications can be drawn out from the perspective of the LDCs. Regarding the EBA initiative, it is important that the trading opportunities provided through the EBA are truly made available to the LDCs. To do so, the EU should minimize the uncertainties and complexities associated with the safeguard measures and rules of origin. The presence of these measures in previous preference schemes (such as the GSP) may have contributed to their poor performance. To avoid

306

W. Yu and T.V. Jensen

this problem reappearing, the LDCs should seriously consider taking action to negotiate for unconditional quota- and duty-free access to the markets of the EU and other developed countries. As many WTO members have proposed the abolition of Article 5 (special safeguard provisions) of the Agreement on Agriculture, the LDCs may take this opportunity to integrate the EBA into the new WTO agreement and to get rid of the attached safeguard provisions altogether. It should be noted, however, that these benefits are bound to be short-lived, and that long-term solutions should be considered. Concerning the possible negative impact of the WTO reforms, the negotiation proposal of the African Group emphasizes the importance of implementing the Marrakesh Ministerial Decision. Although this decision was meant primarily to tackle the problems resulting from the Uruguay Round reform, many of the mechanisms proposed can also help to ease the problems discussed here. For example, technical and financial assistance and cooperation are much needed in solving many supply-side constraints and bottlenecks that the LDCs face. As the LDCs have to compete on level grounds with other WTO members eventually, the key to improving their position is to break these constraints. Certainly, the EU and other developed countries are in the position to offer such assistance as they stand to benefit the most from further multilateral liberalization. If the efforts from the rich countries are indeed forthcoming, and if the cooperation between the rich and the poor is genuine during the Doha Round, there is no reason to fear marginalization of the LDCs. If not, those countries may not have the willingness to participate in the negotiations.

Notes * This

chapter is based on Yu and Jensen (2005). For example, Bora et al. (2002) have shown that the EBA would result in welfare gains of around US$400 m., which is similar to a slightly earlier study (Ianchovichina et al., 2000), while in another study (Trueblood and Somwaru, 2002) the estimated benefits are several times higher. 2 The study by Trueblood and Somwaru (2002), for example, lacks an accurate account of the existing preferences that the LDCs have already enjoyed. Consequently, it overstates the likely impact of the EBA. 3 See Panagariya (2002) for a detailed discussion. 4 Although the GSP applies also to developing countries that are not LDCs. 5 Prior to 2002, Senegal was included in the ACP preferences as a non-LDC country. Since then, it has been included in the UN’s list of LDCs. 6 There is considerable overlapping between country coverage of the ACP and the GSP. In fact there are only nine LDC countries that are included in the GSP but not in the ACP. These countries are mostly South Asian countries, including Bangladesh, Bhutan, Cambodia, the Laos, Maldives, Myanmar, Nepal and Yemen. In general, the ACP preferences are more favourable than the GSP, so it is natural for many LDCs to adopt the ACP preferences when available. 7 Although the EBA was adopted as an extension of the GSP, prior to the EBA, some LDCs actually enjoyed more favourable terms through the ACP preferences. Further, the dutyfree access granted under the GSP covered fewer products than the ACP preferences, as 1

The Everything But Arms Initiative

307

the latter have been extended to all but 857 products at the HS-8 level, whereas the GSP provided duty-free access to all but 919 HS-8 products. Nevertheless, one has to note that the GSP covered more LDCs. In this study we take the view that the EBA is an extension of the preferences that were actually enjoyed by the LDCs at the time of its inception. In this sense, it is worth reviewing both the GSP and the ACP preferences. 8 The GTAP database is a global database containing detailed input–output tables at country or regional level, and bilateral trade flows. By providing a ‘snapshot’ of the world economy at a certain point (the current version has a base year of 1997), it aims to support policy analysis of global trade, agriculture and environmental issues. It has been used extensively by applied researchers. The current version contains information for 66 regions/countries and 57 industries. For detailed documentation of the database, see Dimaranan and McDougall (2002). 9 Panagariya (2002) has pointed out that precisely due to the many restrictions and conditions these preferences have been rendered ineffective. A recent World Bank study (Brenton, 2003) provided an initial evaluation of the impact of the EBA for the year 2001 and pointed out that the low level of utilization of the EBA preferences by non-ACP LDCs might be due to the rules of origin. 10 Bangladesh is recognized as a LDC. It is also listed as an individual region in the GTAP database. Due to our focus on African LDCs, we aggregate Bangladesh into the South Asian Associations of Regional Cooperation (SAARC). 11 The aggregated SSA region contains 43 individual countries, 33 of which are LDCs and the rest are non-LDCs. The GTAP version 5 database does not provide further breakdown of this region. Therefore, we are forced to treat this as a LDC region. The country coverage and the preferences offered in the EBA are therefore extended to the non-LDC countries in SSA in our quantitative analysis. Consequently, the results regarding the impact of the EBA may be slightly exaggerated. However, trade data for 1999 from the TRAINS database indicate that exports of major EBA products from the LDCs in SSA dominate exports from all SSA countries. For instance, only about 15 per cent of sugar exports from SSA were from non-LDCs. 12 Bananas are part of the category of vegetables and fruit in the GTAP database. 13 This involves using a simulation routine called ‘ALTERTAX’ (Malcolm, 1998) to exogenously shock the MFN tariff rates in the GTAP database to the actual preferential rates, while keeping the trade data as close to those in the original database as possible. Note that these pre-EBA tariff rates are taken from the UNCTAD study. After independent replication, we have reached similar estimates using the TRAINS database. 14 These scenarios are formulated based on a recently released policy paper (FOI, 2003) on the proposed negotiation ‘modalities’ by the Chair of the Agriculture Committee of the WTO. 15 We choose a simplistic approach to implement the reduction of domestic support in the EU. As the GTAP database breaks down total domestic support into output subsidies, intermediate input subsidies and land and capital based subsidies, we shock all these instruments by 35 per cent to achieve the overall reduction target. As different instruments have different effects, other approaches may achieve quantitatively different results. 16 This is mainly due to the presence of an output tax on manufacturing goods in Zambia in the GTAP database version 5. Reduced manufacturing output leads to a loss of tax revenue. 17 The CAP consists of export subsidies, high import tariffs and quantitative restrictions (import quotas specified in Tariff Rate Quotas – TRQs for various products and production quotas), as well as domestic support measures that are directly linked to agricultural outputs and inputs and factors used in agricultural sectors (see Chapter 2, this volume).

308

W. Yu and T.V. Jensen 18

The EU sugar regime features restrictions on imports through high import tariffs, domestic intervention price and production quotas within the EU, and preferential access to the EU market through Tariff Rate Quotas to the ACP countries. This policy has resulted in an internal EU sugar price that is about 100–200 per cent higher than the world market price (Frandsen et al., 2003). 19 It is worth noting that in Scenario 3 tariff barriers are maintained for the EU, whereby the EU producers are protected from stronger competition. Removing the tariffs would no doubt lower the import prices in the EU market and would lead to more imports. 20 This estimate is obtained by conducting a separate experiment where only the MFN reforms by the EU are simulated. Adding the welfare results from the EBA scenario (Scenario 1) to the above estimate will lead to the welfare estimate for Scenario 5. Welfare results for implementing the various EU reform scenarios with and without the EBA can be obtained from the authors.

References Armington, P.A. (1969) A theory of demand for products distinguished by place of production. IMF Staff Papers 16, 159–178. Bora, B., Cernat, L. and Turrini, A. (2002) Duty and Quota-free Access for LDCs: Further Evidence From CGE Modeling. Policy issues in international trade and commodities study series No. 13, UNCTAD, United Nations, New York and Geneva. Brenton, P. (2003) Integrating the Least Developed Countries into the World Trading System: The Current Impact of EU Preferences under Everything but Arms. World Bank Working Papers No. 3018, Washington DC. Dimaranan, B.V. and McDougall, R.A. (2002) Global Trade, Assistance, and Production: The GTAP 5 Database. Center for Global Trade Analysis, Purdue University, West Lafayette, Indiana. European Community (2001a) Council regulation (EC) No. 416/2001 of 28 February, Official Journal of the European Communities, Brussels. European Community (2001b) Council regulation (EC) No. 2501/2001 of 10 December, Official Journal of the European Communities. European Commission (2002) The EC’s Proposal for Modalities in the WTO Agriculture Negotiations (europa.eu.int/comm/trade/ issues/sectoral/agri_fish/docs/modalities.pdf).

European Commission (2003a) List of ACP Countries and Regions (europa.eu.int/comm/ development/body/country/country_en.cfm). European Commission (2003b) User’s Guide to the European Union’s Scheme of Generalized Tariff Preferences (europa.eu. int/comm/trade/index_en.htm). FOI (2003) Notes on the Harbinson Draft on Modalities in the WTO Agricultural Negotiations. Danish Research Institute of Food Economics (www.foi.dk). Frandsen, E.F., Jensen, H.G., Yu, W. and Walter-Jørgensen, A. (2003) The EU Sugar Policy: an analysis of price versus quota reductions. European Review of Agricultural Economics 30(1), 1–26. Hertel, T.W. (1997) Global Trade Analysis: Modelling and Application. Cambridge University Press, Cambridge, UK. Ianchovichina, E., Mattoo, A. and Olarreaga, M. (2000) Unrestricted Market Access for Sub-Saharan Africa: How Much is it Worth and Who Pays? Development Research Group, World Bank, Washington DC. Malcolm, G. (1998) Adjusting Tax Rates in the GTAP Database. GTAP Technical Paper No. 12, Center for Global Trade Analysis, Purdue University, West Lafayette, Indiana. Panagariya, A. (2002) EU preferential trade arrangements and developing countries. The World Economy 25(10), 1415–1432.

The Everything But Arms Initiative Trueblood, M. and Somwaru, A. (2002) Trade Liberalization and the Least Developed Countries: Modeling the EU’s Everything But Arms Initiative and a Hypothetical US Response. 5th Annual Conference on Global Economic Analysis, 12–14 June Taipei, Taiwan. WTO (2001a) Ministerial declaration (document WT/MIN(01)/DEC/1). Geneva. WTO (2001b) Committee on Agriculture – Special Session – WTO African Group: Joint Proposal on the Negotiations on

309 Agriculture (document G/AG/NG/W/142). Geneva. WTO (2003) Committee on Agriculture – Special Session – negotiations on agriculture first draft of modalities for the further commitments (document TN/AG/W/1). Geneva. Yu, W. and Jensen, T.R. (2005) Tariff preferences, WTO negotiations and the LDCs: the ‘Everything but Arms’ initiative of the EU. The World Economy 28(3), 375–405.

14

New Regionalism in the Aftermath of Cancún: to the Benefit or Detriment of Developing Countries?

CHANTAL POHL NIELSEN Danish Research Institute of Food Economics, Agricultural Policy Research Division, Rolighedsvej 25, 1958 Frederiksberg C, Denmark

Introduction When multilateral trade talks collapsed at the Fifth WTO Ministerial Meeting in Cancún, Mexico, in September 2003, negotiations of new regional and bilateral trade deals appeared as an alternative to the WTO Doha Round. For instance, in the immediate aftermath of Cancún, WTO members witnessed the negotiation of a bilateral trade agreement between the host Mexico and Japan. Bilateral trade agreements were also in focus on the sidelines of the October 2003 Asia–Pacific Economic Co-operation (APEC) Summit. The host of this meeting, Thailand, was poised to launch bilateral trade negotiations with the USA, Canada, New Zealand and Hong Kong, after having successfully completed a bilateral trade, investment and services agreement with Australia (Bridges, 2003). The USA indicated its intention to move ahead with different regional and bilateral trade agreements, including partners as different as Australia, Morocco and Central America countries, while the EU also continued with the multiple regional negotiations from Eastern Europe to Least Developed Countries. The renewed emphasis on bilateral and regional trade agreements after the failure in Cancún represented, in fact, the continuation of a trend. Since the early 1990s there has been a dramatic increase in the creation of new preferential trade agreements (PTAs) as well as a revival of many existing agreements. Today, virtually all countries are members of a bilateral or regional trade agreement or in the process of negotiating membership, and several are members of more than one agreement. Furthermore, these new regional trade agreements are not merely between neighbours and between ‘equal’ trading partners, but stretch across both continents and different 310

© CAB International 2006. WTO Negotiations and Agricultural Trade Liberalization: the Effects of Developed Countries’ Policies on Developing Countries (eds E. Diaz-Bonilla et al.)

New Regionalism in the Aftermath of Cancún

311

levels of economic development. Developing countries join forces in regional trade agreements of their own, but they are also party to a large number of preferential trade agreements with developed countries. Regionalism has thus become an inherent part of the global trading system and therefore all countries – members and non-members alike – will be affected by this phenomenon. Indeed, using Bhagwati’s (1995) analogy, the global trading system can be characterized as a spaghetti bowl of tariff and non-tariff preferences and other increasingly complex regulations defined in the many overlapping agreements. The agreements differ with respect to product coverage, extent of tariff preferences, time frames for implementation, rules of origin, technical standards, safeguard provisions and customs administration rules, among other things. Moreover, several PTAs go beyond trade policy to include provisions on investment, competition, environment and labour. There are of course many reasons why countries choose to enter bilateral and regional trade agreements. Frustration over unsuccessful liberalization attempts at the multilateral level is one possibility. In general, however, countries create or join regional trade agreements because increased market access is expected to enhance the prospects of economic growth through more competition, utilization of economies of scale, benefits from regional specialization, technical spillovers and possibilities of attracting investment to an expanded regional market. It is also hoped that such agreements will create a more stable and predictable trading environment. Moreover, developing countries’ participation in PTAs seems also to be motivated by the possibility of locking in domestic policy reforms through external commitment. Finally, regional trading blocs may be viewed as being able to exert greater bargaining power in the multilateral WTO negotiations. In light of the ups and downs in the multilateral trade talks that are explicitly aimed at being a Development Round, it is worth considering whether the subsequent renewed emphasis on bilateral and regional trade agreements is to the benefit or detriment of developing countries. This chapter seeks to shed light on the issue by reviewing the most recent theoretical and empirical literature that assesses the economic impact of PTAs1 on developing countries. First, a brief discussion of the theoretically expected impacts of PTAs is presented. This is then followed by an overview of the main results of a large number of empirical assessments of selected PTAs. The chapter concludes by drawing preliminary conclusions about the desirability of bilateral and regional approaches to trade liberalization seen from the perspective of developing countries.

Theoretical Considerations on the Impact of Preferential Trade Agreements2 Old and new trade theory3,4 Trade theorists agree that global free trade is welfare enhancing. Under global free trade, factors of production will be allocated within countries in

312

C.P. Nielsen

a way that achieves structures of production, employment and trade that are consistent with their comparative advantage. Such a reallocation of resources will boost efficiency and thereby increase welfare. Short of global free trade, both reality and theory become more complicated. Preferential trade agreements (even those that are simple, or shallow) introduce an additional complexity because liberalization is taking place in a ‘second best’ world, where some distortions are eliminated (e.g. tariffs within the PTA) while others remain (e.g. tariffs on non-PTA trade and other intra-PTA domestic policies) (Burfisher et al., 2003). The seminal theoretical contributions to this field deal with static welfare analysis of customs unions and date back to the 1950s and 1970s (Viner, 1950; Meade, 1955; Kemp and Wan, 1976). More recently, Krueger (1999) has provided an analysis of the differences between two – theoretically simple – types of PTAs, namely free trade areas (FTAs) and customs unions (CUs). What complicates matters, however, is that many real-world agreements don’t satisfy the textbook definitions of FTAs or CUs. PTAs typically contain selected elements of each of these in addition to various aspects of deep integration. Examples of what deep integration may imply include the removal or disciplined use of non-tariff barriers, as well as provisions governing investment flows, regulatory harmonization, environmental legislation and cross-border labour movements. Economic theory can, nevertheless, provide guidelines for assessing the welfare implications of these intermediate arrangements. The focus of what is becoming known in the literature as ‘old’ trade theory in the tradition of Viner and Meade is primarily on the static concepts of trade creation, trade diversion and terms-of-trade effects. Trade creation occurs when a PTA member increases imports from a lower-cost PTA partner, and its own high-cost domestic production declines. Consumers benefit because they can purchase imported goods at a lower price than the domestic variety. Lower prices increase consumer income and, by increasing demand, this effect may lead to increased imports from both PTA members and nonmember countries. Trade diversion occurs when intra-PTA trade replaces imports from more efficient non-member countries as a result of the PTA’s tariff preferences. Trade diversion is mainly a cost to PTA partners that must pay a higher price for their imports, but it is also costly for outside countries that lose exports, and who may be forced to lower their export prices. When countries involved in a PTA are large enough to affect world market prices, there may also be terms-of-trade effects. A PTA is likely to improve the terms-of-trade for its members and worsen them for nonmembers. Lower demand for non-member imports (because imports from member countries become cheaper due to tariff preference, despite a possible cost advantage of the non-member country) may lead to lower export prices of the non-member country. Furthermore, increased trade within the PTA may lead to a decline in the availability of goods to nonmembers, thereby raising the price for non-members of imports from the PTA (and may force non-members to produce such goods themselves). The deterioration of terms-of-trade of a non-member country is of course the

New Regionalism in the Aftermath of Cancún

313

mirror image of the terms-of-trade improvement experienced by the PTA member. So, even if a PTA member loses tariff revenue in connection with a diversion of trade from non-members to members, these losses may be outweighed by improved terms-of-trade vis-à-vis non-members. The net effect on national, regional and global welfare depends on the relative size of each of these effects, i.e. whether or not trade creation exceeds trade diversion, and what happens to the terms-of-trade. In theory, one can construct cases in which the formation of a PTA is net welfare reducing or enhancing depending on assumptions made about the size of the countries involved in the PTA. See e.g. De Melo et al. (1993), Grossman and Helpman (1995), DeRosa (1998), Panagariya (1999), Burfisher et al. (2003) and Schiff and Winters (2003) for a sample of such varying assumptions. The net effect of trade creation and trade diversion is very difficult to assess empirically, however, because it is difficult to construct the counterfactual of how would trade have looked like without the PTA. These effects, however, are only part of the story. Increased trade is often associated with increased investment opportunities, enhanced productivity, sharpened competition,5 finer specialization and increased utilization of economies of scale. Moreover, trade liberalization takes place in a second-best world characterized by imperfect competition and domestic policy distortions. Hence, ‘new’ trade theory attempts to enhance the understanding of the impact of PTAs on the global economy by incorporating such issues in the analytical models and expanding theory tools to include analyses of rentseeking behaviour, game theory, industrial organization theory (especially imperfect competition) and new growth theory (e.g. understanding the links between international trade and productivity growth). In particular, there is an increasing interest in investigating the trade–productivity link, which seems to be a major source of additional growth and welfare gains. One possible explanation for this link is the so-called ‘investment creation’ effect (Ethier, 1998). PTA involvement may be seen as an integrated part of a development strategy for developing countries, hoping that internal reforms and more open trade regimes will attract FDI from developed countries. ‘Investment diversion’ may also occur, however, if FDI flows are diverted from other developing countries that are potential destinations for these financial resources.6 These positive growth effects may be counteracted, however, by non-transparent rules of origin that result in trade and investment diversion. In principle, both the positive and possibly negative growth impacts of all these effects should be taken into account in addition to the conventional analysis of trade creation, trade diversion and terms-oftrade effects to provide a full assessment of the impact of a PTA.

Regionalism and multilateralism: friends or foes? One of the most contentious aspects of the debate about preferential trade agreements concerns the relationship between regionalism and multilateralism. More specifically, do such agreements tend to reinforce or

314

C.P. Nielsen

derail negotiations at the multilateral level? From the perspective of economic theory, the ‘first best’ approach to global trade liberalization is the multilateral approach. As discussed in the previous section, the welfare implications of regional trade agreements, on the other hand, are ambiguous and depend on the precise context. Hence, economic theory cannot provide clear-cut answers as to whether regionalism reinforces or hinders the move toward global free trade. Table 14.1 provides a list of arguments in the literature of how regional preferential trade agreements may act as building blocks or stumbling blocks7 on the path to global free trade. Naturally, there are elements of truth in both sets of arguments. Harmsen and Leidy (1995) have attempted to identify conditions under which PTAs are most likely to contribute to the overall goal of global free trade and to minimize trade diversion effects. These conditions include: coverage of all sectors; transition periods should be as short as possible; transparent rules of origin; liberal rules of accession; cancellation of the possibility of using anti-dumping laws among members of PTAs; strengthened disciplines on the use of anti-dumping action against nonmembers; MFN liberalization should either precede or accompany new PTAs; in setting the common external tariff, customs unions should adopt the tariff schedule of the least restrictive member in its entirety or – more ambitiously – the lowest MFN tariff among members for each product; and deeper forms of integration are preferred, ceteris paribus, since potential gains from efficient resource allocation within the bloc are maximized. Another relevant issue is the relationship between domestic (traderelated) policies and trade liberalization at the regional or multilateral levels. Most PTAs exclude or limit trade liberalization of sensitive agricultural products. This is because there is an inherent contradiction between trade liberalization on the one hand and trade-related domestic policies that protect ‘sensitive’ sectors on the other. Market forces within a free trade area, for example, will tend to unify prices, thereby rendering efforts of national agricultural policy at maintaining different price levels either ineffective or very costly (Burfisher and Jones, 1998). Precisely for this reason there is a need to understand the impact of protective agricultural policy regimes in the developed countries on their PTA partner countries. Whether initiated unilaterally or forced through multilateral negotiation, policy reforms in countries with protectionist agricultural policies, such as a hypothetical reduction of price support in the EU and the USA, would erode the value of preferential access granted to developing countries through a PTA. A politically contentious issue is whether and how developing countries should be compensated for this effect – introducing a potentially new role for both multilateral and bilateral foreign aid policies. The bottom line is that economic theory cannot provide us with clearcut conclusions about the desirability of regional, preferential trade agreements, and therefore it is ultimately an empirical issue to determine the net impact of a given PTA. Hence, the next section provides an overview of the main results obtained by a large number of empirical studies dealing with PTAs.

Stumbling blocks

PTAs encourage others to the multilateral negotiating table, i.e. the prospect of ‘fortresses’ may help motivate greater efforts to achieve successful multilateral negotiations

Widespread regionalism may lead to a break-up of the world economy into hostile blocs that divert political energies from multilateral initiatives

It may be easier to negotiate multilaterally between fewer and larger PTA-based blocks

PTAs make it more difficult to negotiate at the multilateral level because agreement about positions needs to be achieved within blocks before and during negotiations

Deeper integration within PTAs can help avoid destructive trade wars

PTAs are by definition discriminating, and large PTA blocks may exert market power to improve the terms-of-trade of its members

Expansion of membership based on open membership clauses will eventually lead to global free trade

Closed membership clauses may block additional members in order to preserve trade gains, while open membership clauses seduce members into protectionist regional initiatives and divert political energies from multilateral initiatives.

Adoption of ‘open regionalism’ (i.e. preferences provided on a MFN basis as suggested for APEC) will eventually lead to global free trade

Protectionism of countries not involved in PTAs may increase as regionalism spreads

PTA-induced growth can induce increased demand for extra-PTA imports, thereby benefiting non-members

Use of non-tariff barriers, e.g. anti-dumping and countervailing duty actions, against non-member countries increases as weaker industries struggle to survive regional free trade

PTAs may be able to tackle issues too deep or complex for multilateral negotiations, and may even serve as blueprints for such issues before reaching the global level

Deeper integration of policies and institutions may create or strengthen interest groups that benefit from trade diversion and have incentives to lobby against free trade

Deeper integration of policies and institutions may help lock in complementary market-oriented policies (‘competitive liberalization’, i.e. increasing regionalism creates competition for reform and for membership of PTAs)

Deeper integration may introduce protection in previously unprotected markets through the adoption of common, distorting internal policies

Deeper integration among PTA members (e.g. harmonization of technical standards to international norms) may promote trade both within the PTA and with developing countries.

The spaghetti bowl phenomenon leads to higher transaction costs, reduced and distorted trade flows. For example strict rules of origin may make it too costly to make use of preferences and/or may lead to trade diversion

Membership of a PTA maintains investor confidence

PTAs may lead to investment diversion

315

Building blocks

New Regionalism in the Aftermath of Cancún

Table 14.1. Arguments in the debate on regionalism versus multilateralism.

316

C.P. Nielsen

Empirical Assessments of the Impact of Preferential Trade Agreements In defining the boundaries of the empirical literature review, it has been decided to focus on selected agreements between developed and developing countries8. More specifically, the agreements covered are the following: the Cotonou Agreement between the EU and the African, Caribbean and Pacific (ACP) countries, the EU–South Africa free trade agreement, the EuroMediterranean Association Agreements, the EU–Mercosur agreement, the US Caribbean Basin Economic Recovery Act (CBERA)9, the African Growth and Opportunity Act (AGOA), the Caribbean Basin Trade Partnership Act (CBTPA)10 and, finally, the Generalized System of Preferences (GSP) schemes, including the EU’s Everything But Arms (EBA) initiative for the Least Developed Countries (LDCs).11,12 Table 14.2 provides an overview of the studies dealing with the selected agreements. Most of these agreements are shallow in the sense that they primarily involve the non-reciprocal granting of tariff preferences to developing countries by a developed country, i.e. the GSP schemes, the EBA initiative, the Cotonou, CBERA, AGOA and CBTPA agreements. The other agreements are reciprocal, i.e. the EU–South Africa, the Euro–Mediterranean, and EU–Mercosur agreements, and reflect the most recent tendency, namely that non-reciprocal agreements are moving in the direction of reciprocity. Although currently a non-reciprocal agreement, the trade negotiations under the auspices of, e.g. the Cotonou Agreement aim for a set of free trade agreements between the EU and groups of ACPs. The impact of individual preferential trade agreements can be assessed both ex post and ex ante. Ex post analyses typically attempt to identify the key explanatory factors behind observed changes in trade flows, including the contribution of PTA membership to such changes. Ex ante approaches typically build a more extensive modelling framework that serves as an economic laboratory in which experiments can be conducted to produce some conditional estimates about the expected impact of a PTA on future trade flows, production, consumption and welfare. A number of ex post studies estimate various specifications of a simple econometric model known as the gravity model, to assess whether or not membership of a PTA has had an effect on observed trade flows. Gravity model estimates the flow of trade (exports, imports or the sum of the two) between two countries as a function of supply conditions in the country of origin, demand conditions in the country of destination, and by forces such as PTAs that may either stimulate or restrain the bilateral flow. In addition to gravity models, ex post studies use various other forms of regression analysis as well as calculations of indicators such as the Effective Rate of Protection, export similarity indices and trade concentration indices. The main merits of the ex post approaches taken by the studies included in this survey are their simplicity, their intuitiveness and the fact that they are based on observed data. The main reservation about these types of studies concerns the more or less weak theoretical foundation underpinning the applied models and techniques.13

Agreement / Coverage / Focus

Study

EU–ACP; the Lomé and Cotonou Agreements

Stevens (1990); Raboy et al. (1994); Kerkelä et al. (2000); Wolf (2000); McDonald (2002); Nilsson (2002)

EU–South Africa Free Trade Agreement

Akinkugbe (2000); Andriamananjara and Hillberry (2001); Lewis et al. (2001); McDonald and Walmsley (2003)

Euro-Mediterranean Association Agreements

Brown et al. (1997); Hoekman and Djankov (1997); Konan and Maskus (1997); Rutherford et al. (1997); Dessus and Suwa-Eisenmann (1998); Hoekman and Konan (1998)

EU–Mercosur

Martínez-Zarzoso and Nowak-Lehmann (2000); Martínez-Zarzoso (2001); Teixeira et al. (2002); Harrison et al. (2003)

CBERA

Clark (1997); Coppin (1992a,b)

AGOA

Ianchovichina et al. (2001); Mattoo et al. (2002); Nouve and Staatz (2003); Shapouri and Trueblood (2003)

GSP schemes

MacPhee and Oguledo (1991); Truett and Truett (1993); Oguledo and MacPhee (1994); DeVault (1996); Nilsson (2002); Özden and Reinhardt (2003)

EU’s Everything But Arms Initiative

Ianchovichina et al. (2001); Lewis et al. (2001); Bora et al. (2002); Trueblood and Somwaru (2002); Yu and Jensen (2003)

Hypothetical agreements involving OECD and developing countries, and specific issues related to PTAs

Oguledo and MacPhee (1994); Coe and Hoffmaister (1999); Ianchovichina et al. (2001); Stoeckel and Borrell (2001); Hoekman et al. (2002a,b); Trueblood and Somwaru (2002); Chevassus-Lozza and Gallezot (2003); Freund (2003); Wainio and Gibson (2003)

New Regionalism in the Aftermath of Cancún

Table 14.2. Overview of empirical studies by agreement/coverage/focus.

317

318

C.P. Nielsen

The ex ante studies use either partial equilibrium (PE) or computable general equilibrium (CGE) models. Both types of model can be used to answer ‘what if ’ questions about the impact of a given preferential trade agreement, but they have different theoretical underpinnings. PE models build directly on partial equilibrium theory and focus specifically on those product markets in which the policy changes are taking place. Quantities and prices in all other markets are assumed to be unaffected by the policy change. In contrast, the CGE framework builds on general equilibrium theory, thus focusing on intersectoral linkages, resource constraints and economy-wide welfare analysis. The main merit of the two ex ante approaches described here – PE and CGE – is that they build solidly on classic economic theory. However, it is important to keep in mind that the conclusions drawn from such analyses are ceteris paribus predictions of likely outcomes based on a large range of assumptions. As pointed out by Schiff and Winters (2003), ‘[i]n principle, all the relationships in a model could be estimated from detailed data on the economy over many years. In fact, however, the number and complexity of the relationships far outweigh the data available, and so most CGE models are designed on the basis of theory, intuition and casual empiricism.’ This caveat notwithstanding, both PE and CGE models are useful tools to analyse a large range of issues for which econometric estimation and other ex post techniques cannot be used, such as when analysing the potential implications of PTAs that are yet to be formed. Table 14.3 provides another overview of the empirical studies – this time according to methodological approach. The question is whether it is possible to draw any general conclusions about the impact of PTAs on developing countries from the existing empirical literature. It is, of course, difficult to compare quantitative results of studies that differ not only in terms of which agreements they are investigating, but even for the same agreements there are large differences in methodology, data sets and scenarios. The remainder of this section will nevertheless attempt to provide an overview of the most central results, primarily in qualitative terms, but also by providing illustrative quantitative examples.

Overall impact of PTAs Consistent across both reciprocal and non-reciprocal agreements seems to be the general conclusion that the overall impact of preferential trade arrangements on welfare and trade is positive yet relatively small. To give just a few examples, the PE analysis by Wainio and Gibson (2003), that investigates a whole range of different PTAs in which the USA is involved, shows that even under complete tariff liberalization, US imports from both beneficiaries of non-reciprocal arrangements and FTA-partners would increase by just 3 per cent. Using a CGE framework to analyse the EU–Morocco FTA, Rutherford et al. (1997) report benefits to Morocco of

New Regionalism in the Aftermath of Cancún

319

Table 14.3. Overview of empirical studies by methodological approach. Methodological approach Ex post Gravity models

Other ex post

Ex ante Partial equilibrium General equilibrium Static

Dynamic

Studies Oguledo and MacPhee (1994); Coe and Hoffmaister (1999); Martínez-Zarzoso and Nowak-Lehmann (2000); Martínez-Zarzoso (2001); Nilsson (2002); Nouve and Staatz (2003) Stevens (1990); MacPhee and Oguledo (1991); Coppin (1992a,b); Truett and Truett (1993); Raboy et al. (1994); DeVault (1996); Clark (1997); Hoekman and Djankov (1997); Bora et al. (2002); Chevassus-Lozza and Gallezot (2003); Freund (2003); Özden and Reinhardt (2003) Akinkugbe (2000); Hoekman et al. (2002a,b); Mattoo et al. (2002); Shapouri and Trueblood (2003); Wainio and Gibson (2003) Brown et al. (1997); Konan and Maskus (1997); Rutherford et al. (1997); Hoekman and Konan (1998); Kerkelä et al. (2000); Ianchovichina et al. (2001); Lewis et al. (2001); Stoeckel and Borrell (2001); Bora et al. (2002); McDonald (2002); Teixeira et al. (2002); Trueblood and Somwaru (2002); Harrison et al. (2003); McDonald and Walmsley (2003); Yu and Jensen (2003) Dessus and Suwa-Eisenmann (1998); Wolf (2000); Andriamananjara and Hillberry (2001)

1.5–2.5 per cent of GDP. Another way of appreciating the relative importance of the impact of a PTA is by comparing it to a baseline projection of the future path of the economy (as opposed to the simple comparison with a single baseline year), which provides a useful counterfactual simulation of how the economy would develop if the PTA were not implemented. Andriamananjara and Hillberry’s (2001) study of the EU–South Africa arrangement includes a baseline simulation by which South Africa experiences a 71 per cent growth in GDP over an 18-year period. Seen in this perspective, their estimated 1.4 per cent increase in South African GDP over the same period as a result of its FTA with the EU (and assuming an increase in trade-related manufacturing total factor productivity) is very small (accounting for just 2 per cent of accumulated total growth over the period), and can hardly be called an ‘engine of growth’. The main explanation given in the studies for this relatively small impact of the various PTAs is that the beneficiary countries already enjoy low tariffs, typically through existing preferential arrangements such as the Generalized System of Preferences, which applies to most developing countries members of the WTO, although for a limited number of products and under some conditions (e.g. Hoekman et al., 2002b; Shapouri and Trueblood, 2003; Wainio and Gibson, 2003). Also for this reason, the welfare of participating countries may in fact decrease in some cases. Hoekman and Konan (1998), for example, find that the result of the EU–Egypt FTA is dominated by trade diversion effects resulting in a

320

C.P. Nielsen

welfare loss of
0.14 per cent to Egypt. The reason is that Egypt already has duty-free access to the EU market for manufactures, and therefore the loss in tariff revenue as a result of a comprehensive FTA outweighs the trade creation effects.

Dynamic gains from trade The gains from trade liberalization may also be small because they are spread over a long period of time. Using a quasi-dynamic framework, Andriamananjara and Hillberry (2001) take account of the phased implementation of the EU–South Africa agreement, i.e. over 12 years. The authors find that this planned delay of gains dilutes the benefits of trade when measured against other sources of growth. If the agreement had been scheduled as a ‘big bang’ in 2000, they find that it would have accounted for 6.8 per cent of total growth over the period 2000–2006 as opposed to less than 2 per cent using the long phase-in and adjustment period. Some analysts conclude that one of the reasons why the empirical studies do not show large overall effects of PTAs is that important new trade theory features (such as imperfect competition, increasing returns to scale, trade–productivity links and other dynamic effects) are not being taken into account adequately. Although primarily illustrative, Brown et al. (1997) inclusion of the dynamic effect of FDI inflows in a static CGE model suggests that even a doubling of annual FDI inflows into Tunisia as a result of an FTA with the EU is unlikely to give a significant boost to Tunisian welfare. Their conclusion is that unless FDI flows become considerably larger than the flow observed to date, FDI is not expected to make a noticeable difference to the economic success of the FTA. Others find more significant results: incorporating export-led externalities in their dynamic model, Dessus and Suwa-Eisenmann (1998) find that the welfare gains accruing to Egypt from its Partnership Agreement with the EU is markedly higher than without these externalities. Their dynamic model results show a welfare gain of just 0.49 per cent of GDP without these externalities compared with 5.24 per cent with these externalities. Hence, in some studies the welfare gains from the inclusion of dynamic effects are noticeably larger, while in others they are not.

Impacts on participating countries A pattern seems to emerge with respect to the impact of non-reciprocal and reciprocal agreements. Non-reciprocal agreements lead to (small) welfare gains for the beneficiary countries, while the preference-giving country either experiences no change or loses slightly in welfare (e.g. Ianchovichina et al., 2001; Bora et al., 2002; Trueblood and Somwaru, 2002; Yu and Jensen, 2003). The studies of reciprocal agreements find that in situations where the

New Regionalism in the Aftermath of Cancún

321

developing countries have been able to maintain high levels of protection prior to the FTA, there are considerable gains to be had for the developed country. Trying to find the ‘optimal’ degree of agri-food liberalization between the EU and South Africa, McDonald and Walmsley (2003) find that if South Africa is not able to induce the EU to liberalize agri-food trade, the welfare benefits of any trade agreement are very small. The EU, on the other hand, will gain substantially from a trade agreement with South Africa even if agri-food trade is excluded from the agreement. Along similar lines, Freund (2003) finds that in North–South free trade agreements, a modified form of reciprocity related to country size seems important. A one per cent increase in preferences offered by the big (small) country leads to a more (less) than one per cent increase in preferences received. This confirms the notion that developing countries have less bargaining power and that because developing countries have not been required to make reciprocal concessions in multilateral negotiations, large countries may be using regional agreements to extract concessions. Using the US negotiations with Central America as an example, Audley (2003) warns that poor countries may suffer under such ‘beggar-thy-neighbour’ deals.

Impacts on developing countries Trade diversion is an issue for non-participants to be concerned about. All the EU–EBA and hypothetical Quad14–EBA analyses, for example, show that there will be trade diversion effects, but that these will be small in magnitude due to the small relative size of LDC exports (Ianchovichina et al., 2001; Bora et al., 2002; Trueblood and Somwaru, 2002; Yu and Jensen, 2003). Bora et al. (2002) compute export similarity indices to identify which non-LDC countries are most likely to suffer market share losses as a consequence of the EU’s EBA initiative for LDCs. In terms of exports to the EU market, Bora et al. (2002) find that the highest similarity is found between African LDCs and African non-LDCs. Hence, it is to be expected that trade diversion as a result of the EU’s EBA is more serious for African non-LDCs than for other non-LDCs. Lewis et al. (2001) also find that the trade diversion effects on developing countries are determined by initial trade dependencies. At the sectoral level, Akinkugbe (2000) concludes that although the aggregate amount of trade diversion may not necessarily be large, on a product-by-product basis, individual product lines may be significantly affected. At the global level, most studies conclude that the trade creation generated by a given PTA outweighs the adverse trade diversion effects, and so countries suffering trade diversion could in principle be compensated, and the world would be better off as a whole (Harrison et al., 2003; Lewis et al., 2001). There are, however, other cases where trade diversion exceeds trade creation (Wolf, 2000). The studies also generally conclude that liberalization on a multilateral, MFN basis would eliminate the adverse trade diversion effects (Brown et al., 1997; Harrison et al., 2003).

322

C.P. Nielsen

Importance of sensitive products Several studies confirm the importance of sensitive products that are either excluded from trade liberalization or given long implementation periods. In the studies of the EU’s EBA, the results show that most of the potential gains accrue from the three sensitive products that are subject to lengthy transition periods. Investigating the sources of the welfare gains experienced by the LDCs, Yu and Jensen (2003) show that a substantial part stems from positive terms-of-trade effects, which arise because the LDCs gain preferential access to the highly protected EU agri-food market. Along similar lines, Harrison et al. (2003) explicitly consider the realistic scenario that the EU excludes agriculture from the EU–Mercosur agreement. The gains to Brazil would be reduced to just one ninth of the gains with full preferential market access to the EU. This is because it is precisely in these highly protected, ‘sensitive’ agricultural and food products that the Mercosur countries have a comparative advantage. Also related to the issue of sensitive products, Shapouri and Trueblood (2003) conclude that for the AGOA to have a substantial impact on the SubSaharan African (SSA) countries, tariff concessions must be given on processed agricultural products. Currently less than 10 per cent of SSA’s agrifood exports are processed. SSA countries could potentially increase these exports since they are producers of raw materials such as coffee and cocoa.

Structural issues in developing countries All the CGE analyses of the various agreements with the EU show a shift in resources from the manufacturing sectors to the agricultural and food sectors in the developing countries (Kerkelä et al., 2000; Ianchovichina et al., 2001; Bora et al., 2002; McDonald, 2002; Harrison et al., 2003). The reason for this, of course, is that the EU’s pre-PTA protection was against agricultural exports – and hence most of the action – takes place in precisely this sector. Consequently, Brown et al. (1997) warn that the developing countries may face significant adjustment problems in connection with the substantial intersectoral shifts in labour and capital that these PTAs would imply. McDonald (2002) rightly questions whether these countries are in a position to actually realize the required structural changes. Several of these countries depend heavily on preferential market access and are characterized by limited possibilities for export diversification. Based on a case-study analysis of three ACP countries, Stevens (1990) finds that the production and export structure of these countries have changed in favour of those agricultural commodities that are particularly strongly protected in the EU and other OECD markets. As privileged exporters to these protected markets, ACP exporters gain part of the economic rent embodied in the artificially high prices, thereby making these countries rather sensitive to removal of such preferences. If sufficiently detailed, a CGE analysis can supplement the analysis of sectoral shifts in production and factor allocation by identification of the

New Regionalism in the Aftermath of Cancún

323

distributional and poverty consequences of trade liberalization initiatives. One such detailed study is that of Harrison et al. (2003). Most of the trade policy options considered by this study – regional, multilateral and unilateral – show that there are progressive distribution gains to be had in the medium to long term, so that the poorest households experience the largest percentage increase in incomes. The reason for this is a shift in resources from capitalintensive manufacturing towards unskilled labour-intensive agriculture. Another type of adjustment that has to be made in developing countries participating in reciprocal trade agreements is adjustment to lost tariff revenue, which in a developing country is often a very important source of government revenue. In a study of a partial EU–ACP15 free trade agreement, Wolf (2000) finds that it is primarily the reduction in tariff revenue that causes the decline in real GDP experienced by the ACP countries. This is explained by the adverse effect that the lower tariff revenue has on government transfers to households and government spending, which in turn leads to an overall decline in demand. The reason is that the tariff reductions involved are asymmetrical because the ACP countries have high initial tariffs. Some studies explicitly simulate the use of other taxes to compensate for the lost tariff revenue. Rutherford et al. (1997), for example, find that the valueadded tax (VAT) rate would have to increase by 55–60 per cent in Morocco as a consequence of the EU–Morocco free trade agreement to compensate for lost tariff revenue. On imports this would imply an increase in the VAT from 11 per cent to about 16–17 per cent. On domestic products this would imply an increase in the VAT from 3 per cent to about 4–5 per cent. Wolf (2000) simulates the possibility of compensation by the developed country partner in her analysis of a partial EU–ACP free trade agreement to deal with the associated tariff revenue loss. She finds that such an aid transfer would help maintain domestic demand. Moreover, ACP imports from the EU would actually increase compared a situation without compensation, while EU imports from ACP would remain unchanged. Hence, part of the aid flows may be ‘returned’ through increased trade. There are a number of studies that investigate whether or not the developing countries are in a position to actually realize the structural changes that would be required in order to fully benefit from the PTAs. The results of Kerkelä et al. (2000) show that it makes a difference whether there is a pool of surplus labour in the economy or if the economy is operating at full employment. Brown et al. (1997) find that the static welfare gains for Tunisia of an FTA with the EU range from a minor loss (
0.2 per cent of GDP) to a gain of 3.3 per cent depending on whether capital is assumed to be intersectorally mobile or not.

Policy choices in developing countries In addition to structural constraints, the developing countries’ own policy regimes will affect the extent to which developing countries may benefit

324

C.P. Nielsen

from the PTAs in which they participate. The results of a growth function analysis by Clark (1997), for example, show that availability of capital, labour and infrastructure are important determinants of growth – as is previous growth performance. The latter is interpreted as an indication that government policies that encourage economic growth are an important prerequisite for taking advantage of PTA provisions, in this case those of the CBERA. This conclusion is consistent with the main message of a recent study by Schiff and Winters (2003) on regional integration and development, namely that the general policy stance is crucial for the degree of success with which developing countries can capture the long-term benefits of participating in a regional trade agreement. When considering the impact of Africa’s own trade policy regime on the region’s trade performance, Coe and Hoffmaister (1999) find that trade policies of the African countries are considerably less open than are policies of other developing countries, and that this has contributed to the relatively low levels of bilateral trade between African countries and developed countries. Hence, they conclude that there remains considerable scope for unilateral trade liberalization to spur trade between Africa and the developed countries, thereby increasing opportunities for technology transfers. Harrison et al. (2003) also explicitly consider the impact of unilateral trade reform, in this case in relation to Brazil’s participation in the EU–Mercosur agreement. They find that unilateral tariff cuts in Mercosur would yield significant benefits to Brazil. Despite these beneficial prospects of unilateral trade liberalization, how willing are developing countries to initiate such reforms? Özden and Reinhardt (2003) developed an econometric model to test the hypothesis that removing GSP preferences induces liberalization of the recipient country’s trade policy regime. Their results strongly support the hypothesis that GSP recipients are roughly 15–20 per cent more protectionist than countries no longer eligible. The results of this study have strong policy implications, suggesting that the efforts being made to enhance the ‘special and differential treatment’ of developing countries (such as through the GSP) within the multilateral trading system may be counterproductive for the establishment of a more liberalized trade regime. Özden and Reinhardt (2003) recommend that developing countries should give up GSP and instead move in the direction of reciprocal trade agreements. Yet, as discussed above, reciprocal agreements introduce new challenges (i.e. potentially significant losses of tariff revenue and adjustment costs) for those developing countries that have not been required to bring down trade barriers at the same pace and to the same extent as developed countries through the multilateral WTO process. Moreover, this is not an either/or question. There are, in fact, two examples where, in the formation of reciprocal agreements, the GSP was the point of initiation and was consolidated as a permanent structure of the agreements, in addition to the provision of lower tariffs on non-GSP products. The two cases are the US agreements with Chile in 2003 and with Central America and the Dominican Republic in 2004.

New Regionalism in the Aftermath of Cancún

325

Benefits and costs of developed country (agricultural) reform Some developing countries have become heavily dependent on preferential access to developed country markets. Through PTAs developing countries are able to exploit the high internal prices that are created by the developed countries’ domestic agricultural policies, such as the EU Common Agricultural Policy (CAP) and the US Farm Bill, and so they enjoy benefits over other low-cost producers in the rest of the world. In Chapter 13, Yu and Jensen explicitly considered the impact of reforms of preference-granting countries’ agricultural policies on preference-receiving countries. Their results justify the concerns that developing countries have about preference erosion. However, retaining the CAP in its current form in order to keep the LDC preferences meaningful is clearly not the most efficient way to proceed, since this entails a significant distortion of world agricultural trade. Yu and Jensen (Chapter 13) show that global welfare gains from EU reform (combining changes in the three areas of export subsidies, domestic support and marker access) would be almost US$15 bn.

Benefits and costs of multilateral liberalization Explicitly considering the relationship between regionalism and multilateralism, Harrison et al. (2003) find that the apparent Brazilian strategy of simultaneously pursuing regional trade liberalization and supporting multilateral liberalization within the WTO is an adequate one. The reason is, of course, that Brazil is more likely to find the most efficient world supplier in the combined economic areas and therefore reduce any trade diversion effects. Given the likely exclusion of agriculture in the EU–Mercosur agreement, their results show that multilateral trade liberalization will provide larger global welfare gains by a magnitude of more than four times that of any of the regional initiatives investigated in this study. Moreover, as mentioned earlier, unilateral trade policy reform would add to these gains. Harrison et al. (2003) thus conclude that Brazil will fare best if pursuing all avenues, i.e. regional, multilateral and unilateral trade liberalization. Developing countries are worried, however, that MFN liberalization will harm them as the value of their preferences are eroded. Indeed, the results of several of the studies reviewed here do warrant this concern. Wainio and Gibson (2003), for example, find that countries that are highly dependent on preferential access to the US market will suffer from trade diversion and thereby lose market shares as a consequence of MFN liberalization. Conversely, countries that do not currently enjoy preferential access will, of course, gain from an MFN liberalization. Generally, the forward-looking PE and CGE analyses suggest that erosion of preferences is something to worry about. What can be learned from past experience? Here the answer is more inconclusive. In their

326

C.P. Nielsen

gravity model analysis of Africa’s trade with developed countries, Coe and Hoffmaister (1999) find that ‘… Africa actually overtraded with the North relative to other developing countries in the early 1970s, but the degree of overtrading has steadily declined over the past 25 years’. The authors suggest that this is partly explained by the fact that Africa’s trade preferences in Europe have eroded over the past two decades. This result stands in contrast to that of Nilsson (2002), however. Also using a gravity model, he finds that the estimated gross trade creation effects of the GSP and the Lomé Convention have not been eroded as a result of the declining EU MFN tariffs in connection with multilateral trade liberalization. The author contends that the reason for this is that the late 1980s saw a reduction in the EU’s use of non-tariff barriers, e.g. voluntary export restraints and anti-dumping measures. In any case, it is a fact that national markets continue to be insulated from one another through a range of barriers other than tariffs and traditional trade instruments, e.g. tax policies, health and safety regulations, environmental standards, etc. Schiff and Winters (2003) assert that cooperation on domestic policies can substantially increase the gains from forming PTAs by bringing down these barriers, but at the same time they point out that it is not necessary to link policy integration to trade preferences in order to achieve these benefits. For instance, individual developing countries can e.g. adopt international standards and recognize regulatory norms of their major trading partners such as the EU and the USA. Moreover, Schiff and Winters (2003) suggest that the WTO can play a large role in facilitating such policy integration by, for example, adopting rules that extend the MFN principle to policy integration initiatives such as customs clearance documentation and procedures. Finally, concerning the relationship between regionalism and multilateralism, Schiff and Winters (2003) contend that the current WTO rules covering preferential trade agreements cannot be relied on to guarantee benign regionalism. That responsibility lies solely with the participating governments. The WTO can forbid some destructive forms of regionalism, but its main role is to act as an instrument for pursuing global liberalization on a MFN basis.

Conclusions Referring to the question posed in the title of this chapter, economic theory cannot provide clear-cut conclusions about the desirability of preferential trade agreements as seen from a developing country perspective and in light of the ongoing multilateral WTO negotiations. Based on the survey of recent empirical studies of selected PTAs, however, a number of preliminary conclusions can be drawn: ●

The welfare impact on participating countries is generally positive, yet small.

New Regionalism in the Aftermath of Cancún ● ● ● ● ●

327

Trade diversion can be an issue, particularly for specific sectors and countries. The greatest gains for developing countries lie in liberalizing trade in products that are sensitive for the preference-giving countries. Reciprocal agreements pose new challenges for developing countries, especially loss of tariff revenue. Multilateral liberalization is preferred, but erosion of preferences is an issue. OECD agricultural policy reform is also desirable, but erosion of preferences is again an issue.

Hence, drawing on the current theoretical knowledge and the available empirical evidence, the overall advice to be given to developing country policy-makers is that they should direct their negotiation resources toward achieving truly comprehensive trade liberalization at the multilateral level through the WTO. Notwithstanding the economic rationale behind this conclusion, the fact of the matter is that the present world trading system is characterized by a myriad of regional, preferential trade agreements that have been concluded based on political rather than economic objectives and that are nowhere near the textbook definitions of free trade agreements and customs unions. Nevertheless, both economic theory and quantitative empirical analyses can provide guidelines as to how real-world PTAs should be constructed so as to avoid the worst pitfalls and help reap the promises of trade liberalization in a second-best world. In particular, for developing countries to capture the long run gains from any trade agreement – bilateral, regional or multilateral – trade liberalization must be accompanied by a sound general policy stance.

Notes 1

In the literature reference is often made to the concepts of ‘preferential trade agreements’, ‘regional trade agreements’ and ‘regional integration agreements’. In this chapter the term ‘preferential trade agreement (PTA)’ will be used for the following reasons. First, many trade agreements are between countries that are not necessarily geographically contiguous, i.e. they are not necessarily ‘regional’ in a traditional sense. Second, the focus of this chapter is on trade, acknowledging of course the fact that many agreements go beyond commodity trade. They extend into areas such as trade in services, foreign direct investment, labour mobility, domestic policies, domestic regulation and national and international standards, among other things, thus explaining the use of the term ‘integration agreements’. The term PTA will be used to include trade agreements with differing degrees of preferential treatment granted to members relative to non-members – agreements that may be reciprocal or nonreciprocal. Notwithstanding the above considerations, the term regional trade agreement will be used interchangeably with the term PTA, since much of the discussion concerns the relationship between regionalism and multilateralism. 2 This and the following section draw on parts of a report by Nielsen (2003), which reviews more than 40 recent empirical studies analysing the impact of selected PTAs. 3 The terms ‘old’ and ‘new’ trade theory are commonly used in the literature and will

328

C.P. Nielsen therefore also be used in this chapter without making any prejudgement about the relevance and applicability of either strand of theory. 4 See e.g. Panagariya (1999) for a more comprehensive treatment of key theoretical contributions that address both old and new themes. 5 Procompetitive effects are thought to have operated in the case of European economic integration, but there is not yet empirical evidence that developing countries will be able to benefit strongly from such effects, the main reason being that developing countries have fewer goods for which product differentiation and economies of scale are important (Schiff and Winters, 2003). 6 There are very few empirical studies of the impact of PTAs on investment. Most agreements are so new that the data are simply non-existent. Furthermore, Schiff and Winters (2003) point out that the real key to increasing investment in developing countries is a sensible general policy, i.e. a sound macro policy, protection of property rights and the presence of efficient financial and banking sectors. 7 It was Bhagwati (1991) who originally introduced this terminology. 8 This delineation means that questions about which types of regional agreements are better for developing countries – North–South or South–South – are not addressed here. It is worth noting, however, that Schiff and Winters (2003) find that if a developing country is to pursue regionalism, it is better to sign up with a large, rich country than with a small, poor country. The main reason is that the former is a more efficient supplier of most goods and a source of greater competition for local producers. 9 This agreement is also known informally as the Caribbean Basin Initiative (CBI). 10 The AGOA and CBTPA are Titles 1 and 2, respectively, of the US Trade and Development Act of 2000. 11 The scope of this chapter is furthermore delineated by including only those empirical studies that are either economy-wide in scope or have a specific agricultural focus. Studies that highlight trade in manufactured goods will only be referred to if particularly relevant to the issues discussed here. Finally, the review generally limits itself to studies conducted within the past decade. 12 See Appendix B of Nielsen (2003) for summaries of the selected agreements and references for more information. 13 Gravity models have been subject to various criticisms (see Nielsen, 2003). 14 Quad: EU, USA, Canada and Japan. 15 It includes only part of the ACP countries included in the West African Economic and Monetary Union (UEMOA).

References Akinkugbe, O. (2000) The European Union and South Africa Free Trade Agreements. An explanatory partial equilibrium analysis of potential effect on ACP–EU trade flows. Journal of World Trade 34(3), 95–109. Andriamananjara, S. and Hillberry, R. (2001) Regionalism, Trade and Growth: The Case of the EU–South Africa Free Trade Arrangement. Office of Economics Working Paper No. 2001-07-A, US International Trade Commission (accessed from internet).

Audley, J. (2003) Bad Bilateral Trade Deals are no better than Bad Multilateral Deals. Trade, Equity and Development Issue Brief, Carnegie Endowment for International Peace (CEIP), Washington DC. Bhagwati, J.N. (1991) The World Trading System at Risk. Princeton University Press, Princeton, New Jersey. Bhagwati, J. (1995) US Trade Policy: The infatuation with Free Trade Areas. In: Bhagwati, J. and Krueger, A.O. (eds) The Dangerous

New Regionalism in the Aftermath of Cancún Drift to Preferential Trade Agreements. American Enterprise Institute for Public Policy Research, Washington DC. Bora, B., Cernat, L. and Turrini, A. (2002) Duty and Quota-Free Access for LDCs: Further Evidence from CGE Modelling. Policy Issues in International Trade and Commodities, Study Series No. 14, UNCTAD/ITCD/TB/15, UNCTAD, Geneva. Bridges (2003) Bridges Weekly Trade News Digest 7, No. 35, p. 4. Brown, D., Deardorff, A. and Stern, R. (1997) Some economic effects of the Free Trade Agreement between Tunisia and the European Union. In: Galal, A. and Hoekman, B. (eds) Regional Partners in Global Markets: Limits and Possibilities of the Euro-Med Agreements. Centre for Economic Policy Research and The Egyptian Centre for Economic Studies, London. Burfisher, M.E. and Jones, E.A. (1998) (eds) Regional Trade Agreements and US Agriculture. Economic Research Services, US Department of Agriculture Agricultural Economic Report No. 771, Washington DC. Burfisher, M.E., Robinson, S. and Thierfelder, K. (2003) Regionalism: Old and New, Theory and Practice. Paper presented to The International Agricultural Trade Research Consortium (IATRC) Conference, Capri, Italy. Chevassus-Lozza, E. and Gallezot, J. (2003) Preferential Agreements – Tariff Escalation: What are the Consequences of the Multilateral Negotiations for the Access of Developing Countries to the European Market? Paper presented at the International Conference on Agricultural Policy Reform and the WTO: Where are we heading? Capri, Italy. Clark, D.P. (1997) A diffusion model of the process of implementing the Caribbean Basin Economic Recovery Act. The Developing Economies 35, 185–195. Coe, D.T. and Hoffmaister, A.W. (1999) North–South Trade: Is Africa Unusual. Journal of African Economies 8, 228–256. Coppin, A. (1992a) Recent US economic policies and the Central Caribbean economies. The Review of Black Political Economy, 55–71.

329 Coppin, A. (1992b) Trade and Investment in the Caribbean Basin Since the CBI. Social and Economic Studies 41, 21–43. De Melo, J., Panagariya, A. and Rodrik, D. (1993) The new regionalism: a country perspective. In: De Melo, J. and Panagariya, A. (eds) New Dimensions in Regional Integration. Centre for Economic Policy Research, Cambridge University Press, Cambridge, UK. DeRosa, D.A. (1998) Regional Integration Arrangements: Static Economic Theory, Quantitative Findings, and Policy Guidelines. World Bank, Development Research Group, Policy Research Working Paper No. 2007, Washington DC. Dessus, S. and Suwa-Eisenmann, A. (1998) Trade Integration with Europe, Export Diversification and Economic Growth in Egypt. CEPR Discussion Paper No. 1938, Centre for Economic Policy Research, London. DeVault, J. (1996) Political pressure and the US generalized system of preferences. Eastern Economic Journal 22, 35–46. Ethier, W.J. (1998) The New Regionalism. The Economic Journal 108, 1149–1161. Freund, C. (2003) Reciprocity in Free Trade Agreements. Mimeo, World Bank, Washington DC. Grossman, G. and Helpman, E. (1995) The politics of Free Trade agreements. The American Economic Review 85, 667–690. Harmsen, R. and Leidy, M. (1995) Regional trading arrangements. In: Kirmani, N. et al. (eds) International trade Policies: Volume II: The Uruguay Round and Beyond: Background Papers. Series: World Economic and Financial Surveys, International Monetary Fund, Washington DC Harrison, G.W., Rutherford, T.F., Tarr, D.G. and Gurgel, A. (2003) Regional, Multilateral and unilateral Trade Policies of MERCOSUR for Growth and Poverty Reduction in Brazil. World Bank Policy Research Working Paper No. 3051, World Bank, Washington DC. Hoekman, B. and Djankov, S. (1997) Towards a Free Trade agreement with the European Union: issues and policy options for Egypt. In: Gala, A. and Hoekman, B. (eds)

330 Regional Partners in Global Markets: Limits and Possibilities of the Euro-Med Agreements. Centre for Economic Policy Research and The Egyptian Centre for Economic Studies, London. Hoekman, B. and Konan, D.E. (1998) Deep Integration, Nondiscrimination, and Euro–Mediterranean Free Trade. Paper presented at the Conference Regionalism in Europe: Geometries and Strategies After 2000, Bonn, Germany. Hoekman, B., Ng, F. and Olarreaga, M. (2002a) Reducing Agricultural Tariffs versus Domestic Support: What’s More Important for Developing Countries? Policy Research Working Paper No. 2918, Development Research Group, World Bank, Washington DC. Hoekman, B., Ng, F. and Olarreaga, M. (2002b) Eliminating excessive tariffs on exports of Least Developed Countries. The World Bank Economic Review 16, 1–21. Ianchovichina, E., Mattoo, A. and Olarreaga, M. (2001) Unrestricted market access for Sub-Saharan Africa: how much is it worth and who pays? Policy Research Working Paper No. 2595, Development Research Group, World Bank, Washington DC. Kemp, M.C. and Wan, H.Y. (1976) An elementary proposition concerning the formation of customs unions. Journal of International Economics 21, 99–110. Kerkelä, L., Niemi, J. and Vaittinen, R. (2000) Renegotiating the Lomé Convention – Trade Policy Schemes and Their Effects for African Regions. Working Paper W-261, Helsinki School of Economics and Business Administration, Helsink. Konan, D. and Maskus, K. (1997) A computable general equilibrium analysis of Egyptian trade liberalization scenarios. In: Galal, A. and Hoekman, B. (eds) Regional Partners in Global Markets: Limits and Possibilities of the Euro-Med Agreements. Centre for Economic Policy Research and The Egyptian Centre for Economic Studies, London. Krueger, A.O. (1999) Free Trade Agreements Versus Customs Unions. Working Paper No. 5084, National Bureau of Economic Research, Cambridge, Massachusetts. Lewis, J.D., Robinson, S. and Thierfelder, K. (2001) Free Trade Agreements and the

C.P. Nielsen SADC Economies. TMD Discussion Paper No. 80, Trade and Macroeconomics Division, International Food Policy Research Institute (IFPRI), Washington DC. MacPhee, C.R. and Oguledo, V.I. (1991) The trade effects of the US generalized system of preferences. Atlantic Economic Journal 19, 19–26. Martínez-Zarzoso, I. (2001) Trade in Distancesensitive Products: Explaining Mercosur Exports to the EU. Paper presented at the conference IV Encuentro de Economia Aplicada, Departament D’Economia, Universitat Rovira I Virgili, Tarragona, Spain. Martínez-Zarzoso, I. and Nowak-Lehmann, F. (2000) Augmented Gravity Model: an Empirical Application to MercosurEuropean Union Trade Flows. IberoAmerican Institute for Economic Research Discussion Paper, Berlin. Mattoo, A., Roy, D. and Subramanian, A. (2002) The Africa Growth and Opportunity Act and Its Rules of Origin: Generosity Undermined? IMF Working Paper No. 158, International Monetary Fund, Washington DC. McDonald, S. (2002) The commodity protocols of the Lomé Convention: the case of beef exports from Botswana. Journal of Agricultural Economics 53, 407–427. McDonald, S. and Walmsley, T. (2003) Preferential Trade Agreements and the Optimal Liberalization of Agricultural Trade. Paper presented to the 6th Annual Conference on Global Economic Analysis, Scheveningen, The Hague, The Netherlands. Meade, J.E. (1955) The Theory of Customs Union. North Holland, Amsterdam. Nielsen, C.P. (2003) Regional and Preferential Trade Agreements: A Literature Review and Identification of Future Steps. (Commissioned by the OECD); Report no. 155, Danish Research Institute of Food Economics (FØI), Copenhagen. Nilsson, L. (2002) Trading relations: is the roadmap from Lomé to Cotonou correct? Applied Economics 34, 439–452. Nouve, K. and Staatz, J. (2003) Has AGOA Increased Agricultural Exports from SubSaharan Africa to the United States? Contributed paper presented at the

New Regionalism in the Aftermath of Cancún International Agricultural Trade Research Consortium (IATRC) Conference, Capri, Italy. Oguledo, V.I. and MacPhee, C.R. (1994) Gravity models: a reformulation and an application to discriminatory trade arrangements. Applied Economics 26, 107–120. Özden, Ç. and Reinhardt, E. (2003) The Perversity of Preferences: GSP and Developing Country Trade Policies, 1976–2000. Working Paper No. 2955, World Bank, Washington DC. Panagariya, A. (1999) Preferential Trade Liberalization: The Traditional Theory and New Developments. Technical Paper No. 6, Department of Economics, University of Maryland. http://www.bsos.umd.edu/econ/ p a n a g a r i y a / a p e c o n / t e c h p a p e r. h t m (accessed May 2003). Raboy, D.G., Simpson, T.L. and Xu, B. (1994) A transition proposal for Lomé Convention trade preferences: the case of the EU banana regime. World Economy, 565–581. Rutherford, T.F., Rutström, E. and Tarr, D. (1997) Morocco’s free trade agreement with the EU: a quantitative assessment. Economic Modelling 14, 237–269. Schiff, M. and Winters, L.A. (2003) Regional Integration and Development. World Bank, Washington DC. Shapouri, S. and Trueblood, M. (2003) The African Growth and Opportunity Act (AGOA): Does It Really Present Opportunities? Paper presented to The International Agricultural Trade Research Consortium (IATRC) Conference, Capri, Italy. Stevens, C. (1990) ACP Export Diversification: Jamaica, Kenya and Ethiopia. ODI Working Paper No. 40, Overseas Development Institute, London. Stoeckel, A. and Borrell, B. (2001) Preferential

331 Trade and Developing Countries: Bad Aid, Bad Trade, Rural Industries. Research & Development Corporation (RIRDC), Publication No. 01/116, Canberra, Australia. Teixeira, E., Cypriano, L. and Pinto, W. (2002) Impacts of AFTA and MERCOEURO on Agribusiness in the MERCOSUL Countries. Paper presented at the 5th Annual Conference on Global Economic Analysis, Taipei, Taiwan. Trueblood, M. and Somwaru, A. (2002) Trade Liberalization and the Least Developed Countries: Modeling the EU’s Everything But Arms Initiative. Paper presented at the 5th Annual Conference on Global Economic Analysis, Taipei, Taiwan. Truett, D.B. and Truett, L.J. (1993) Trade preferences and exports of manufactures. Weltwirtschaftliches Archiv 129, 573–590. Viner, J. (1950) The Customs Union Issue. Stevens, New York. Wainio, J. and Gibson, P. (2003) The Significance of Nonreciprocal Trade Preferences for Developing Countries. Paper presented to The International Agricultural Trade Research Consortium (IATRC) Conference, Capri, Italy. Wolf, S. (2000) From Preferences to Reciprocity: General Equilibrium Effects of a Free Trade Area Between the EU and the UEMOA. Paper presented at the Third Annual Conference on Global Economic Analysis, Monash University, Australia. Yu, W. and Jensen, T.V. (2003) Tariff Preferences, WTO Negotiations and the LDCs: the case of the Everything But Arms Initiative of the EU. Paper presented to the 6th Annual Conference on Global Economic Analysis, Scheveningen, The Hague, The Netherlands.

This page intentionally left blank

Index

Page numbers in italics refer to tables and figures access to food 165 to markets 228–229, 231–232 under Everything But Arms Initiative 296–300 in Harbinson Proposal 143–144, 145, 146, 147 Africa modelling of trade with EU-25 basis of model 66–68 EU exports to North Africa 68–69 EU imports from rest of Africa 69–71 impact on African agricultural production 71–73 welfare implications for Africa 73–74, 75 national trade policies 324 positions in Doha Round 137 cotton 97 sub-Saharan countries calories per capita vs. GDP 189 FAO food stock recommendations 198, 199 food consumption data 190–191 impact of potential EU ban on GM crops 270 model of consumer preferences re GM crops 274–275 negotiating strategies 8 probability of food deficit 191–196

Africa, Caribbean and Pacific (ACP) preferences (EU) 284, 323 Agenda 2000 see under Common Agricultural Policy Agreement on Sanitary and Phytosanitary Measures 266 Agreement on Subsidies and Countervailing Measures (GATT) 251 Agreement on Technical Barriers to Trade 266 agreements, preferential trade aftermath of Cancún meeting 310–311 conditions for contribution to free trade 314 developing countries policy choices 322–323 structural issues 322–323 dynamic gains from trade 320 empirical impact studies 316–318, 319 methodologies 319 impact of agricultural reform 325 impacts on participants 320–321 impacts on developing country non-participants 321 and multilateral liberalization 326 overall impact 318–320 sensitive products 322 theoretical considerations 311–313 333

334

Index aid food aid in Harbinson Proposal 148 proposals of developing countries 233–234 Albania 176–177 Amber Box 96, 113 in Harbinson Proposal 149–150 Asia 4, 5, 8 models of GM technology impacts 274–276 see also China; Japan

Barbados 180 biosafety: Cartagena Protocol 265–266, 267 Blue Box 95, 96, 99, 113 changes in WTO member positions 127 in Harbinson Proposal 125, 149 boxes Amber Box 96, 113, 149–150 Blue Box 95, 96, 99, 113, 125, 127, 149 criticized by developing countries 207 Development Box proposals 126, 208, 210 for developed countries 230–232 for developing countries 226–230 general issues 232–234 Development/Food Security Box vs. Green Box 211–220 Food Security Box proposals 208 Green Box 113, 125, 127, 148–149, 253 Harbinson Proposal 125, 148–150 WTO member positions 127 butter see milk and dairy products

Cairns Group creation of G-20 12 negotiating strategies 6, 8 Canada exports and imports 4 labelling regulations 265 Nominal Assistance Coefficient (OECD) 5 position in Doha negotiations 135 Cancún meeting, September 2003 96–97

Cape Verde 178, 180 Caribbean see Latin America and Caribbean Cartagena Protocol on Biosafety 265–266, 267 cereals costs of maintaining stocks 200–201 estimate of minimum safe level of world stocks 186–188 EU market regime 47, 288 modelling of effects 64 impact of GM technology 269–274 see also wheat CGE models 267–268, 318 China food security 177–178 impact of GM technology 268–269 positions in Doha negotiations 125, 136 Club of Rome, 1972 report 186 cluster analysis identification of potential coalitions 128, 130–135 methodology 129–130, 164–166 insecure countries 171–172, 176–177 neutral countries 177–178 secure countries 178–179 typology of countries 168–171, 173–175 Codex Alimentarius 266–267 Cold War 243 Committee on World Food Security 186 commodities: prices 233 Common Agricultural Policy 243 Agenda 2000 36–37, 54 budgetary expenditures 42 intervention prices for wheat 200 market policy 40–42, 44, 46 acreage support 47 crops and associated products 47, 49–51 dairy products 51–52 level and composition of support 48 meat 52–53 McSharry reform 35, 36 mid-term review and 2003 reform 37–39, 54–55, 58–59 modelling assumptions 62

Index

335 modelling of effect on EU-25 production 63–66 objectives 34–35 primary aim 186 rural development policy 42–44, 55 main objectives 45 competition, export 144, 148, 149 conservation: US programmes 91–93 Conservation Reserve Program (USA) 92 Conservation Security Program (USA) 92, 93 consumers: expression of preferences 271–272, 274–276 consumption, food correlation with GDP 189 FAO database 190–191 indicators 164 modelled effect of protection vs. R&D investment 217 corruption 245–246 Cotonou Agreement 284 cotton Doha Round negotiations 97 impact of GM technology 268 credits export credits in Harbinson Proposal 148 role in alleviating food insecurity 201 Cuba: position in Doha negotiations 136–137

dairy products see milk and dairy products decoupling 2003 CAP reform 38–39, 59 modelling assumptions 62, 109, 112 Derbez Text (2003) 12–13 developed countries agricultural and industrial value 247 agricultural history 242–243 Development Box policy proposals 230–232 food security 181 developing countries agricultural and industrial value 247 criticism of boxes 207

Development Box policy proposals 226–230 domestic policy considerations 28–29 effects of liberalization on income 247, 248 GM products 263, 276, 277 and Harbinson Proposal 137–138 inadequacy of WTO categories 179 industry 238–242 modelling of policy reform scenarios 114–117 negative effects of protection 220–222 not helped by legal exemptions 210 preferential trade agreements policy choices 323–324 structural issues 322–323 proposals for Development Box 126 proposals on export subsidies 229–230 WTO categories 163 see also Africa development, rural (EU) 42–44, 45, 55 Development Box 126, 208 comparison with Green Box 211–220 general proposals 232–234 proposals for developed countries 230–232 proposals for developing countries 226–230 and Special Products concept 210 distributions, statistical 192–196 diversion, trade 312 Doha Round negotiations see also Harbinson Proposal African cotton case 97 EU proposals 126 isolation of 139 failure to achieve consensus 95–96 G-20 proposal 97 July 2004 framework agreement 98–100 lack of convergence 127 movements in members’ positions 126–127 negotiation issues 124 potential coalitions: identification through cluster analysis methodology 129–130

336

Index Doha Round negotiations continued rationale 128 results 130–135 proposals for Development Box 126 scoring of members’ positions 124–125 US-EU proposal 96–97 US proposals 81, 94–95, 126 Dominican Republic 172

Eastern Enlargement (of EU): modelling of scenarios derived impact for Africa 71–76 experimental design 60–63 results for EU-25 trade with Africa 66–71 results for production in EU-25 63–66 Ecuador 177 Egypt 319–320 employment: modelled effect of protection vs. R&D investment 216 Environmental Quality Incentives Program (USA) 92, 93 Europe: agricultural policy concerns 243 European Agricultural Guidance and Guarantee Fund (EAGGF) 42, 43 European Union 2 Africa, Caribbean and Pacific (ACP) preferences 284 Eastern Enlargement see Eastern Enlargement Everything But Arms initiative see Everything But Arms initiative (EU) export subsidy rates 108 exports and imports 4–5 Generalized System of Preferences (GSP) 284–285 import-weighted tariffs 107 labelling regulations 265 model of GM crop ban 270–271, 272–273 Nominal Assistance Coefficient (OECD) 5 positions in Doha negotiations see Doha Round negotiations regulations on GMOs 264–265

Uruguay Round negotiations 36 use of PSE data in models 110 see also Common Agricultural Policy Everything But Arms initiative (EU) 70–71 concessions offered 286, 288–289 countries and products covered 286, 287 impact assessment domestic support reforms 302–303 effects on trade and production 291–293 elimination of export subsidies 300–301 macroeconomic effects 293–294 market access reform 296–300 methodology and data 289–290 policy implications 305–306 policy scenarios and results 290–291 role of sensitive products 294–295 welfare effects 293–294, 299, 303–304 safeguards 289 significant step to fulfilling Doha Development Agenda 282 exponential distribution 195–196 Export Enhancement Program (USA) 93 export subsidy rates 108 exports effect of protection 219 Harbinson proposals on competition 144, 148, 149 from least developed countries to EU 288 effect of Everything But Arms initiative 291–293 externalities 236–238, 244–245

FAIR Act, 1996 (USA) 82–83 comparison with 2002 Farm Bill 85 government payments to farmers, 1996–2000 84 groundnut programme 89 Farm Security and Rural Investment Act (USA, 2002)

Index

337 comparison with 1996 Farm Bill 85 conservation programmes 92–93 criticisms 80–81 dairy programme 88–89 groundnut programme 89–91 limitations on payments 86, 88 sugar programme 87–88, 91 target prices 85, 86 three-tiered support structure 84 trade provisions 93–94 and WTO considerations 86, 93 farms: size and concentration 4 Food and Agricultural Organization (FAO) estimation of safe level of cereal stocks 186–188 food consumption database 190–191 opinion on GM crops 278 recommended food stock policies 197–199 alternative strategy of finance not stocks 201–204 evaluation from statistical analysis 199 food security 18–21 inadequacy of WTO categories 179–181 indicators at national level cluster analysis methodology 166–168 definitions 164–166 insecure countries 171–172, 176–177 neutral countries 177–178 secure countries 178–179 typology of countries 168–171, 173–175 modelled effect of protection vs. R&D investment 216–218 prime political concern 185–186 rationale for international financial fund 201–204 see also stocks, food Food Security Box 208 fruit: EU market regime 51, 288

G-20 creation 12 Doha Round proposals 97, 126 GATT: different treatment of agriculture

and industry 249–253 GDP (gross domestic product) impact of Harbinson Proposal 155–159 modelled effect of protection vs. R&D investment 215 and per capital calories consumption 189 Generalized System of Preferences (GSP, EU) 284–285, 319, 324 genetically modified products see GM (genetically modified) products global computable general equilibrium (CGE) models 267–268 Global Trade Analysis Project (GTAP) model and database 59–60, 104–107, 268, 290 application to enlarged EU derived impact on Africa 71–76 experimental design 60–63 results for EU-25 trade with Africa 66–71 results for production in EU-25 63–66 exogenous and endogenous agricultural programmes 109–114 policy reform scenarios 114–117 GM (genetically modified) products 22–23 arguments for and against 262 benefits for developing countries 263 different perceptions of GM foods 265 extent of production 262–263 objectives 263 potential impact and implications see under models regulations: international 265–267 regulations: national 264–265 Green Box 113, 253 changes in WTO member positions 127 comparison with Development/Food Security Box 211–220 in Harbinson Proposal 125, 148–149 Green Revolution 240 groundnuts: US provisions 89–91

338

Index Harbinson Proposal 12–13, 17, 209 appraisal based on members’ positions 150–152, 153 domestic support 148–150, 151, 152 export competition 144, 148, 149 market access 143–144, 145, 146, 147 model impact on GDP and welfare 155–159 impact on trade 160 scenario and assumptions 153–155 positions 125, 131 in respect to central players 135–137

Import Substitution Industrialization 239, 241–242, 254 imports effect of protection on food imports/total exports ratio 217, 218 EU import barriers 41 and food security 180 income, distribution of 246, 247, 248 India food security 176, 177–178 position in Doha negotiations 126–127, 136–137 industry developing countries 238–242 treated differently to agriculture in GATT and WTO 238–242 International Fund for Agricultural Development (IFAD) 186 investment: effect of prices 26

Japan agricultural policy concerns 243 exports and imports 4 import-weighted tariffs 107 Nominal Assistance Coefficient (OECD) 5 use of PSE data in models 110 July 2004 framework agreement (Doha Round) 98–100 July Package (2004) 13

Korea, Republic of exports and imports 4 Nominal Assistance Coefficient (OECD) 5 kurtosis 193 Kyrgyzstan 177

labelling country of origin indication 94 for GM products 265 Latin America and Caribbean: negotiating strategies 8 legumes: EU market regime 49 liberalization 314 and erosion of tariff revenues 138 and global welfare 155–159 impact on income distribution 247, 248 negative welfare effects 25–27 and preference erosion 137 Lomé Convention 284

maize: impact of GM technology 269–274 Maldives 178, 180 Malthus, Thomas Robert (1766–1834) 186 markets: access to see under access Marrakesh Ministerial Decision 306 McSharry reform see under Common Agricultural Policy meat EU market regime 52–53 modelling of effects 64 milk and dairy products EU market regime 51–52 modelling of effects 64 US provisions 88–89 models 14, 15–16 Development/Food Security Box vs. Green Box basis of model 211–213 implications 220–222 results 214–220 simulations 213–214 of enlarged EU derived impact on Africa 71–76 EU-25 production 63–66

Index

339 EU-25 trade with Africa 66–71 experimental design 60–63 Everything But Arms Initiative: impact assessment domestic support reforms 302–303 effects on trade and production 291–293 macroeconomic effects 293–294 market access reform 296–300 methodology and data 289–290 policy implications 305–306 policy scenarios and results 290–291 role of sensitive products 294–295 welfare effects 293–294, 299, 303–304 global computable general equilibrium (CGE) models 267–268 GTAP see Global Trade Analysis Project (GTAP) model and database of Harbinson Proposal impact on GDP and welfare 155–159 impact on trade 160 scenario and assumptions 153–155 impact of GM technology China 268–269 consumer preferences 271–272, 274–276 implications 276–277 maize and soybeans 269–274 of OECD policy impact on developing countries agricultural domestic support 107–109 exogenous and endogenous agricultural programmes 109–114 policy reform scenarios 114–117 trade and agricultural programmes 104–107 of preferential trade agreements 318 use of PSE data 110 modernization: of agriculture 81–82 modulation 39, 59 Morocco 323 multifunctionality 21–22

based on idea of joint production 244 economic framework 236–238 European Union 42–44 of industry 238–239 justifies separate treatment for agriculture 235–236 a separate policy issue: pros and cons 244–246 multilateralism vs. regionalism 313–315, 325–326 multinationals 241 Myanmar 178, 180 Namibia 176 nationalism 238 Nominal Assistance Coefficient (OECD) 5 normality test 193 Norway: position in Doha negotiations 136 OECD modelling policy impact on developing countries agricultural domestic support 107–109 exogenous and endogenous agricultural programmes 109–114 policy reform scenarios 114–117 trade and agricultural programmes 104–107 Nominal Assistance Coefficient 5 Producer Support Estimate 107, 109 mapping of data in model 110–111 oil, olive: EU market regime 50 oilseeds EU market regime 49 impact of GM technology 269–274

Philippines 176 Poisson distribution 192–193 power, political 185–186 preferences 23–25 erosion 137 preferential trade agreements see agreements, preferential trade

340

Index prices commodities 233 effect on investment 26 EU market price support 40, 41 probability, statistical 191–196 Producer Support Estimate (OECD) 107, 109 product tracing 266 production, agricultural modelling of EU-25 production 63–66 regional comparison 10 productivity, total factor 106 products, primary 249 protection impact on preferential trade agreement partners 314 model of effects basis of model 211–213 implications 220–222 results 214–220 simulations 213–214

quotas, tariff rate 144

Regional Economic Partnership Agreements 284 regionalism 311 vs. multilateralism 313–315, 325–326 regulations: for GM products see under GM (genetically modified) products Republican Party (USA) 83 rice EU market regime 288–289 impact of GM technology 268–269 Rome, Ancient 185–186

security, food see food security seeds, grass 50 Sensitive Products concept 209 setaside: of farmland 47 skewness 193 South America: model of consumer preferences re GM crops 274–275 South Asia: negotiating strategies 8

soybeans: impact of GM technology 269–274 Special Products (SP) concept 209, 210 Special Safeguard Mechanism (SSM) 209, 252 non-agricultural goods 249 stocks, food costs of upkeep and maintenance 200–201 estimation for sub-Saharan Africa data on food consumption 190–191 probability of food deficit 191–196 rationale 188–189 size of stocks required alternative strategy of finance not stocks 201–204 FAO recommendations 186–188, 197–199 from statistical analysis 196–197 statistical evaluation of FAO recommendations 199 statistical analysis of critical deficits 189–190 strategies, negotiating 5–6, 8 stresses, biotic 263 structure, agricultural: regional comparison 9 sub-Saharan Africa see under Africa subsidies: differences between agricultural and industrial products 250–253 sugar EU market regime 49–50, 286, 288 US provisions 87–88, 91 Swaziland 177, 231–232

tariffs erosion of revenue 138 Harbinson proposals for reduction 143–144 import-weighted 107 terms-of-trade effects 312–313 tobacco: EU market regime 51 total factor productivity 106 traceability 266 trade impact of Harbinson Proposal 160

Index

341 impact of negotiations on food security 165 regional comparisons 4, 11 theoretical considerations 311–313 Trinidad and Tobago 177, 180

urbanization 238–239, 241 Uruguay Round negotiations Agreement on Agriculture 7(box), 252–253 compromise by USA and EU 36 unfinished business 27 and world food stocks 187 non-agricultural goods 250 USA 2–3 2002 Farm Bill see Farm Security and Rural Investment Act (USA, 2002) agricultural policy concerns 242–243 and the Codex Alimentarius 266 conservation programmes 91–93 cotton programme 97 Doha Round negotiations see under Doha Round negotiations export subsidy rates 108 exports and imports 4 FAIR Act, 1996 82–83, 89 government payments to farmers, 1996–2000 84 import-weighted tariffs 107 labelling regulations 265 modernization of agriculture 81–82

Nominal Assistance Coefficient (OECD) 5 suit against EU moratorium on GMOs 265 Uruguay Round negotiations 36 use of PSE data in models 110 Uzbekistan 177

value: agricultural and industrial 247 vegetables: EU market regime 51, 288 Venezuela 180 Vietnam 176

Ward’s method: cluster analysis 129 welfare, global effect of Everything But Arms Initiative 293–294, 299, 303–304 impact of agricultural policies 245 impact of GM technology 270–271 impact of Harbinson Proposal 155–159 modelled effect of protection vs. R&D investment 214–216 wheat CAP intervention prices 200 EU stocks 196, 196 FAO stock policy recommendations 198, 199 evaluation from statistical analysis 199 wine: EU market regime 50 World Food Conference, 1974 186